Podcasts about UI

Georgi Petrov, CEO and co-founder of Uxify, joins Karl and Taylor to talk about the half of web performance nobody optimizes for: the shopper's device. After building SMS Bump (acquired by Yotpo), ConsentMo, and NitroPack (acquired by WP Engine), Georgi is now tackling device-specific friction, the laggy buttons, layout shifts, and interaction delays that vary wildly from one phone to the next. We get into how Uxify profiles a visitor's capability without touching your code, the specialized agents that fix problems on the fly, the ClickHouse-powered data layer behind it all, and how to think about performance heading into BFCM. Plus the Dev Changelog and Picks of the Week.SponsorThe Support Heroes: https://www.thesupportheroes.com/Subscribe to Liquid WeeklyDon't miss out on expert insights and tips. Subscribe to Liquid Weekly for more content like this: https://liquidweekly.com/Find Georgi OnlineUxify: https://uxify.com/integrations/shopify/LinkedIn: https://www.linkedin.com/in/georgipetrov/Timestamps(00:00) Open(00:29) Intros(02:11) Meet Georgi Petrov of Uxify(02:54) Origin story: hardware hacking to his first software(06:03) Finding Shopify: SMS Bump, ConsentMo, and NitroPack(12:23) Making fast decisions and failing cheaply(15:32) Why Uxify: performance is half website, half device(20:11) How it works: profiling devices and adapting on the fly(26:17) Sitting alongside your theme without touching your code(28:26) Real user monitoring and specialized agents(37:47) The tech stack: ClickHouse and heavy browser testing(41:57) Shopify cart variations and a clever checkout fallback(46:24) Prepping for BFCM and who Uxify is for(51:14) Dev Changelog(58:22) Picks of the WeekDev Changelog[action required] Variants now support multiple barcodes: https://shopify.dev/changelog/product-variant-barcode-is-being-replaced-by-barcodesCreate and delete dev stores in Shopify CLI: https://shopify.dev/changelog/create-and-delete-dev-stores-in-shopify-cli[action required] Discounts Allocator Function API developer preview has ended: https://shopify.dev/changelog/discounts-allocator-function-api-developer-preview-has-endedDelivery options in Shopify Functions input now expose their metafields: https://shopify.dev/changelog/delivery-options-in-shopify-functions-input-now-expose-their-metafieldsPrepare your app for the Shopify admin's new look: https://shopify.dev/changelog/prepare-your-app-for-the-shopify-admins-new-look[action required] Updates to Events payloads and subscription configuration: https://shopify.dev/changelog/updates-to-events-payloads-and-subscription-configurationNew Pricing audit trail for contextual product variant prices: https://shopify.dev/changelog/pricing-audit-trail-for-contextual-product-variant-prices[breaking api change] [action required] Removed session.currentSession.staffMemberId from POS UI Extensions (2026-10): https://shopify.dev/changelog/removed-session-currentsession-staffmemberid-from-pos-ui-extensions-2026-10Request a bundle size exception for existing UI extensions: https://shopify.dev/changelog/request-a-bundle-size-exception-for-existing-ui-extensionsPicks of the WeekKarl: Keto Oreo-style sandwich cookies (about 50 calories each), part of his ongoing tour through packaged keto snacks: https://www.choczero.com/products/keto-sandwich-cookiesGeorgi: Calisthenics between AI prompts. Knocking out push-ups and squats while waiting on an agent, which he says burns the time well and sharpens his focus when he comes back.Taylor: Ted Lasso, season 4 on Apple TV+. Still, in his words, one of the most wholesome shows around: https://www.imdb.com/title/tt10986410/

#Localization #Globalization #ServiceNow #AI A translation can be word-for-word accurate and still be completely wrong. In this episode, Bobby Brill sits down with Sam Smyth, Product Manager on the globalization side at ServiceNow, to get into what happens after you decide to go multilingual — the tools, the governance, and the quiet work nobody sees until it breaks. Sam spent fifteen years in localization operations and management before joining ServiceNow, and he's blunt about what actually goes wrong: it's rarely the translator. It's the content you sent them. We get into Localization Workspace, the new globalization agent that builds a glossary from your knowledge articles in minutes, why every customer still insists on a human in the loop, and the one mistake that will break your product UI in six languages at once. Past Localization episodes: https://youtu.be/4ka4qLk8Uj0 https://youtu.be/cGXCnABXKow Guest: Sam Smyth, Product Manager - ServiceNow Chapters 00:00 "People don't see localization until it goes wrong" 00:22 Translation vs. localization — a quick recap 00:56 Meet Sam Smyth: fifteen years in localization 01:37 When a perfect translation is still wrong 03:34 Why customers don't want to leave the platform 04:07 Localization Workspace: no more spreadsheets and email chains 05:13 The surprise: everybody still wants a human in the loop 05:38 Speed, cost, quality — picking all three 06:55 Inside the globalization agent 08:50 Who approves the terms? (And which shade of green?) 09:55 Language governance and the invisible costs 11:49 AI, volume, and doing the work upstream 13:02 Where to actually start with a glossary 16:28 A day in the life of a localization program manager 18:56 Shifting left: from firefighting to strategy 20:35 The one mistake: never hardcode your UI 22:10 Wrap-up ServiceNow Training and Certification: https://www.servicenow.com/services/training-and-certification.html ServiceNow Insights podcast playlist: https://www.youtube.com/playlist?list=PLCOmiTb5WX3qvGq7Cp3o2KkCiplJyqQOK Servicenow.com: https://www.servicenow.comSee omnystudio.com/listener for privacy information.

#Localization #Globalization #ServiceNow #AI A translation can be word-for-word accurate and still be completely wrong. In this episode, Bobby Brill sits down with Sam Smyth, Product Manager on the globalization side at ServiceNow, to get into what happens after you decide to go multilingual — the tools, the governance, and the quiet work nobody sees until it breaks. Sam spent fifteen years in localization operations and management before joining ServiceNow, and he's blunt about what actually goes wrong: it's rarely the translator. It's the content you sent them. We get into Localization Workspace, the new globalization agent that builds a glossary from your knowledge articles in minutes, why every customer still insists on a human in the loop, and the one mistake that will break your product UI in six languages at once. Past Localization episodes: https://youtu.be/4ka4qLk8Uj0 https://youtu.be/cGXCnABXKow Guest: Sam Smyth, Product Manager - ServiceNow Chapters 00:00 "People don't see localization until it goes wrong" 00:22 Translation vs. localization — a quick recap 00:56 Meet Sam Smyth: fifteen years in localization 01:37 When a perfect translation is still wrong 03:34 Why customers don't want to leave the platform 04:07 Localization Workspace: no more spreadsheets and email chains 05:13 The surprise: everybody still wants a human in the loop 05:38 Speed, cost, quality — picking all three 06:55 Inside the globalization agent 08:50 Who approves the terms? (And which shade of green?) 09:55 Language governance and the invisible costs 11:49 AI, volume, and doing the work upstream 13:02 Where to actually start with a glossary 16:28 A day in the life of a localization program manager 18:56 Shifting left: from firefighting to strategy 20:35 The one mistake: never hardcode your UI 22:10 Wrap-up ServiceNow Training and Certification: https://www.servicenow.com/services/training-and-certification.html ServiceNow Insights podcast playlist: https://www.youtube.com/playlist?list=PLCOmiTb5WX3qvGq7Cp3o2KkCiplJyqQOK Servicenow.com: https://www.servicenow.comSee omnystudio.com/listener for privacy information.

This is also the first time we've released a podcast in full video. So even if you're typically an audio podcast listener, we highly recommend you watch the full video on Youtube (click the link above). Let us know if you enjoyed our live show, and if you'd like to attend one in the future. "Don't override your players' experience with your version of the story." Emotional, hilarious, and... sexy?! Our very first REPOD Live Show was full of unexpected turns involving McDonald's drive thrus, a romance that blossomed amidst the most terrifying experience I've ever played, and some truly memorable moments. Oh, and it also had more references to sex than we've ever had on twelve seasons of this show. I guess that's what happens when you're recording live. We've been wanting to take Reality Escape Pod live for a long time. What better place to finally do it than at RECON, our own escape room convention? For this experiment, we needed the right guest. Gijs Geers of DarkPark Escape Rooms in the Netherlands was perfect for this live show. Gijs is an escape room creator, owner, and former magician who has experienced plenty of the industry's highs and lows. What I appreciated most, though, was how candid Gijs was willing to be. He wasn't banging the drum of his self-importance. He got real with us. At times, he even became emotional, particularly when talking about the excitement and anxiety the TERPECAs bring. It made for a surprisingly personal conversation, especially considering it was all happening live on the RECON stage.   Full Show Notes   Episode Sponsors We are immensely grateful to our sponsors this season: Buzzshot Bookings, Ubisoft VR Escape Games, Immersif, and Patreon supporters like you. We truly appreciate your support of our mission to promote and improve the immersive gaming community.   Buzzshot Buzzshot is proud to announce their new "all-in-one" Booking System designed from the ground up for one type of business, escape rooms. Offering: Easy to use booking UI with no fees or hidden charges. Customer waiver capture Branded Team Photos Player Messaging Review Managment Automations Customer Surveys Broadcast Emails Special Offer for REPOD Listeners: REPOD listeners get an extended 21-day free trial plus 20% off your first 3 months, with no set-up fees or hidden charges. FREE Trial Account with full features* including SMS messaging, and 'Buzzshot Bookings' trial upon request. Visit buzzshot.com/repod to learn more about this exclusive offer.   Ubisoft VR Escape Games Harness the power of VR Escape games to maximize your venue's potential. Deliver blockbuster adventures and impossible experiences, all brought to life in Ubisoft's most popular worlds, including Assassin's Creed, Prince of Persia, and more. Location based games optimized for entertainment venues Exclusively available only in escape room venues, VR arcades, and theme parks. Games can accommodate anywhere from 2-8 players No start up fees; you only pay when players play Special Offer for REPOD Listeners: REPOD listeners get an extended free trial for an entire month of royalty-free operation. To hear more and claim your free offer, send a message via their contact page and be sure to use our code REPODxUBI26. Enter the code into the "description" field on the contact form.   Immersif Business Consulting Are you looking to unlock more revenue and streamline efficiency? Sam Wai is helping creatives and suppliers in the immersive industry earn more with specialist accounting and mentorship. Offering: Business Mentoring and Consulting: Growth focused Strategic Planning Actionable Advice Performance Intelligence Reports: Analysis of your game booking performance Learn how to optimize your marketing Customize your business based on player habits and booking preferences Special Offer for REPOD Listeners: REPOD listeners get a free one hour Business Mentoring Call. This is a focused one-to-one session to help you work through a challenge, make a difficult decision or bring more structure to your next stage of growth. We'll explore what's holding the business back, weigh up your options, and agree on practical next steps. You'll leave with clearer thinking, renewed confidence and a written summary to keep you moving forward. Book through immersif.co/REPOD or email Sam hello@immersif.co and be sure to mention REPOD to claim your free business consultation.   Support Us On Patreon Today Love escape rooms as much as we do? At Room Escape Artist, we've been analyzing, reviewing, and exploring the world of immersive games since 2014. We help players find the best experiences, and push the industry forward with well-researched, rational, and reasonably humorous escape room and immersive gaming content and events. By becoming a Patreon supporter, you're not just backing a blog — you're fueling a mission to make the escape room and immersive gaming community stronger, more thoughtful, and more connected. Access exclusive Patreon content such as: The Bonus Aftershow The Spoilers Club Early access to escape room Tour tickets and REA articles. Your Patreon support goes toward our mission: paying our contributors, funding our infrastructure, and supporting deep research and industry advocacy.   Our Other Podcast PG's Playhouse If you love wordplay, puzzles, and trivia, this is the podcast for you! PG's Playhouse recreates a fun game night, all in a short, 30-minute format. Of course, what's game night without making new friends? We bring on different guests for the different episodes. Each episode features a puzzle packed with wordplay and trivia, a short chat with the guest, and a segment exploring an interesting topic. I hope you'll take a listen and play along with us at PG's Playhouse.   Production Credits Hosted by David Spira & Peih-Gee Law Produced by Theresa Piazza & Sheena Vira Supported by Lisa Spira Edited by Steve Ewing Music by Ryan Elder Logo by Janine Pracht

We are excited to have Anthropic share their latest AI x Finance work at AI Engineer New York, coming up in 2 weeks!In case you've been under a rock, here's a non-exhaustive list of what Anthropic has been shipping since closing the largest fundraise of all time in May at $47B ARR:* June: Launched Claude Tag and Sonnet 5 and Fable 5* July: Opus 5, /checkup. crossed $65B ARR* Last month: Fable/Mythos 5.1, and EFS (upcoming pod)* IPO target $2T, end 2026 ARR estimated $100B* Cowork/chat merged before did* Claude Mods* Dario endorses the same Pacing the Frontier message cosigned by all labs* Last week: Opus 5.5, Plugins portal, Cloud Sessions/Claude Projects* Today: Sonnet 5.5!Today's episode should catch you up, with Thariq Shihipar, the explainer-king of Anthropic, who we last caught up on Fable launch day with The Field Guide to Fable:The Future of Mutable SoftwarePay special attention to Claude Mods (especially the cheatsheet):In general this is also the inverse of the other viral tweet from Thariq:Cloud Brain, Local HandsAnd give a try to Claude Projects:The “hands” terminology is not just an analogy for the local/cloud paradigm that is being built up at frontier coding agent companies like Cognition, but is ALSO particularly relevant to the safety systems discussions that we'll be discussing with Anthropic in an upcoming episode as they prepare to pace to frontier with responsible AI deployment.For those who want Thariq's writing tips we teased at the start of the pod, watch the full video here:From the rapid rise of Claude Code to a future where agents can rewrite their own harnesses, collaborate across teams, and operate across cloud and local environments, the way we build software is changing extraordinarily fast. In this episode, Anthropic's Thariq Shihipar joins swyx and Vibhu to unpack how power users are actually working with Claude Code today, why prompting remains a high-skill discipline, and where Anthropic thinks the agent harness is headed next.We go deep on Claude Code's evolving interface: Ask User Question and elicitation, artifacts as persistent generative interfaces, Claude Tag for multiplayer agent workflows, Projects, model effort, implementation notes, and the new Claude Mods system for customizing the harness itself. Thariq explains why Claude.md may eventually disappear, why the smartest model could also become the cheapest model for many tasks, and why mutable software could become a new paradigm for how applications are built and customized.The conversation then turns to agent security and Anthropic's “Pacing the Frontier” argument. Thariq walks through recent incidents where agents discovered unexpected ways to communicate, exploit infrastructure, reverse-engineer benchmark scorers, and chain vulnerabilities together. We discuss sandboxing, prompt injection, autonomous agents, interpretability, constitutional classifiers, probes, fallbacks, Auto Mode, and why securing increasingly capable agents may become one of the defining engineering problems of the next few years.We discuss:* Why agentic coding went from controversial to the default in less than a year* Why prompting is still one of the highest-leverage skills for working with Claude Code* How expert users build a mental model of Claude and what it can reliably one-shot* Why discovering your “unknown unknowns” matters more as agents become more capable* Artifacts as persistent, generative interfaces between humans and agents* How Claude could split into a cloud-based “brain,” local or remote “hands,” and dynamic interfaces* Claude Tag, Projects, and multiplayer agents and how collaborative agent workflows could evolve* Why spending more time on the initial prompt can dramatically reduce wasted agent work* When to use low, medium, high, or max effort for different engineering tasks* Why frontier models may eventually outperform smaller models on both intelligence and token efficiency* Why implementation notes can expose decisions the model considered but chose not to make* Why Claude.md may eventually disappear — and why starting without one can sometimes be better* Claude Mods: customizing the execution loop, UI, subagents, routing, and behavior of Claude Code* Model routers, forked agents, and supervisor agents that automatically improve agent workflows* Why Claude Mods may be an early preview of “mutable software”* The bitter lesson of harness engineering and why agent architectures go out of date so quickly* How Claude Tag is becoming an organizational harness for multiplayer work* Why giving agents access to company data creates an enormous new security surface* The Exploit-Bench incident where agents discovered ways to communicate and collaborate* Why agents hacked Hugging Face for scorer code rather than benchmark answers* How agents chained sandbox and infrastructure vulnerabilities in unexpected ways* Why increasingly capable agents make traditional security assumptions harder to maintain* The argument behind Anthropic's “Pacing the Frontier” proposal* Why software engineers are increasingly doing two jobs: engineering and keeping up with AI* Constitutional classifiers, probes, and fallbacks and what interpretability looks like in production* How Auto Mode checks whether an agent's actions actually match the user's permissions* Why Thariq can see serious AI risks while still having a relatively low p(doom)Thariq Shihipar* X: https://x.com/trq212* LinkedIn: https://www.linkedin.com/in/thariqshihiparTimestamps00:00:00 Introduction00:04:12 Ask User Question and the Future of Agent Interfaces00:08:29 Artifacts, Projects, and Multiplayer Agents00:15:37 Prompting as the Core Claude Code Skill00:21:52 Context, Effort, and Smarter Model Usage00:28:10 Is Claude.md Going Away?00:32:49 Claude Mods: Customizing the Claude Code Harness00:36:35 Model Routing and the Rise of Mutable Software00:44:40 The Bitter Lesson of Harness Engineering00:50:49 Claude Tag as an Organizational Harness00:55:59 Pacing the Frontier and Autonomous Agent Security00:58:22 Agents Hack Hugging Face for the Scorer01:05:34 What Happens When Agents Need More Compute?01:10:32 AI Coding Is Changing Faster Than Engineers Can Keep Up01:17:17 Probes, Fallbacks, Interpretability, and Auto Mode01:28:32 AI Risk, p(doom), and Closing ThoughtsTranscriptIntroduction: Life at Anthropic and the Pace of ChangeSwyx [00:00:00]: We're here in the studio with our friend Thariq from Anthropic, and I guess generally the Claude Code, I-- there's, there's so much, merging of boundaries and you've been so on top of everything since you joined Anthropic. You have been early to Claude Code itself, but then also, and you've told that story in other podcasts, and you've also been talking about seeing like an agent. Most recently you did the top AIE World Tour talk, Field Guide to Fable, which obviously you guys launched Fable, so that was-- that's cheating. And mostly you most recently also launching Claude Tag, and we're also gonna be talking about Pacing the Frontier. There's a lot going on in Anthropic. I guess top of the question is, what's it like being at Anthropic when there's so much going on?Thariq Shihipar [00:00:48]: I think that It is, like. I think you can get whiplash sometimes. I think, like, going. When I joined Anthropic, I joined because of Claude Code. Like Claude Code had just come out and I was like, “This is so good.” And Opus 4 to me was like just, I could not imagine, like, how good it was? And that was, like, a real moment for me. But I was, like, trying to convince, like, my startup friends to use agentic coding, and they're like, “Oh, no, like, our engineers don't think it's good enough,” or something. And I was like, “That's insane.” and now you, like, fast-forward, 12 months, less, and, like, it's just like, yeah, the default way that everyone codes, right? And I think that, like, just having to go from, like, selling it to, like, now, teaching people how to be. make the most use of it and be more efficient and things like that is just like a big, like big change. And, yeah, I think, like, it's just hard to stay on top of everything as a human? Like, I think things happen so fast and likeSwyx [00:01:51]: You just throw more agents at it.Thariq Shihipar [00:01:52]: Yeah, like that's like the agentic stuff scales much better than the, like, human stuff where it's like, oh, like, there are three things happening right now and, like, they're all emergencies and, like, how do you, like, respond to it? Yeah.Teaching People to Use Claude CodeVibhu [00:02:05]: What do you split your time on? You do a lot of technical writing, engineering work.Thariq Shihipar [00:02:10]: Yeah, so I think that, like, when I joined the Claude Code team, I wanted to teach people how to use Claude Code and I think that, like, that has been something that, like, I thought, like, maybe I would spend a little bit of time on it or, like, I'd, like, do. I was spending some time on the agent SDK first, and I wasn't exactly sure, like, how the bitter lesson would go, when it comes to, like, harnesses, right? Like, I think sometimes we were like, “Oh, like, what's after Claude Code?”? And so initially I was like, I just wanna teach people how to use Claude Code and make it easier to use Claude Code. And I think that has just, like, as the harnesses have gotten better and better, that's like the dominant problem now is, like, how do you use the agents, right? Like, it's like such a high skill expression thing. So I do that and then I do engineering work. I give talks, but I think, like, when I'm doing engineering work, my goal is to take that feedback that we get from users and also, like, then be able to talk about, like, hey, how to use Claude Code to do engineering. So there's like a good loop there. Yeah.Swyx [00:03:07]: Yeah. I'll-- For listeners, we'll attach, the talk that you did with Sarah for the Dev Writers, meetupThariq Shihipar [00:03:13]: Oh, yeahSwyx [00:03:13]: Which we talked a little bit about, well, first you do the work and then you talk about the work.Thariq Shihipar [00:03:16]: Right.Swyx [00:03:16]: Something like that.Thariq Shihipar [00:03:17]: Yeah.Swyx [00:03:17]: It's sow and reap orThariq Shihipar [00:03:19]: Yeah, reap and. Sow and reap.Swyx [00:03:21]: Something like that. Something like that. Yeah, so, and then just to preview a little bit, we are gonna talk about the evolution of the harness. It has come a long way from just being a CLI. We're gonna talk about, Claude Mods, which is starting to leak today, because you couldn't keep it secret.Thariq Shihipar [00:03:36]: Yeah. yeah.Swyx [00:03:39]: Yeah, there's, there's a lot, there. I think you started off with, like, adding ask user question tool, which people love and hate.Thariq Shihipar [00:03:48]: Yeah.Swyx [00:03:48]: Like, I thought it was, like, very innovative, and then now I have, like, my own version. You have your Interview Me version.Thariq Shihipar [00:03:55]: Yeah.Swyx [00:03:56]: And, yeah, everyone just has, like, their own stuff. And, like, it no longer matters ‘cause now you're supposed to, write prompts that create other prompts and loops and all these things.Ask User Question and Human-Agent InteractionThariq Shihipar [00:04:05]: Sure, yeah.Swyx [00:04:06]: So what's the state of the art, today? Like, what are people. what are you, like, telling people to do today?Thariq Shihipar [00:04:12]: Yeah, ask user question was the first time that the model was good at elicitation. I think this was, like, an emergent behavior that I, like, wanted to see if the models could do. I have, like a human-computer interaction background, so I, like, did that in undergrad and grad school. And so this was like. I think it's like human-agent interaction to me, like, trying to figure out, like, how can the agent communicate with you and extract, the requirements, right? I think that, like, one of the things about, like, that's difficult as Claude Code has gone broader and broader is that everyone has, like, their own way of using it, and it's very hard to, like, change the default behavior. So for example, like, if someone asks Claude Code to do something,Thariq Shihipar [00:04:59]: Sometimes they just want them to do the work, ‘cause they're, like, maybe a very good prompter, and sometimes they want. like, are not good at prompting? And you need. like, the agent needs to, like, clarify? And so that's, like, a good split. Like, and the ask you the question tool like, splits along that side where, like, are-- do you feel like you're good enough to instruct the agent as it is, or is the agent able to, like. does the agent need to, like, pull out more requirements and, like, collaborate with you more and really understand your preferences?Thariq Shihipar [00:05:27]: I, on the whole, believe that pretty much everyone is more on the latter than the former, that they, like, have more ambiguity and they know less than they want, than they, like, think they know about the problem. but, like, it's like a interface design problem to make that easy? And so, like, if you're designing a problem, like, or if you're going through a problem, like, things like what's the schema or, like, what's the call stack and things like that are really important. like, the details in the design are important. Ideally, you want to figure out some of these, like, hard problems ahead of time before starting implementation. And yeah, that's why they call, like, unknowns, right? And so I think that this will forever be, like, a skill in agentic coding is, like, figuring out your unknowns. So, like, because even if the model is, like, super intelligent- It, like, needs to know what you want? And, like, you have preferences. like, you need to like, pull the, pull that out. and so that's, like, I think how I'm, what I'm pushing. the question then is, like, how does the agent interact with you? And I think that has been HTML, has been, like, the big way of doing that. And we've recently added artifacts, right? And artifacts, I think we've done a bad job of, like, or, like, I've done a bad job of, like, explaining how to use them fully. We have a lot of property capabilities. They have a database associated with them? And so every artifact can store and write persistent data. They can, like, feed back into Claude? And so, like, one thing that, like, people are not doing yet that I'm trying to, like, encourage is, like, this idea of a dashboard artifact. So you have, like, Claude working on a project long-term. Maybe it's like a kanban or something. it can store that kanban data in its database. Multiple Claudes can access that data via, like, the artifact MCP, and, like, that artifact can, like, talk to those Claudes as well. And so, like, the. We're building the primitives for you to be able to have this, like, generative interface via artifacts that will, like, let you surface more of that rich detail from the agents. And I think that, like, almost everything with agents right now is, like, this problem of, like, you think what you want, but you don't really know what you want, and, like, the agents need a lot of detail, and collaborating with them in the loop is really important. And so artifacts are, like, the, like, way that we're trying to evolve there. But there's a lot of work to do because it's so much more complicated than, like, a multiple-choice question? there's a lot more, like, detail in terms of, like, diagrams and code snippets and schemas or, like, whatever it is for that problem. But, like, artifacts is, like, the mo-more AGI-pilled way of, like, doing ask user question. So yeah.Artifacts as the Interface to the HarnessSwyx [00:08:15]: I think one thing that's unclear to me about these, the artifact stuff is, like, what feedback should go in through the artifact and what feedback should go through a Claude, a chat? Because the more AGI-pilled one is to just feed everything to the Claude.Thariq Shihipar [00:08:29]: I think the more AGI-pilled one is to go through the artifact. Like, and I think that, like, we imagine in the limit, I think that artifacts will be your interface into the harness? You can, like, comment on this, like, live, like, document of your plan, of the work. you can see maybe, like, multiple agents and different agents are doing this, and that artifact is built for the current work that you're doing, right? And so, like, each one has, like, slightly different. I think we're still, like, getting there from, like, an infrastructure perspective. But yeah, I think, like, on-the-fly interface for your harness is probably where things are headed.Vibhu [00:09:03]: Is there a version of it that's an abstraction from CLI or chat and you. Because right now, a lot of it is, okay, you're interfacing with Claude Code, you're having HTML given back for a mockup. It's pretty rich. There's diagrams. Artifacts are ways to connect these together. Why not just do everything that way?Separating Brain, Hands, and Surface UIThariq Shihipar [00:09:22]: Then it becomes, like, separating out, like, where is the inference happening? Where is the intelligence happening? Where is the work happening? like, I think this is like, difference between, like, or, like, some of the distinction between local and cloud, right? And so, I think right now, if you use Claude Code, it's, like, local and, like, you can spin off remote control, for example, to get some cloud behavior, or you can spin off Claude Code in the cloud, right? We're moving towards a place where instead of Claudes, like, you message a local Claude, it starts a session locally and it executes, to more like you have a Claude that you message that's in the cloud that's running. it can run, like, local, or, like, cloud sessions. This is how Claude Tag works. But, like, over time, we'll add, like, local hands as well. And so, like, local hands will be the ability for that agent to access your computer if it's online, and be able to, like, work there. And so it can spin off many different subagents. It can, like, commu- those subagents can communicate with each other, and that's where the artifact comes in to display all of that work. So you can imagine, like, the. You're separating out these things. So there's, like, the surface UI display that's an artifact and hosted somewhere and has a database and everything. There is the inference intelligence, right, that's happening on the cloud, and you don't have to worry about shutting off your computer or whatever, right? and then there's the, like, hands. Like, and it can be local, it can be in, like, a remote sandbox or wherever you need your work to be done. That's like unpackaging, like, the Claude Code experience right now where, like, right now it all happens in one place, right? So.Multiplayer Agents, Claude Tag, and ProjectsVibhu [00:11:00]: How do you see, like, the multiplayer side of that? So say teams want to work in this way. Right now it's very individual, but how do you see the future of multiplayer? Like, right now, I guess there's Claude Tag, which is a version, but.Thariq Shihipar [00:11:12]: We're launching projects. And so projects is the, like, this abstraction that's like Claude Tag, but on our Claude products, right? So you can message it and, like, it will do the Claude Tag-like stuff, like spinning off subagents. So We think with multiplayer. Like, Claude Tag is, like, a little bit more native multiplayer because it's just, like, in your Slack and the permissions are all figured out and stuff like that. But I do think multiplayer is, like, an important part of the story and, like, that will need to get tied together more. Like, you can imagine how complicated it gets when you're like, oh, you have hands, but now you have other hands in other people's computers too, and, like, you need to, like, permission them or, like, you have, like, your MCP and someone else's MCP, and how do you figure out how to use them, right? It gets, like, quite complicated. And Claude Tag does a good job of, like, sanding down all of these issues, right? So that, like, when you have, yeah, Google Docs, how does it access Google Docs, right? Like, it accesses through the shared Claude MCP, or it can access through your local credentials as well if it doesn't have access. But yeah, I think Claude Tag is our multiplayer, product, and it's really useful for these, like, things that are inherently multiplayer. Like, okay, like on-call, for example, incidents are inherently multiplayer. You want to tag Claude, you want multiple people to log in, you want it to be able to find context. I think whenever I'm, like, working on something and I want, like, privacy or security or, like, I want other people to review it's really nice to, like. I'll have a channel per project and I'll, like, at legal, for example, be like, “Hey, like, I want to ship this. Can you, like.” Like, here's. Like Claude knows everything, just chat with it. And that way legal gets precise answers, on like what exactly is shipping into the code, and I don't need to be in the loop, right? So I think like multiplayer is getting like more and more like, yeah, everyone can participate with Claude. I think Claude Tag is like that product and like projects will start off single player and will like, expand.Swyx [00:13:14]: I think there's a question about like maybe dual questions about identity and the unit of isolation.Identity, Permissions, and IsolationThariq Shihipar [00:13:20]: Yeah.Swyx [00:13:20]: Claude Tag, you specifically chose to make it its own identityThariq Shihipar [00:13:26]: Yes.Swyx [00:13:26]: Which is like, a controversial choice. There's, there's other ways to do it.Thariq Shihipar [00:13:30]: Yeah.Swyx [00:13:30]: Claude Projects probably it sounds like, if it's anything like ChatGPT Projects, it is, the isolation is that artifacts, that cloud instance, everyone's collaborating on this. It'll. It sounds like, it should be like if you're, if you're collaborating with legal on a thing, like that channel should be a project, right? Like it's not yetThariq Shihipar [00:13:50]: Yes.Swyx [00:13:50]: But it. that's the natural next step.Thariq Shihipar [00:13:53]: Yeah, like I think in Claude Tag, it's effectively. Like Claude Tag, you have to do your own arrangement. And so Claude Tag, yeah, each channel is like you can name it as you want, and I nameSwyx [00:14:04]: Yeah.Thariq Shihipar [00:14:04]: Like each featureSwyx [00:14:06]: Yeah.Thariq Shihipar [00:14:07]: As a channel.Swyx [00:14:07]: And, but I think like there is some trans- like it's unclear when there is transference, because let's say it is. if you have a coworkerThariq Shihipar [00:14:14]: Yeah.Swyx [00:14:14]: Who is tagging on all these things, yes, there is transferThariq Shihipar [00:14:16]: Yeah.Swyx [00:14:16]: Because it's the same person. but with Claude, it's unclear if it's like necessarily like, well, no, you don't know any of. you don't know about the other stuff. You should only use this stuff.Thariq Shihipar [00:14:25]: It's like the tip of the iceberg meme, right, where you can like. This is what we spend so much time onSwyx [00:14:31]: Yeah.Thariq Shihipar [00:14:31]: Is like there is like infinite surface area of like, okay, you want Claudes to. Not infinite, but like there's like surface area, a lot of like, surface area to figure out of like permissions and visibility and like how can you let Claude operate as well as you can, as safely as you can? And obviously, this is very important to us because like security for our code base is very important. And so we've put a lot of time into this. Yeah, there's so many like edge cases you can figure out where it's like, oh, like, yeah, this Claude in this channel has different permissions, but it can message another channel, and can't it exfiltrate data that way? Or like can you like. What if it uses your MCP and then messages someone else? Like there's like so much, and we've like really put a lot of work into sanding it down.Swyx [00:15:14]: Yeah. Lots of work. okay. Fable?Fable and the Meta-Skill of PromptingVibhu [00:15:18]: Fable, you wrote two good articles. you've written many good articlesThariq Shihipar [00:15:22]: Yeah.Vibhu [00:15:22]: But on, Field Guide to Fable, Building Claude Code. I'm curious from what you've seen, is there any common patterns that you see in like top users at Anthropic externally? Like what are best practices for getting the most out of Claude Code?Thariq Shihipar [00:15:37]: The like meta skill I say is like prompting is like very important? And like that. Like I think this is like not trivial to say because I think a lot of people are like, “Oh, prompting doesn't matter. It's just like I can just say a sentence and Claude will do it.” And I think prompting is really this like, this. It's like public speaking, like, or writing or something, and for a specific audience, and that audience is Claude. And you need to like build a mental model of Claude and how it thinks and how it works, right? And so that's like the most important skill in working with Claude Code is like having this mental model, right, of Claude and like what it can do well, what it can one-shot, what it can't. And so many people when you see prompting, they're just like, they're short prompts, but they have such a good mental model of Claude and of like the code base and things like that like it's effortless? But it's like high skill ceiling. So like that work of like, spending a lot of time prompting and building mental models of how, and intuition for how the agents work is really important. And then I think like the next thing is like the unknown stuff we talked about earlier, where it's like being able to find out like your, what you don't know or what you haven't written down, learning about like different things. I think as Claude can do more and more things, the likelihood of you doing something out of distribution for you and like you have low domain knowledge on is very high? And the more you can like learn the vocabulary to be able to prompt Claude, it becomes really important. And so like I think the most important unknowns are the unknown unknowns, where you're like, I just like don't even know that this exists, right? Yeah, exactly. I think that's like a illustration of like the map and the territory, right, where you're like, “Okay, this is my prompt,” and the territory is like the actual like work that the agent needs to do, right? And if you are like very precise, you can give more precise things, right? So like for example, in design, I'm not very precise. I'm not a designer, so I say like, “Give me like eight different mock-ups.” But if I was a designer, maybe I'd be like, “Oh, hey, here are some reference sites.” Like, “I want this type of font and this type of like look to it, and here's like a few different components to like visualize. Here's a Figma MC board to bring in,” like. And so you can just be so much more precise with that language. And if you're not a designer, you just need to like try and learn the language or learn the unknown unknowns. And this is true of like everything, I think. Like the more, like you can work with Claude to learn like how things work, the better your prompting will be. I think another good example of this is like game design, like where a lot of people are like, “Oh, like I can vibe code a game now.” And they're like, “It's not fun.” And like it's just like the thing about game design is like every one of these choices has like a lot ofTaste, Domain Knowledge, and Learning the VocabularySwyx [00:18:25]: Variations.Thariq Shihipar [00:18:25]: A lot of like craft to them. So it's like, oh, okay, like when you're making a flying game, the feel of the plane and the like, way it responds to your controls has a lot of like. Like, a game designer would spend like days on that. Do? and likeSwyx [00:18:44]: To me, that's what taste is, right?Swyx [00:18:45]: Like it is like from the possible space of one thousand mathematically valid answersThariq Shihipar [00:18:49]: Yeah.Swyx [00:18:49]: Here's the one that is the humans will like.Thariq Shihipar [00:18:51]: Yes. Yeah.Thariq Shihipar [00:18:52]: I think with taste, I'm like torn on this word ‘cause I think you're right, but everyone has different definitions, and it sounds kind, sounds like low skill or like elitist almost, where you're like, oh, like there are certain people with taste?Swyx [00:19:06]: It's like taste is what I call taste.Thariq Shihipar [00:19:07]: Yeah, exactly.Swyx [00:19:08]: And it's like these guys don't have taste.Thariq Shihipar [00:19:09]: Yeah, exactly. Oh, like an engineer doesn't have taste. Like I, the like founder, have taste.Thariq Shihipar [00:19:14]: ? And I think that's not true. Like I think like the engineers have a lot of taste for these particular like problems? And I think everyone has taste for particular problems. I think like Jason Liu, like say like in order to, yeah, have taste, you have to eat?Thariq Shihipar [00:19:32]: And I really like that, where it's like, okay, you have to like do a lot of things. You have to like iterate and figure out what you want, what you like, and, like build that like domainSwyx [00:19:41]: YesThariq Shihipar [00:19:41]: Domain vocabulary. And then when you're prompting, you're like synthesizing all of that for a product.Swyx [00:19:46]: Isn't it annoying when someone else says it better than you?Swyx [00:19:48]: It's just like, f**k, I have to quote this guy forever.Vibhu [00:19:51]: Having to quote Jason Liu forever.Vibhu [00:19:53]: He's gonna love this.Thariq Shihipar [00:19:55]: So I get prompts, more than that.Vibhu [00:19:57]: And sometimes it's not even that. Sometimes it's just intuitive, right? Like you don't realize you even want something till a model puts it out, and you're like, “Oh, this just feels immediately better,” right?Voice Prompting and Information DensityThariq Shihipar [00:20:07]: Yeah, exactly.Swyx [00:20:09]: One thing I go back and forth on is I feel like the way I prompt half the time, let's say I use voice.Swyx [00:20:16]: Did I say voice? Other people have voice. that is the opposite. That is just like me rambling for like two minutes Pressing down the function key and then let go, and then like hopefully it figures it out. And oftentimes it does.Thariq Shihipar [00:20:26]: Yeah.Swyx [00:20:26]: But it's not as thoughtful as like a structured prompt with like Well-run communication as though it's a PRD or a memo. Is that in line with how people do this? There's like bimodal prompting where there's some prompts where you spend a lot of time upfront and other prompts you just dash it off?Thariq Shihipar [00:20:43]: I don't think the voice is necessarily low. Like I think it's like more like how much information is in the prompt. like the model can. Like you can and like add some sentencesSwyx [00:20:53]: RightThariq Shihipar [00:20:53]: And be like, “Oh, like I changed my mind,” like in the middle of the prompt, and it will be able to follow that perfectly? So I think the like actual format of the text is less important, but then like the ability to. Like how much information is in it, right? And I think for voice, a lot of times, going back to like human-agent interaction and like for a lot of people, it's just way easier to talk than to like type? and I. If that gets more information out of you, like that's better.Vibhu [00:21:21]: At some level, it feels like just giving the model as much contextThariq Shihipar [00:21:24]: YesVibhu [00:21:24]: Over prompting before you kick off is a best practice. I don't know. A lot of the times, like when I was first trying out Fable, I spend a solid 30 minutes like really crafting a long prompt. This, I think, is a response of models running for longer and longer, right? It's still a little difficult to nudge them as they're in like, in the loop, but I just like intuitively spend more time kicking off that first prompt and working with it a lot.Spend More Upfront, Iterate LessThariq Shihipar [00:21:52]: My personal opinion is that if I was a software engineer, if I was like, just running my own startup, for example, I think I would mostly fit, stick to a max 20x? like maybe verification and so code review are like separate things. But I think like what I see a lot of times is people hit rate limits when they're doing this like, oh, like it did a lot of work and you're like, “Oh, I don't like this.” Like, “Can you like undo this and redo it?” And then you're like iterating on this like thing that the model could have done if you had like spent more upfront time or given it better context? And instead it's like you're like, “Nope, don't like that design. Try this.” Or like, “You messed this up,” or something like that. And then that just eats up so much more of like, your usage. And so that's like, I think maybe like a key like tip both for like efficiency as well, right? And yeah, I think like context, and not just like context on like what the goal is good, right? Like are you building a prototype or is it like a production thing? Like where can you spend compute or when, where can you not spend compute? Like I think you have to give the model permission or like not permission to do things sometimes where, like it doesn't know intuitively how much you want to spend on this task, right? And you can use effort for this. So I did-- I'm working on a blog post about that where it's like, if you want. For like we see that effort scales with the complexity of the task. So for security, effort gets like way more results. Like high effort versus like low effort gets, like changes the evals a lot. But for software engineering, it doesn't change it a huge amount because effort is mostly spent on the verification and the like edge case testing and things like that. And so like being able to like give the model that guidance of like, “Hey, this problem is something that I think I want you to spend a lot of time verifying and edge case testing,”?Effort, Model Choice, and VerificationVibhu [00:23:43]: How about model in the mix? So, there's Opus and Fable with effort.Thariq Shihipar [00:23:47]: Yeah.Vibhu [00:23:48]: There's also Haiku in there.Thariq Shihipar [00:23:49]: Yeah. It's not quite true yet, but it's very close where I think the frontier models will be Pareto dominant over like almost everything. like maybe. And sometimes I think Opus might be Pareto dominant. Do? Like I think depending on like how things, like shake out if it's like a newer version of Opus. But I think that like increasingly it's just going to be like the smart model is going to be able to like do the simple task for less tokens than the like the other models because of verification. With verification, in the limit, your model doesn't need to verify, right? If it's a perfect model, it just does the work once and it's like, okay, like you, I did it? And increasingly with Fable, I'm like, I'm like, “Dude, you don't need to spin up Chromium and screenshot all of these things.” Like I see it. Like you did it, right? And so a lot of the. At higher effort, you spend more of those tokens verifying. But if you're working on simpler problems, and a lot of software engineering is like well, like in Fable, like low and medium stability, it can spend less tokens verifying. And as the models get smarter and smarter, they will just be able to like, “All right, done.”? Like, I can run the lint for sanity's sake, but, like, I, like, know it lints? Like, you don't even need to do that. And that will be so much more token efficient than, like, the smaller models. Yeah.Swyx [00:25:15]: Is there a good, practice on our side that we can use to see if we're using too much effort? Like, I freakingThariq Shihipar [00:25:23]: YeahSwyx [00:25:23]: Hate wasting time on that stuff.Thariq Shihipar [00:25:24]: Yeah. I know what you mean. I think, like, so in this blog post, my rough distribution is, like, code review and security should be, like, high or max and, like, software engineeringSwyx [00:25:37]: You said recommend mix settings per domain.Thariq Shihipar [00:25:37]: Yeah. I think, like, if you're doing, like, UI or something like that, like low and medium, I think is you're building, like, an API and you want to make sure, like, you cover enough edge cases? And so I think building, like I said, that mental model of, like, how things work across these distributions is, like, yeah, part of the job.Implementation Notes and Decision LogsVibhu [00:25:56]: This is more intuition-driven or eval? Because I'm guessing this would change as you go.Swyx [00:26:00]: He has evals.Thariq Shihipar [00:26:01]: Yeah. So what I did in the blog post is I go over all of the terminal bench evals. So there are, like, 70 problems and I'm show that, like, okay, like, in the security problems it does more. and then I also, like, look at some of the transcripts just in terms of, like, how-- what does it answer, what does it forget or something. And a lot of times, this is another prompting tip I have, is, like, asking it to make decision notes or implementation notes because, in every eval problem that it faces, it thinks about the correct solution, and decides not to do it. it's like, oh, like, here is the answer. What if I did this? And then it's like, oh, probably not? and then keeps going. And this is, like, the majority of the failures, at, like, a higher max level. It's very rare that the model just doesn't know how to do something. If you just have these implementation notes, then you can review and you can be like, “Oh, I want you to do this thing that you didn't do.” The models are getting better at surfacing that overall. Like, I see in the transcripts of Fable 5.1, like, when it does this output, it will call out its decision-making as well. but making this more explicit in the harness is better. And now we're, allowing ways of you modifying the harness so you can, like, add someVibhu [00:27:23]: Ooh.Thariq Shihipar [00:27:24]: Calculate with there. Yeah.Swyx [00:27:25]: Yeah. So I do wanna call out two things that you mentioned that I think exist outside of prompting. One is like, let's, let's call it the prompt that is so important that it shouldn't be in a prompt. It is in Claude.md or Agents.mdThariq Shihipar [00:27:38]: YeahSwyx [00:27:38]: Which is like goals, right? Like your situation, your goals, the things that you want, the thing. and then second of all is the decision log or the experiment log or whatever log of traces that you might want to survive the current session to do those things. Those are, like, externalities that there's no standard. There's no-- It's not like skills. It's not like MCP. There's no standard. It's, it's just like it's a markdown file. first of all, is that right? Is Claude.md going away? You have a documented dislike of, Agents.md, but you're gonna do it?Claude.md, Agents.md, and Model-Specific InstructionsThariq Shihipar [00:28:10]: Yeah. Okay. So Agents.md, yeah, like, we're, we're gonna do it. I think it's just, like, different models are very different from each other? But I realize that it's, like, such a pain to, like, maintain different ones? And yeah, like, as the models get better and better, the floor of how they accomplish the simpler task is better. And so I do think in the limit, Claude.md goes away, and maybe not even, like, that far. Like, I think, like, I think that right now it might be better to start a new project without a Claude.md.Swyx [00:28:44]: Yes.Thariq Shihipar [00:28:44]: I think that, like, maybe if you see very repeated failure modes, you add them to your Claude.md. The really tough thing is that this changes per model. And so, like, if you've added a bunch of failure modes or, like evenSwyx [00:28:57]: So you need Fable MD, you need Opus MD.Thariq Shihipar [00:28:59]: Or well, even Fable 5.1 versus Fable 5.Swyx [00:29:03]: Yeah.Thariq Shihipar [00:29:03]: Like, it is annoying. Like, I'm not like,Swyx [00:29:05]: YeahThariq Shihipar [00:29:05]: Like, we don't, like, do this on purpose? It's just, like, how the models work, right? And so, like, maybe, like, Fable 5 had this, like, failure mode that Fable 5.1 doesn't. And if you keep this context, this running log of a bunch of different failure modes, they will probably over constrain Claude? And so this is like. we just added evals plugins for skills.Swyx [00:29:28]: Yeah.Thariq Shihipar [00:29:29]: And so now you can eval if a skill is better. I think Daisy on our team did this. And so, yeah, this is like we're trying to work on this. We know it's, like, you still have to spend tokens on it and, like, it's not, it's not perfect, but it's, like, we're trying to help out with this problem.Swyx [00:29:44]: And so, and as far as prompting goes, the one tip I wanna offer is, something I have told people a lot is sufficiently advanced prompting is indistinguishable from sufficiently advanced executive communication. So I've referred to-- This is an executive comms workshop from Heavybit that is the best I've ever seen in my career. And they teach this thing called the SCQA model. Just Google it. It's a, it's a thing. Like, people have done prompting for decades. It's just called executive communication. It's like when one person has to communicate to thousands of people down the org chart, this is what you do. so situation, complication, question and answer, is how you write the memo. but obviously sometimes you don't have the answer, but you can at least list out the SC and Q, and then they have some examples in there. So just leaving breadcrumbs for people if they want to explore.Underrated Prompting Patterns and ELI5Vibhu [00:30:31]: Before we move on, I wanna ask you, any other underrated tips, ways people could get a lot of value from Claude Code that they're not using?Thariq Shihipar [00:30:41]: Yeah, I think a lot of them are in the, this unknowns, like, doc. Like, I give a bunch of example prompts, like, using it for brainstorming, using it to quiz you after. we added this, like, explain it like I'm five skill which is a very short prompt. And it doesn't even say explain it like I'm five. It's like the key word of this prompt is big pictures, few words. like, that's like the main thing. And it is shockingly good? Like, you, like, I think I tweeted about this and it's like /eli5, and, like, you can install it as a plug-in. But yeah, it's, like, way better at just cutting through the BS and being like, yeah, exactly right here. So the diagrams are, like, quite clear. I think one of the things that is true with artifacts is, like, they put too much text in and people are not reading the artifacts? And so, like, this simplifies it a lot more. And, yeah, this came out of, like, just people at Anthropic, like, going through very complicated incidents and being like, “What is happening?”? So, this one I think is great, yeah.Swyx [00:31:47]: My version of this is the, it's like test your understanding. Give you a few choices and then, like, if you get it wrong, you have a mismatch between what you think is happening versus what's happening.Thariq Shihipar [00:31:58]: Yeah. I think this is one of those things that everyone loves talking about, and then very few people really do. Like, I thinkSwyx [00:32:05]: Really helpful.Thariq Shihipar [00:32:07]: Yeah. But most people just don't want to get quizzed about something? Unfortunately, I think this is one of the, like, things that we need to, like.Swyx [00:32:16]: What's the opposite of ask you the question or ask you the question before the thing?Thariq Shihipar [00:32:19]: Yeah.Swyx [00:32:19]: This is after the thing.Thariq Shihipar [00:32:20]: Exactly. Yeah.Vibhu [00:32:21]: It's a good way to stay grounded of, like, do you even know what you're doing, right? The worst case is when people send you slop and they haven't understood what they're asking for or what the output is, and it's like, “Dude, I don't wanna read this. Do you even know what it is?” So, you make it a rule for yourself that before you send stuff, you should at least know what's implemented.Claude Mods: Customizing the HarnessThariq Shihipar [00:32:41]: Yes, but so you could make this a mod and you could build your own mod to, like, make sure you test it. So yeah, you can do that.Swyx [00:32:49]: All right. Let's get right into it. What is Claude Mod, and what is this diagram showing?Thariq Shihipar [00:32:54]: Yeah. Okay, so Claude Mods is you can customize the entire Claude Code harness, and we're going to. If you have requests, we will, like, let you, like, please let us know. We'll add more and more. This works for CLI, it works for desktop. maybe it will work for Claude Tag in the future. I don't know. Like, we're trying to make this very extensible. You can see this reference sheet. I don't want people to get overwhelmed by it? At a high level, you can customize both the execution of the harness, and the UI of the harness. And so, like, you say on that Tetris example from Boris, that's like customizing the UI, right? Like showing, like, Tetris in the game.Thariq Shihipar [00:33:35]: But, like, let's say that you wanted to do this thing where you had. you tested your assumptions or, like, tested your understanding after every project, right? What you would do is you would ask Claude to make this plug-in. It would spin a classifier after every prompt. And so, like, at the end of each turn, you would spin off a sub-agent or, like, a forked agent. A forked agent is, like, maintains the prompt cache, right? So it's like a, like one of those unintuitive things where you can fork and do, like, a little request, and it'll be very cheap because the entire prompt cache is, like, done. And so you can be like, “Has this task been completed?” likeSwyx [00:34:18]: This is how you do BTW and all those.Thariq Shihipar [00:34:20]: Yeah. The underlying forked agent, yes. But so you can, in the f-fork sub-agent, you can say, like, “Has this task been completed? If so, return true.” And then in your hook, or in your, like, plug-in mod, or sorry, like, in the sub-agent probably, you would say, like, “If true, give me a quiz.” give me questions and answers, and then, like, in a JSON format, and then you'd parse it, and then you display above the prompt input, this list of questions, right? And so this is something that's, like, slightly token-intensive because, like, you have to do it after every end of the assistant turn. But it's, like, a lightweight classification, and then you can, like, get this quiz, and then you'll see, like, Claude will always do it for you. You don't need to remember to do it. There are lots of these, like, tips that we've talked about, right, where it's like, oh, implementation notes. You can also add a tool for implementation notes now. And so, like, this tool that I'm adding is, like, register, like, I think assumption is what I'm calling it, but, like, maybe I'll change it around. And this is a mod. And so, like, you give it a register assumption tool, and then it will keep a list. It'll. Every time it does it'll keep a, like, add to the list, and then at the end it will display those assumptions? Another mod I'm working on is a model router. And so, like, internal, like, Claude model routing, right? So it's. This is, I want to say the reason we don't do model routing by default is, like, it's a hard problem? And likeForked Agents, Assumption Tracking, and Model RoutingSwyx [00:35:51]: You will get it wrong.Thariq Shihipar [00:35:52]: Yeah, you, like, yeah, you will, like, accidentally use, like, Fable for a hard problem or Sonnet forSwyx [00:35:57]: Yeah, if you have auto approve, but you don't have auto mode.Thariq Shihipar [00:36:01]: Well, you will have auto. Like, you don't have, like, auto routing or something.Vibhu [00:36:04]: You don't have auto mode for model picker.Thariq Shihipar [00:36:06]: Yeah, exactly. SoVibhu [00:36:07]: I'm getting the rough question of, like, how much do you open this up and how much do people have to think about this? Like, when you talk about prompt caching and building a router, it seems like you could easily build a mod that routes per query, and I'm just killing my plan very fast, right? I guess my question is more so, like, what is, like, a product talk like this look like, right? Who is it for? Is it for power users? Is it everyone should be able to go throughSwyx [00:36:33]: Oh, definitely power users, right?Thariq Shihipar [00:36:35]: Yeah, I think it is power users, but, like, the nature of Claude Code is that so many people are power users? Because it's easy to share things, like you can. Like, one person can make a good model router thing that doesn't break prompt cache all the time, and then you can, like, compose them. Another cool thing about the plug-ins is that they can hook into and compose with each other. And so I have, like, a mod that will, like, create a mode selector at the top, and any plug-ins can register to be a mode. And so, like, the auto router can be a mode, right? Or, like, you can have a mode that's, like, artifact mode, where it's like it primarily talks to you in artifacts. like, you can toggle between plan mode? And so, like, you can create more and more of these modes. But the ability to create modes is in it itself a mod? And so there's a lot of richness here, but we do want to make it fairly easy. We want to be-- make it so that you can just, like, install someone else's. You can ta-- you can chat with Claude and, we'll, like, make sure that it understands the nuances of things like prompt caching and stuff, so it can, like, warn you. This is, like, not extremely complicated behavior for Claude, I think, but we should have just a good skill on how to make mods. and yeah, we'll see how we go. But I do think that this is, like, a preview of, like, mutable software, and, like, how, like, generative software, just like you can customize safely. If enabled, you could customize any piece of software. And I think that more and more apps ideally do something like this?Power Users, Modes, and Mutable SoftwareSwyx [00:38:13]: And by the way, you, we have, you have another cool tweet about how, there's the infinite money button, which is like make your SaaS, consumable by agents. I think mutable software is interesting and, other people have also tried to do it. I think the hurdle comes when you can do everything, then people, users get, tend to get confused. So usually the stuff that works is just like one opinionated flow. This is in the side of less opinionation. It's just like, well, more power to power users. And I think probably unlocked by AI, where, like, you can just prompt for whatever the thing is.Thariq Shihipar [00:38:47]: Yeah, or there can be a skill that gives the opinions?Mods vs. Hooks vs. ArtifactsSwyx [00:38:50]: Yeah.Thariq Shihipar [00:38:50]: And then, yeah.Swyx [00:38:51]: So knowing a little bit about, like, TypeScript and build systems and all these things, the closest-- I'm very curious that the team who worked on this, if, I don't know how close you were to them, if they drew any inspiration from build systems like Babel, Webpack, all these, like, old school things. Because it sounds very similar, like the plug-in ecosystem of those things where they can compose with each other.Thariq Shihipar [00:39:11]: Yeah, I'm not deep in the technical details, but I do know it was a collaboration with someone on the Bun team and someone on the Claude Code team.Swyx [00:39:17]: Yeah, it's a build system mecca.Thariq Shihipar [00:39:19]: Yeah. Exactly. It's, it's very exciting. But yeah, like, agents can just do this very complicated like, extensibility into your software now. And so, yeah, like, another reason to, like. If you run a startup, like, you can just prompt Claude and be like, “Hey, like, could we make an extension system? Like, what would that look like?”?Swyx [00:39:37]: Yeah.Swyx [00:39:38]: And I just really wonder, like, you had hooks in the past and plug-ins, all these things. So what specifically will mods be able to do that those things could not do?Thariq Shihipar [00:39:47]: Internally, we were originally calling this function hooks. And so, like, that's, like, gives you a little bit of an idea where, like, hooks register a, like an event to happen and then, like, a script to call. And this inside of the, like, TypeScript runtime is running things. And so, like, you get some benefits of just, like, it has a bunch of things in the Scope with, like, for example, like how many turns is in this conversation, right? Like, how many tokens have been used? Like, et cetera. Like, what are the messages? Things like that. So it has a bunch of messages that can be used. And then it's just, like, a lot more hooks. So we have, like, or a lot of, lot more, like, things you can register on. And then you can do because of the. because it's all happening in process, you can, spawn sub-agents, with four contests and contexts and stuff. And, like, that will return. You can parse the results of those. You can use structured output to like, return them. and then you can modify the UI, which you can never do in hooks. So, yeah.Swyx [00:40:50]: Yeah. Yeah. So modify UI, this is why you showed the Tetris example. Does it also ex-extend to artifacts? I assume it does.Thariq Shihipar [00:40:57]: You-- Like, artifacts are like a different way of customizing it. like, you can definitely. One of the mods I'm working on is, like, this dashboard mod, which will, like, prompt Claude to maintain a dashboard, that's an artifact. But they're like, slightly orthogonal, or not orthogonal. They compose with each other in different ways. Like, mods are, like, a little bit more, like, in your Claude Code harness, changing the agent loop? And, like, the UI is, like, an added benefit. and then artifacts are just like you want to, see things at a high level, very inter- highly interactive. like, the affordances can be a lot bigger than, like a TUI or even in our desktop.Next Steps, Supervisors, and Persistent GuidanceVibhu [00:41:40]: I'm guessing you'll have a good blog post on the differences, because right now you can also, make a loop that outputs to an artifact that's an interactive dashboard, but you can also do it with a mod. There's just some thinking about making a hacking on a harness when we don't know much about the harness, right?Thariq Shihipar [00:42:00]: Well, something I'm excited about with mods is, like, there's so much things with Claude Code that you just have to remember? You're like, “Oh, like, let me do this, and then let me call the dashboard skill that does the loop,” and things like that. And, or like, “Let me test my assumptions afterwards.” And I think, like, if you do all of these things using these little classifiers and stuff, and you're like, “These are the things I care about. This is what I want to do,” you can, like. You don't have to remember as much. One more, like, mod I'm working on is a next steps mod thatSwyx [00:42:28]: I have-- I was gonna say, I have a next step skill. I always run next steps.Thariq Shihipar [00:42:32]: And does it have access to your skills? Like, this is one of those things where I'm like.Swyx [00:42:37]: I think so.Thariq Shihipar [00:42:38]: Okay. Yeah, probablyVibhu [00:42:39]: Do skills need specific access toThariq Shihipar [00:42:41]: Well, I think there'sSwyx [00:42:41]: Don't they always haveThariq Shihipar [00:42:42]: I think there's, like, specific prompting, I guess, to, like, know your skills. Like I think Claude forgets them sometimes throughout, like, the thing. But anyways, the idea of, like, yeah, next steps that also are like, “Oh, hey, this has happened. Use the explain skill to explain to you what happened because this seems, like, quite complex,”? Or, like, yeah, “Use your unknown skill. It looks like you are, like, asking the model to, like, iterate on these small changes. It seems like you could prompt better.” like, “What if you did this?” Right? So, I think, yeah, like spending more compute there. Yeah.Swyx [00:43:20]: And it should always come out as multiple choice. we have, I haveVibhu [00:43:23]: We have his skill.Swyx [00:43:24]: My next step skill is like this.Thariq Shihipar [00:43:26]: Okay, perfect. Yeah.Swyx [00:43:27]: You can steal it.Thariq Shihipar [00:43:28]: Yeah.Swyx [00:43:29]: Like, but like, for me, it's all-- I think models really always need to be reminded, what are you trying to do here?Thariq Shihipar [00:43:35]: Yeah.Swyx [00:43:35]: Look at the whole transcript and go like, oh, was this original goal? Did your solution solve it? Were you lazy? If you're lazy, maybe there's a reason. Maybe you needed approval from me. Maybe you needed, there's two things you wanna suggest. So it's, it's a little bit like the modification of the ask user question or interview me skill. so it's next steps.Thariq Shihipar [00:43:55]: Yeah, exactly. And again, the benefit of doing it with mods is you can do it as a fork sub-agent, and so it doesn't remain in the context afterwards. So you have this, like, idea of like, okay, the model is doing its execution and you have this almost like supervisor, like, that is like making sure that you can do like the next steps well. So yeah.Swyx [00:44:15]: Yes. I do have two panels and like I often try to have a supervisor thing, keep the high-level context and then the implementationThariq Shihipar [00:44:21]: YeahSwyx [00:44:22]: Detail in another agent.Vibhu [00:44:23]: I feel like a lot of this abstracts away as models change? The, like, half an hour ago you said bitter lesson of harness engineeringThe Bitter Lesson of Harness EngineeringThariq Shihipar [00:44:31]: YeahVibhu [00:44:31]: And we're on the other extreme right now, I feel.Swyx [00:44:33]: Well, so yeah, exactly. If everything's customizable, what is Claude Code, right?Thariq Shihipar [00:44:37]: Yeah.Swyx [00:44:37]: And which I talked to you about last night.Thariq Shihipar [00:44:40]: Yeah, I think that this is. I think the bitter lesson is unintuitive? In terms of like. Also, like we're misusing a little bit of the bitter lesson here where it's like, it's more about like scaling and compute and stuff. But like, I think there is something where it's just like. I think I use it as an approximation here to say that harnesses go out of date very quickly? And like how, but how they change is unintuitive? And so like the big obvious example is like from chat to like agents where you had to give them entirely new tools, right? But like, I think this new version of like, oh, it can modify its own harness, right? This is like, an own harness loop is like a way of using its capabilities, right? Or like it can build an artifact. And like, I think the way I think about it is like the models have more and more intelligence, and they're like so much more intelligent now than like the average software engineering task. Like, you look at the like terminal bench ones and they're like solve like the Jacobian conjecture. Not really, but like, it's like they're, they're quite complex. Like, I would not have been able to do this really as a software engineer.Swyx [00:45:42]: And you said TB4 or TB2?Thariq Shihipar [00:45:43]: TB3. TB3.Swyx [00:45:44]: TB3.Thariq Shihipar [00:45:44]: Yeah. They're quite complex, but the goal is still to deliver user value, right? And like you said, there's like this infinite space of things to do. And so the ways like you spend compute are to keep the user in the loop and make sure that like you're getting to the right decision in the end of the day and like the right output. And artifacts and mods are this way of like spending that intelligence. and I think that's like, yeah, the next step. And so, yeah, I think Claude Code is like, has the core things of agent loop which are, have gotten more complicated. It's like, it needs a sandbox to operate safely. It needs auto mode to like make sure like the permissionsVibhu [00:46:21]: Approvals.Thariq Shihipar [00:46:21]: Yeah, approvals. it needs computer use and MCPs and like all of these like ways of accessing your data, and it needs web search and web fetch. And like, so the-- as the models can do more and more, the core harness has to be like quite complex and very secure. But then like how you interact with it can change quite a lot.Vibhu [00:46:42]: What other harness engineering best practices have you, from the Claude Code team itself? I feel like, there was a phase of plan mode, which is not as used. We now have auto mode. at a point you cut the majority of the system prompt, you got rid of examples. What other best practices are there for harness engineering?Core Harness Primitives and Managed AgentsThariq Shihipar [00:47:02]: I think there is like a forking path where at some point, eventually, yes, the model will just be able to like vibe code the exact version of Claude Code, even describing all this complexity that I've talked about, right? Like auto mode and computer use and stuff. Eventually, the models will just be able to do that in one shot. But I think they can one shot simpler harnesses? And so like, I think some people. Sometimes you don't need this full, like if you don't need computer use or like all this like more complicated stuff. I think before we, you had to use things like the agent SDK, which was like Claude Code wrapped, in order to like. And I would, like suggest people do that because there was so much complexity into building a harness. And now as that's got more abstracted, we have like, Claude managed agents, which lets you have that complexity, but still like, right, like a very bare bones like harness that's scoped to your task. Yeah, I think there's like this barbell effect where like for like very complex, for like coding task and like these like complex things, you should use our harness. And then for like a lot of like simpler or like, more domain-specific things, you can build your own harness because Claude has gotten better at building harnesses, and we have these harness primitives like managed agents. So yeah.Swyx [00:48:18]: Yeah. Is there a general progression? Let's say chapter one was ultra code dynamic workflows, then chapter two was cloud mods. Where is this going?Swyx [00:48:29]: Where you're, you're, you can customize the thing on demand.Thariq Shihipar [00:48:36]: Yeah. I do think that like this evolution of projects and like artifacts and splitting out like brain and hands and, surfaces is like where things are going more. And like, I think it's like not all quite there. partially it's like a, it's just like more token expensive? And like, I think likeProjects, Local Hands, and Cloud-to-Local HandoffsSwyx [00:48:59]: Why would projects be more token expensive? I understand mods would be slightly more token expensive. No, not something I'm worried about.Thariq Shihipar [00:49:06]: Yeah.Swyx [00:49:06]: But whatThariq Shihipar [00:49:07]: You're asking Claude to do. It's like creating loops. Like you're asking Claude to do more work for you. And so like it's managing the sub-agents and reviewing it, versus where you would be doing that work normally. And so that's like gonna be a little bit more intensive, like. Outputting to an artifact is gonna be a little bit more token-intensive than, like, outputting normally. I don't think it's too much more, but like, it's like combining all of these together well, like I think we're, we're still working on like local hands and things like that, I think is like, yeah, where things are headed, yeah.Swyx [00:49:37]: Yeah. Claude and local is, handoff is very interesting. I was thinking about this as reverse cloud remote.Thariq Shihipar [00:49:44]: Yeah.Swyx [00:49:45]: Because it's like remote, it's you're handing off to cloud, but here the cloud is handing off to local, right?Thariq Shihipar [00:49:49]: Yeah, exactly. Yeah, remote control is also another way of doing it. And I do want to say this is like how I think about it and like what the things that I'm most excited about this, but like there are, just like lots of different ways to work with Claude. Like some people use remote control a lot, some people use Claude Code on the web a lot. Obviously, like at Anthropic, we use Claude Tag a lot, and like what's great about Claude Tag is we set up all this stuff for our own execution. And I do think if you're an enterprise, that's still the best way to go. but if you're like an individual, Projects is this way of like, getting some of that like niceness of Tag, which has like that like supervising agent and yeah, adding artifacts and stuff, but like without having that whole like admin setup. And so there will be many ways to use Claude, I think. I think it's probably not just one like single.Claude Tag as an Organizational HarnessSwyx [00:50:36]: You had the multiplayer thing here. Let's, let's just check in on Claude Tag. it's been about two-plus months. Lots of, public, adoption and trying it out.Thariq Shihipar [00:50:45]: Yeah.Swyx [00:50:45]: What's new? What's, what have you found since the launch?Thariq Shihipar [00:50:49]: Like, Claude Tag is how we useSwyx [00:50:51]: It's like 80% of your

Every product team is shipping AI. But the dirty secret is that, for many companies, very few users are adopting. The feature launches, the people who try it love it, and adoption stalls anyway. Our guest today thinks that's because teams obsess over building AI and barely think about adoption. Charanya Kannan is VP and GM of Navan Anywhere, which puts Navan's AI travel and expense tools inside Slack, Teams, and Gemini, where people already work, instead of asking them to open another app. She joined Navan when the company was still called TripActions and doing under $50 million in revenue... now it's well on its way to a billion. Her take: if customers have to work harder to use your AI, it isn't actually better. In this episode, Charanya shares: How Navan Anywhere was built distribution-first, bringing booking and expenses into the tools people already use every day Why Navan's margins went up in the AI era, while the rest of SaaS braces for compression And why she believes PMs who mostly manage process and Jira tickets will fade away, while those who can actually drive user benefit and business outcomes will matter more than ever Links LinkedIn: https://www.linkedin.com/in/meetcharanya/ Navan: https://navan.com/ Chapters 00:00 Introduction 04:17 Why AI features don't always get adopted 06:30 Building Navan Anywhere distribution-first 10:13 When to use AI vs. deterministic systems 12:01 Conversational cognitive load and why good UI still matters 15:57 Evals, distilled models, and the AI reliability pyramid 20:14 Product owners vs. product managers in the AI era 26:10 How Navan is growing margins while SaaS faces AI compression 31:40 Conclusion Follow LaunchPod on YouTube We have a new YouTube page! Watch full episodes of our interviews with PM leaders and subscribe! What does LogRocket do? LogRocket's Galileo AI watches user sessions for you and surfaces the technical and usability issues holding back your web and mobile apps. Understand where your users are struggling by trying it for free at LogRocket.com.Special Guest: Charanya Kannan.

Danny Peña and Parris Lilly dive into The Witcher 3: Wild Hunt Remastered! Over a decade after its original release, CD Projekt Red has delivered a major update packed with PC Path Tracing, DLSS 4.5 support, reworked combat animations, UI updates, transmog, auto-looting, and accessibility features.   In this episode, Danny and Parris break down their hands-on impressions across PC and Switch 2, discuss why the inclusion of Hearts of Stone and Blood and Wine for free makes this the ultimate definitive edition, and share their thoughts on the upcoming "Songs of the Past" expansion along with CDPR's target release for The Witcher 4.

It's an all new That Real Blind Tech Show as Allison, Brian, David, and Jeanine gather around the campfire to discuss the latest tech news and usual nonsense.   But before you listen go and get the That Real Blind Tech show iOS app. What are you waiting for? Grab it here.   After disclosing way to much personal information about someone, we kick the show off discussing the return of the Brooklyn Sewer Men, and of course Brian has a very interesting conspiracy people about all of it.   And we are only five weeks out from the 19th annual Laugh For Sight at Gotham Comedy club go get your tickets here.    Google is getting ready to send an A.I. not a Super Intelligence satellite up in to space next week. We ask what could go wrong?   Then we're talking baseball, and Dummy Hoy, Baseball's first deaf pioneer and the push to get Mr. Hoy in to the Baseball Hall of Fame.   We then dive in to our thoughts and review of iOS 27 and the issues that some of us have been having with the new operating system. Brian starts with an awful review of his experiences with the new Siri. Jeanine then dives in to her experiences with iOS 27. Brian then brings up the massive security issue in iOS 27, the entering of your password with VoiceOver and being told the wrong number of characters entered. Next up Brian discusses the disaster that Downcast has become in iOS 27. And then there's the Notification Center, yuck.   Jeanine then gives us her thoughts of Mac OS golden Gate.        The long rumored Apple Home Pod is rumored to be arriving next month. Would you want Siri runninng your household?   The article title says CahtGPT's Voice mode is getting some big upgrades, yet we couldn't find anything big about this at all.   The Colonel, Bernie Sanders is proposing the elimination of the development of all Super Intelligence and jail time for people who do develop it. And yes, we hate the usage of Super Intelligence.   Remember the Rabbit R1? We do, well now you no longer need the defunct device to run their software.   Meta launched Muse their new A.I. about a month ago, it caught some of us by surprise, we're hearing it's not bad, and that it is currently the most downloaded app in the App Store. Allison then fills us in on what she has been using Muse for and how it has worked.   Brian then dives in with a disappointing new UI design of Claude.   Muse did not get off to a banging start on the Mac as it launched with a one click vulnerability. Is that bad?   At Meta Connect, Zuckerberg ignored all the idiot Privacy Wonkers out there, made us happy.   Meta announced the new Meta Ray Ban Perv Gen 3 starting at $249.   Meta then announced a pair of cameraless Ray Bans, and on behalf of That Real Blind Tech show, if you know a blind person that buys these, please slap them on our behalf.      Meta then closed with the announcement of the Meta A.I. Charm, and since these devices will be able to communicate with other Charms, and will have cameras, Privacy Wonkers must be spiraling even more.   And it's more of Watcha Streaming, Watcha Reading.   To contact That Real Blind Tech Show, go download our new That Real Blind Tech show app, and please for the love of all that is holy, rate and review us here.    you can email us at ThatRealBlindTechShow@gmail.com    join our Facebook Group That Real Blind Tech Show, join us on the Twitter @BlindTechShow

A weekly news show informing you on the latest in Bitcoin, privacy and open source tech, hosted by Ungovernables, Max and Q.AOBMax: flooded the downstairs after leaving a sink running over the weekend, spent Saturday ripping out old timber and clutter he'd been meaning to clear out anyway, then a full kids' day that left them sick afterwardsQ: spent the weekend building his own local, voice-controlled AI assistant after watching the new Spider-Man filmQ: read this week's letter #7 from Keone, "Notes from the Inside", posted on The Rage -- a tough one on conditions moving between facilities; audio version out later this week, timed with Rick's Free Samurai prize drawQ: enjoyed last week's Freedom Tech Friday listener questions episode, ranging from Bitcoin to Monero to privacy, multisig and local AINEWSResearch lab [alloc] init, founded by Misha Komarov with Clara Shikhelman as Head of Protocol Research, published Shielded Bitcoin, a proposal for Zcash-style private transfers needing no soft fork; shielded transactions are encrypted data blobs carried in OP_RETURN or witness data that Bitcoin only orders and timestamps, while separate indexer software checks zero-knowledge proofs and stops double-spends; value sits in encrypted "notes" spent by publishing a one-time nullifier that proves the spend without revealing which note it is; the peg moving real BTC in and out isn't designed yet and rests on an unproven witness-encryption scheme called PIPEs v2; CoinDesk reports fees around 4x normal, a trusted setup and no launch date -- Bitcoin Magazine, CoinDesk, allocinitBlockstream published its own post-mortem of the 6 September Liquid exploit: a rangeproof-cache flaw dating to April 2018 ("Bug A"), responsibly disclosed 2 August and patched by 11 August, introduced a second flaw ("Bug B") where unprefixed cache keys let two different proofs collide; the attacker used it to mint about 4,000 unbacked LBTC and peg out 3,996 BTC through SideSwap, draining the reserve from about 4,205 BTC to 197 BTC; 3,400 BTC was returned and the network resumed 9-10 September, but about 602 BTC remains with the attacker and peg-outs stay paused with no date; the 11-of-15 federation multisig worked exactly as designed, the failure was in the software deciding what counted as a valid transaction -- BlockstreamBitget detected unauthorised transfers from its hot and warm wallets at 18:31 UTC on 24 September and froze withdrawals platform-wide; CEO Gracy Chen says the attacker compromised a backend system, spoofed transaction data and triggered Bitget's own approval process, with private key compromise ruled out; the loss was revised from $351.6M to about $387.5M once Zcash and TRON assets were counted, mostly ETH, TRX and USDT with no bitcoin taken; Circle and Tether froze about $318K, while roughly $83M in native XRP moved beyond Ripple's power to freeze; North Korea attribution comes from Bitget, Elliptic and MetaMask's Taylor Monahan, not yet from any government; Bitget says its User Protection Fund covers the loss and is reopening withdrawals in phases from 28 September -- CoinDesk, TFTC, Bitcoin Magazine, CoinDesk (freezes), CoinDesk (XRP), BitgetAMLBot traced part of the Bitget haul from TRX to USDT, bridged to Ethereum, swapped to about 145 ETH, then through THORChain into about 4.59 BTC, with roughly 4 BTC of that linked to an unnamed Wasabi coinjoin round and the addresses "blacklisted"; most of the stolen funds have not moved -- AMLBot on X, crypto.newsMatt Morehouse disclosed two denial-of-service bugs in Eclair v0.13.1 and earlier, fixed in v0.14.0: oversized feature-bit init messages could allocate about 300MB per message and crash a node, and zlib-compressed channel queries could inflate 64KB into 64MB; his smite fuzzer found the first, an LLM-assisted search for similar code patterns found the second -- Delving BitcoinLightning Labs published four security advisories: a High-rated bug let an invoice be marked settled after an interceptor had already cancelled the HTLC, affecting tapd 0.5.0 and earlier and lnd 0.18.4 to 0.18.5; three Low-rated DoS bugs covered a gossip stall, a panic on a malformed DNS seed response, and memory exhaustion from Brontide write allocations; nodes on current lnd (0.21.3 or 0.20.4) are unaffected -- Lightning Labs securityGalaxy's Alex Thorn disclosed that 52.37 BTC from weak-entropy Coldcard addresses, about 2.8% of the total taken and roughly $4.5M, had been moved into an address belonging to a Wyoming "Crypto Recovery Trust" carrying an OP_RETURN reading "claim:cryptorecoverytrust.com"; the trust says owners can reclaim coins by proving control, but the white-hats are unnamed and its legal documents are unverified -- CoinDeskSEC Commissioner Hester Peirce announced she resigns effective 2 October after nearly nine years to join Regent University School of Law, leaving the SEC with two Republican commissioners and no replacement nominee named; two days earlier, at SIFMA's Digital Assets Conference, she argued for replacing KYC document collection with zero-knowledge proofs and attribute-based credentials, telling regulators "we build ever bigger data haystacks on the theory that we will find a needle or two inside" -- CoinDesk, TFTC, TFTC (speech)New York Attorney General Letitia James and Governor Kathy Hochul sued Polymarket US in state court on 24 September alleging unlicensed gambling and underage betting, seeking at least $4.6B in fines; Polymarket moved the case to federal court and countersued, arguing the Commodity Exchange Act gives the CFTC exclusive authority; the next day a unanimous Sixth Circuit panel ruled Kalshi's sports contracts are subject to state gambling law, splitting with the Third Circuit and making a Supreme Court case more likely -- CoinDesk, CNBC, CoinDesk (Kalshi)BitMEX stopped trading, deposits and new positions at 04:00 UTC on 23 September after 11 years, with API withdrawals ending 28 September and website withdrawals staying open; from 1 October verified accounts with a balance pay the greater of 1% a year or $50 a month; owner HDR Global Trading cites a strategic review, with no legal or regulatory issues behind the closure -- CoinDeskThe x402 protocol, reviving HTTP 402 "Payment Required" for machine payments, merged Ben Carman's spec for paying with Lightning: the server issues a BOLT11 invoice whose description hash commits to the exact request, the client pays and returns the preimage, and the facilitator checks it against the payment hash and invoice signer without querying the receiver's node -- GitHub PR #2861RELEASESAm I Exposed v0.36.0 -- 2026-09-26Detects Whirlpool tx0 premix outputs and Wasabi 1.x coinjoins it previously missed or misgraded, and flags input-side address reuse as a leak instead of scoring it as good.Jam v2.0.0-beta.4 -- 2026-09-24Fourth beta of the JoinMarket web UI: adds Sign Message, pins the exact UTXOs shown when sweeping, and lets you freeze or unfreeze several UTXOs at once.Shhark v0.8.1-shhark-preview.5 -- 2026-09-12A self-hosted, privacy-focused Ark wallet preview adding optional Payjoin v2 on Signet, Tor-only networking that fails closed, Silent Payments and experimental post-quantum messaging.ZEUS v13.2.2 -- 2026-09-23Stable release embedding LND v0.21.3 with SATS Routing and Coinos as swap providers, plus a critical fix moving iOS wallet data out of iCloud-synced Keychain.Blockstream Green Android 5.7.0 -- 2026-09-24Adds manual coin selection filters, transaction notes and a price chart, and restores 2-of-2 and 2-of-3 multisig account creation.umbrelOS 2.0.0 -- 2026-09-22Stable release of Umbrel's home-server OS adding a Photos app, multiple user accounts, virtual machines, FailSafe RAID storage and a redesigned App Store.SignerOS v1.3.0 -- 2026-09-23Fixes a bug where one valid input in a multi-input transaction could make a fake change output look legitimate; every input is now checked against the actual cosigner keys.Blockstream Green Desktop 3.6.0 -- 2026-09-21Adds paying to Lightning addresses and LNURL-pay from the send flow, redesigns manual coin selection, and restores 2-of-2 and 2-of-3 multisig account creation.Everything ElseAmber v6.6.5 -- 2026-09-21Nostr signer: relay backups are now encrypted with a separate derived key, previously readable by any app with a remembered decrypt permission.Arkade TS SDK 0.4.76 -- 2026-09-25Developer SDK patch release for the Arkade (Ark) protocol, following 0.4.75 earlier in the week.Bisq Easy (Android) 0.14.1 -- 2026-09-26Security release: embedded Tor updated to 0.4.9.13, closing high-severity Tor issues, now built from Bisq's own Tor fork.Sister app: Bisq Connect 0.10.0 (2026-09-26), same Tor upgrade.Breez Spark SDK 0.26.0 -- 2026-09-23Adds receiving USDT and USDC, and instant or expedited claims for on-chain deposits.Cashu TS v4.11.0 -- 2026-09-22Backported fixes: melt preimages checked against the invoice hash, requests default to a 5-minute timeout, and closed subscriptions report an error.cln-nip47 v0.2.1 -- 2026-09-26Nostr Wallet Connect plugin for Core Lightning. Dependency updates.Core Lightning v26.06.8 -- 2026-09-22Security release fixing responsibly reported vulnerabilities, confirmed to include the dual-fund drain reported in issue #9498. Upgrade.Ditto v2.42.2 -- 2026-09-27Nostr social server: verified-link badges, muted users blocked from push notifications.Also in window: 2.39.2 to 2.42.0 (posting streaks, emoji packs, push).JoinMarket-NG 0.40.0 -- 2026-09-27BIP-329 labels now distinguish coinjoin output, coinjoin change and deposits; adds PSBT v2 signing and warns when the wallet daemon listens in plaintext off localhost.

In questa puntata parto da un caso concreto nato preparando una sessione su Blazor e AI: cosa succede quando un processo non può più vivere comodamente dentro una singola request?Da Task a IAsyncEnumerable, fino a processi in background, Channel e SignalR per aggiornare la UI: il punto non è complicare l'architettura, ma scegliere il modello giusto in base al lifetime del processo.Perché una REST API va benissimo in moltissimi casi… ma non tutto deve necessariamente vivere dentro una request HTTP fino alla fine.#dotnet #aspnetcore #blazor #restapi #signalr #backgroundservice #channel #iasyncenumerable #softwarearchitecture #softwaredevelopment #developers #dotnetinpillole #podcast

The Fork In Your Ear Ep#222 "Ionic The Hedgehog" - Podcast Show Notes & Summary 9-26-26 Tim and Nate spend the first ten minutes fighting a Discord refresh, a hiss they blame on forced suppression, and a Windows accessibility crosshair that PowerToys cannot kill. Then the ring lands on Life: first day of fall floods Tim's plugged gutters, a USPS driver does an Austin Powers turn in his driveway because Washington will not let carriers reverse more than a foot, and Nate trades a Mustang insurance bill for a 2019 Hyundai Ioniq they immediately name Ionic the Hedgehog. Dream Nate gets a tattoo without telling Mrs. Foo. Amazon cargo e-bikes with fake pedals unload patio furniture in Burbank heat. Games dump after the last episode posted — Diablo 5, StarCraft Dominion, more No Man's Sky, Famitsu wishlists, Kojima's canceled Metal Gear-shaped "Fizzint," Bungie saying "we're listening" again, and Xbox shoving Halo under an Activision Call of Duty team while Halo Studios shrinks to "community support." Tim's HTML RTS playtest is up to 0.58 with dumber-AI rehab. Tech is iPhone Duo vs Z Fold 8 and a missing telephoto that keeps Tim off the fold. Entertainment is Lanterns going feral, Mayday being exactly Ryan Reynolds, Paramount-Warner getting waved through, Darkwing Duck coming back without Jim Cummings, Kara's Spaceballs DVD, and a Reddit stranger who archived Wizard's Rule episodes 0–40. Get forked.   Detailed Show Notes  

Earlier this month, world model company Runway introduced GWM Worlds 2, a research preview that “turns high-fidelity video and audio generation into real-time interactive simulation.” Runway calls this an “autoregressive diffusion” model; with autoregressive describing how it generates over time.One new feature in particular caught our eye: WorldPrompt, a proposed input format for specifying a generated world and the actions within it. It allows you to fix some aspects of a simulated environment — including the first frame — and then create a series of timestamped events. The events, or actions, can even be prompted in real-time.To understand the implications of WorldPrompt, we spoke to Kamil Sindi, Runway's CTO, and Robin Kahlow, its Principal Research Scientist for generative video and multimodal AI. We also have exclusive comments from Anastasis Germanidis, co-founder & co-CEO of Runway, courtesy of a podcast swyx and Vibhu did with him.Who's building real-time interactive world models?First, some context about world models that can generate interactive video and audio in real-time.Runway is reportedly valued at $5.3 billion, based on its most recent fund raise of $315 million in February. Its first release, GWM Worlds, was launched last December.Alongside Runway, there are several other notable projects in this domain: Google DeepMind's Genie 3 (which also generates at 720p and 24 fps), Odyssey-2 Pro, and World Labs' RTFM (Real-Time Frame Model). We've summarized their differences in the following table:Given the complexity and massive latency demands of real-time video and audio generation (which we'll get into below), all of the projects listed above have limitations. For instance, Google notes that Genie 3 “can currently support a few minutes of continuous interaction, rather than extended hours.”But as our interviews with Runway show, real progress is being made.The central idea of WorldPromptWorldPrompt, a new feature in GWM Worlds 2, helps differentiate Runway from its competition. You can think of it as a control layer for characters, cameras and the environment. As Kahlow put it, it's a way to “control all the different subjects in the world” — similar to a computer game.“Like, if there's an NPC [Non-Player Character] somewhere, the NPC might walk up to you and say something. So you could achieve the same thing with this kind of model, where you can have very detailed control over everything in the scene.”As the name suggests, WorldPrompt is a prompting mechanism — not a programming language. So, unlike virtual world games like Minecraft or Roblox, GWM Worlds 2 doesn't offer scripting capabilities or the ability to control state. But there's a power to that, as Sindi pointed out.“You can create promptable worlds on-demand with video and audio in sync, across all these different domains and environments. That's not a distant-future hypothetical thing,” he said.But there are also limitations to prompting a world model. We asked how reliably the model would follow an instruction to create, for example, a law of gravity or a certain ability in a character?“Yeah, so it's a research preview,” Kahlow replied. “So it's not perfect, of course, and there are still flaws. It really depends on how difficult the action is. I would say movement works quite reliably.”Sindi added that more training plus scaling the data and models is resulting in “better following.”How a video model becomes a real-time runtimeDespite the current limitations of GWM Worlds 2 — especially if you compare it to pre-designed and scriptable worlds like Minecraft or Roblox — the true promise of world models like Runway is that they'll eventually lead to fully self-generated, real-time games and experiences. Which is an extremely hard engineering problem, as Kahlow reminded us.“There are two challenges. One is making the model not generate a whole clip at once. So instead, you want it to generate frame by frame while you're looking at it. And the other challenge is actually making the generation fast, so you can play it in real time.”GWM Worlds 2 offers real-time interactive worlds streamed in continuous 720p video at 24 frames per second (fps) and audio at 48,000 Hz.Runway achieved this firstly by taking its foundational audio-video generation model and fine-tuning it to the new WorldPrompt format, so the model can follow that. It then post-trains the model to generate autoregressively.“And after that, we work on making it real-time through distillation methods,” Kahlow added.Co-CEO Anastasis Germanidis offered more technical details in our podcast with him. He told us that the process starts from “bidirectional diffusion that basically generates an entire video at once and [makes] it autoregressive.” This allows the model to “generate one frame or a few frames at a time.”Germanidis described two possible forms of distillation in order to make it real-time: distilling a larger model into a smaller one or reducing its diffusion steps. As a general example, he said a model might go from around 50 denoising steps to four, with some quality loss but potentially comparable results.The challenges of real-time generationGermanidis admitted that there were issues with how it generates real-time interactive video.“The biggest challenge with autoregressive models is error accumulation,” he said. “You're feeding generated frames back into the model to generate the next frames, and if there are any small errors, they accumulate over time.”Sindi told us there are also challenges dealing with “infinite generations” of content.“There's all these challenges around what context to keep, what to discard that's not important. And so there's all these optimizations we have to think about, so we're not blowing up our GPU memory.”Another current limitation is long-term memory. “The model does not have perfect memory,” Kahlow said. “That's still an open research problem.”Causality and correctnessWhile performance is the primary challenge for Runway at this time, its world model also has to produce plausible consequences when a user takes different actions.Germanidis used the example of simulating football; he pointed out that online video training data contains more successful goals than failed goal attempts, so a video model might render the first more convincingly.“If I take this action versus this action, you want it to generate equally realistic outcomes,” he told us. “That's, I think, the big gap between video models and world models: that idea of counterfactual generation.”Sindi told us that evaluation gets harder the more complex interactions get.“If you have this multi-prompt, multi-character, multi-scene [environment], how do you really understand what was causal and what was not?”To try and solve that, Runway has some automated verifiable tests. But since GWM Worlds 2 is a research preview, Kahlow noted that doing tests yourself is also advisable — “trying out your model to see what doesn't work is really important.”More than gaming — there are agent use cases tooGaming is the obvious use case for what Runway is building, but there are others. Kahlow mentioned robotics — for example using a simulated environment to test how a robot works.Another, more intriguing, use case is to use it to test agents at scale.“Having thousands of simulated environments is much less challenging if you have a suitable model like GWM Worlds,” Kahlow said.But how does an agent know what's changed in the world — is there a structured state that it can read, or is it just the generated video and audio that it's consuming and understanding?“So there's no structured state here,” Kahlow replied. “It's just observing the same thing you might observe in real life, just [in this case] from cameras.”Sindi noted that GWM Worlds can also be used for “synthetic data generation for agents.”Finally, Germanidis suggested there's potential to use these world models alongside reasoning models.“You're maybe using some reasoning [for] planning of the scene, and then you're passing it into the diffusion head that's actually generating the pixels.”Anastasis Germanidis* LinkedIn: https://www.linkedin.com/in/agermanidis/* X: https://x.com/agermanidisTimestamps00:00:00 Introduction00:05:17 Runway's Origins and the Bet on Generative Video00:12:23 The Stable Diffusion Story00:18:44 Gen-2, Controllability, and the Weekend Hack00:23:02 From Video Generation to World Models00:28:03 Learning From the World, Not Just Language00:35:04 Sora, Runway's Existential Crisis, and Gen-300:39:39 Why Real-Time Video Is Inevitable00:43:06 Interface World Models: Software Without Code00:50:25 The Fully Neural Operating System00:55:11 World Models for Robotics01:02:32 Robot Policies and World Action Models01:07:47 The Lucid Dream Test01:11:41 Video Agents and Omni Models01:23:12 Artists, AI, and Creative Workflows01:27:14 Physical AI and the Future of World ModelsTranscriptIntroduction: Runway, Creative AI, and the Early ThesisSwyx [00:00:00]: Okay, we're here with, Anastassios from Runway, with, me and Vibhu in the studio. Welcome.Anastasis [00:00:08]: Good to be here.Swyx [00:00:09]: Congrats on all your success and progress with Runway. You're opening offices all over the world. Did you envision this when you first started out?Anastasis [00:00:16]: Not quite. I think even when we started, we had this idea that, It was more a matter of when, not if, we were seeing the early generative models of 2016, 2017, and just extrapolating, assuming, we resolution, quality increases predictably over time. There's gonna be a point where most of content will be generated, and that was maybe the initial thesis of Runway was we will need, as a result of those generative models, rethink how creative tools are made. and as we built out the research behind, our generative models, it then became clear that they were useful far beyond that as well.Anastasis' Background: Art, Simulation, and Machine LearningSwyx [00:00:57]: And it is more obvious now with, like, the real-world stuff and the world models that we'll talk about later. I'm just kinda curious how you go from a background in, like, Zocdoc and, computer vision into Runway. Like, take us back to that early conversations with Chris and, whoever else is on your founding team.Anastasis [00:01:14]: I was always splitting through those two worlds. One was the I had my own art practice. I was making a lot of interactive art, I think for a long time. and then on the other side, I was working in startups, and I was working as a ML engineer, as a backend engineer at different companies. I've always been interested in, coding and computation, and especially interested in simulation and brought it back into my early artwork as well. And at the same time, I was interested inSwyx [00:01:43]: The personal site has a few, right?Anastasis [00:01:44]: Yeah.Swyx [00:01:45]: Is there one that we should pull up? Just in case there's something that's like. I just like to go down memory lane.Anastasis [00:01:50]: Yeah.Swyx [00:01:50]: Okay, what is this?Anastasis [00:01:51]: So this was, a project that I made, I think back in 2015, where I built this software that would give, voice instructions to people in a gallery space. So it would coordinate interactions between people. And so it will first give you an identity, like you're an, architect, you're 30 years old, and, you like sports. and then it would match you with another person, and you have this completely generated interaction. language models were not quite there at the time, and so it was it was a mix of some templates and some, like, some Markov chain-generated text, and it would just completely simulate these small talk conversations between, everyone in the gallery space. so was always very fascinated on the one hand with, generative models and, like, the early machine learning work that was being at that time. But at the same time, there was this separate thread of simulation and what it means. Like, what can we learn about humans by creating those very simple models of their interactions and their behavior?Early Generative Art: pix2pix, GANs, and Uncanny ValleyVibhu [00:02:56]: Did you generate the prompts or, the 30-year-old, whatever? Was it you generating them? How'd you, how'd youAnastasis [00:03:03]: Exactly. So the program would just generate- those, from. Yeah, a lot of it would be Mad Libs style of justVibhu [00:03:10]: YesAnastasis [00:03:10]: You have lists of different professions, lists of different,Vibhu [00:03:14]: HobbiesAnastasis [00:03:15]: Personality types, lists of different, ages, things like that. And then it would just combine those things together. And then maybe the next project we go is, Uncanny Valley, Uncanny Road, which wasSwyx [00:03:27]: GansAnastasis [00:03:27]: One of the first projects that, we built with, one of my two co-founders, Chris. This was taking, pix2pixHD, which was one of the early image-to-image models that NVIDIA released back in 2016 or 2017. and it was a model that would take a semantic map of a scene and then generate a photorealistic, let's call it, output. very early days, so it was not very high-fidelity outputs, but it w I think was the first image-generation model that could generate at 1K resolution. And it was all trained on self-driving datasets. So the semantic categories it would support were only, things you would encounter on the road. So it would be pedestrians, traffic signs,Vibhu [00:04:16]: StoplightsAnastasis [00:04:17]: Bikes, stoplights. And so that was one of our first indications that we built this and people were making all this, like, very surreal imagery of, yeah, a million plus a million pedestrians or a million traffic signs or, like, gigantic humans. And it was a indication that you could take a model that was trained on this very boring dataset, essentially, of, like, not that many interesting things happen when you're on the road, and then you can repurpose it and go very out of distribution and make something that was artistically compelling. And that was It's a summary of the thesis of Runway in some ways, that you can take the same generative models, and if you look at them from another direction, if you build interesting tools around them and you give them to artists, they're gonna do things that you don't expect.Vibhu [00:05:02]: Very cool. I like the, UX of it. You're just given an empty canvas, try whatever, do whatever. And then the other one, like, you see everyone with wired headphones? Like, that's, that's a sign that it's, it's veryAnastasis [00:05:16]: The AppleVibhu [00:05:17]: YeahAnastasis [00:05:17]: Apple, your version.Vibhu [00:05:17]: Original ads. Yeah. Take us to today. You've been doing this for seven years at Runway. How have we got to this? Like, how do we go from driving simulator data to all this? And you cover the whole stack of generative media?From Creative Tools to a Research LabAnastasis [00:05:33]: Interestingly, we're almost back in, we're, we're full circle. We're, we're now applying our models and beyond creative tools into real-world scenarios. But it was a, it was a long journey. It was very early on we realized the first version of Runway was a way to easily use the, all the open source model of the day, things like pix2pix to. and give them to artists. That was the initial idea, is those models are too difficult to use if you're not a machine learning engineer. Like, what happens when you give them to artists? Very quickly, we realized we needed to build a research org, inside of Runway, and that happened maybe on year one. And, a lot of the mandate there was. The image-generation models of the time, the video generation models of the time, or there were barely any video generations all the time, but they were not quite there where they could be productionized and brought into tools that would be part of creative workflows. so we need to push the frontier of the research. And so maybe the first four years of Runway, research was almost happening on the background until there was a moment in 2022, with latent diffusion, with, DALL-E 2, where, there was that step function change, and you guys maybe remember around the time.Swyx [00:06:49]: I started in this space because of latent diffusion and Stable Diffusion.Anastasis [00:06:54]: Yeah.Swyx [00:06:54]: Because I was like, “Wow, this is not only, like, feasible, it is doable on consumer hardware.”Anastasis [00:07:01]: Exactly, yeah.Vibhu [00:07:01]: I think the delta is also huge. Like, I learned pix2pix. Like, this was intro to ML, the TensorFlow, like, Jupyter, Google Colab notebooks were like this, and then you have a sudden step function change, with diffusion and whatnot. Any other ones since that. Like, there were clear examples of what early diffusion were to get to here. Any other changes in key technology research?Green Screen, Rotoscoping, and Early RunwayAnastasis [00:07:26]: Between, 2018 when we started and 2022?Vibhu [00:07:29]: Yeah.Anastasis [00:07:29]: So one of the early work that we did in Runway was solving segmentation, image and video segmentation. It was a very important problem because most VFX involves essentially separatingSwyx [00:07:42]: RotoscopeAnastasis [00:07:42]: Subjects. Yeah, rotoscoping. Extremely manual process. Nobody enjoys doing that. and so a lot of the early days of Runway was building this tool. It was called Green Screen, and it was for a long time the main thing that people were using Runway for. It ended up being used in, Everything Everywhere All at Once and a bunch of other high-visibility films and series. But that was essentially, Runway for a long time was a post-production tool until latent diffusion and generat- Gen-1, Gen-2, happened.Swyx [00:08:12]: Cool. let's, let's go past that moment. You've come a long way. Then you started releasing your own models. Maybe describe that journey as well.Scaling Video Models and the Bet on 1,000 A100sAnastasis [00:08:20]: Yeah, so we go to the other point, yeah, in mid-2022 when it became clear that we're doing research at a fairly small scale of compute, and it became clear that, like, scaling laws would apply to, image and video gen in the same way that we're applying to language generation. So we made a big bet, and I think at so at the time, we signed this deal to build a cluster of a thousand A100s, which at the time we were a Series B startup. That was a almost, slightly irrational decision maybe, but we really believed that if we trained a video model at a large scale, we would get, like, a great model at the end. And at the time, the goal or we set the goal around fall of 2022 of what is, what does the latent diffusion, Stable Diffusion moment look like for video? And at the time, the best model of the time was called CogVideo. it was one of the early video models. It was very 256 by 256 resolution, very not very high quality. and so we decided we're gonna build out this cluster, and we're gonna just invest in, like, in building out our own video model. it became clear as we're training Gen-1 that it was difficult to get to fully. we wanted to build text-to-video, but it became clear to us that an easier starting point would be to start from video to video. Because when you have a stronger conditioning, it's, it's an easier problem to restylize an existing video versus generate the video from scratch. And so we released Gen-1 first back in, it was January of, 2023. Yeah.Vibhu [00:10:04]: It's just a fun visual podcast, honestly. Like, if we can see February 2023, what was the state of stuff?Gen-1: Video-to-Video and Depth ConditioningAnastasis [00:10:10]: It's so interesting ‘cause at the time when you see those results, you think this is so incredible, and this is like, it's almost like image generation or video generation is solved. And then you look back a few years after, and it's like, it's It's just like you get used to the results very quickly, with those models. But at the time when we started seeing those results, it was, it felt quite incredible, and the level of, like, quality that you could get. And, so the Gen-1 was a depth-conditioned video model, so it would turn. it would take a input video, it would predict. it would it would first convert it into the depth map, and then we would generate, pixels with a latent diffusion model.Swyx [00:11:01]: Yeah, very effective.Vibhu [00:11:02]: Yeah. I didn't realize how distracting the blog post would be. Sorry.Anastasis [00:11:05]: Yeah, but, one of my favorite examples of on those, on Gen-1 was both, if you go up to mode three or mode two, there was this storyboard use case where people would makeVibhu [00:11:18]: OohAnastasis [00:11:18]: WouldVibhu [00:11:20]: You can mess around with theAnastasis [00:11:20]: Make a city out of books or out of boxes, and then they would shoot a video with their phone and then translate it into a photo-photorealistic output. There was all these ways in which those models were starting to be used for storyboarding and also for really. and then if you go to mode four, like, of taking untextured 3D scenes and then turning them into photorealistic output. So we saw a lot of use cases early on where people that were familiar, were power VFX editors would just take a blender, render, and then they would get translated in with Gen-1 or create a scene in Unity and then take a capture a video of it and then translate into, restylize it. So I still think video to video is powerful. I think we had a recent video-to-video model as well, and it's one of my favorite ways of using those models is essentially using them to use ground truth video as, like, the initial inspiration and then translate into different styles or different outputs.Stable Diffusion, Stability AI, and Open SourceSwyx [00:12:23]: But I think we're gonna go into, like, the rest of Runway and catch people up to speed today. I did wanna cover the, let's call it the Stable Diffusion controversy, or, what happened with Stability AI, whatever. I think there was a two sides of the story. I think there's part of that is a normal thing of, like, people, join and leave companies, but what is the, retrospective now that, there's been some years behind it?Anastasis [00:12:49]: Yeah, it's a very, it's a very long story to go into. I think it wouldSwyx [00:12:53]: Which I remember you wrote a really long post about.Anastasis [00:12:56]: We would probably cover the whole hour to go into it in more detail. But, essentially, there was the latent diffusion paper that came in, I think that was at the end of, 2021. And then Patrick Esser, who was one of the researchers behind, latent diffusion, and he worked at Runway at the time, he built latent diffusion in collaboration with Robin Rumbach and a few other folks back, in the in, CompVis, which was, a labSwyx [00:13:26]: Like a research group, yeah.Anastasis [00:13:27]: And, after releasing the early latent diffusion model, they, essentially they were. the goal was to keep working on versions of the model, scale it up, incorporate new data, incorporate new tasks. And Stable Diffusion was the same model, but trained on more compute, and then with a few more tricks, like a classifier-free guidance paper came at some point, I think in the early 2022. And thatSwyx [00:13:52]: Which, like, was a big prompting improvement.Anastasis [00:13:55]: Yeah.Swyx [00:13:55]:?Anastasis [00:13:56]: That improved results. it was trained on better data, so like, the esthetic subset of LAION, but it was effectively, the same underlying architecture. And there was that big training run, that, happened on Stability's cluster. Stability financed that run. And looking back at that story, I think it was the work to build and train that model was done. It was a, it was a research project. It was done as part of, like, continuation of the latent diffusion work. It then, I think it the model became very successful, and it, I think there were the. And I think as a result of its success, other companies tried to, figure out the commercialization path for it. But for us, it was very important that we try to, we make sure that we. It was meant to be an open source research project, and so the we decided that we should continue releasing versions of it, since that was the original goal of Stable Diffusion, and that led to releasing Stable Diffusion 1.5. There was maybe a day of, a bit of, miscommunication there, but ultimately that was resolved very quickly within hours. so yeah, there wasSwyx [00:15:12]: OkayAnastasis [00:15:12]: Not a niceSwyx [00:15:13]: I just wanted to. you have toAnastasis [00:15:15]: Yeah.Swyx [00:15:15]: You're one of the main players in that journey, and so it's nice to hear from the source of, like, what happened. Yeah.Anastasis [00:15:22]: Yeah. I think it's all, it's all in the past nowSwyx [00:15:26]: YeahAnastasis [00:15:26]: I would say. and, like, both companies, Stability took its own path, Runway took its own path.Swyx [00:15:32]: Yeah. There's still. James Cameron is backing the new Stability, whatever they're doing with the Hollywood studios.Anastasis [00:15:38]: Right.Swyx [00:15:38]: I don't know what they are doing. I think one thing that impresses me, and I'm happy to move on, is that back in the that time, let's say, like 2021, 2022, there was this community of people that you were involved in that was researching all this stuff, right? And, like, from everyone I talked to who was active then, it seemed like it was fairly obvious that somebody would do the hero training run that would produce Stable Diffusion. So, like, I guess the question is, like, you had the you were you had made investments. You were you had the foresight. Is it accurate to say, like, that is reflective of, like, what people were thinking at the time? Or was it still very much like, “Well, we'll use it as, like, a post-production tool or something. I don't know.”? Like, where in the sentiment were we that maybe you can think back to, like, what the community was like back then?The Early Creative AI CommunityAnastasis [00:16:28]: I reminisce and I think very fondly those early years, from like 2018 to 2022, because it was a very small community that, as you said, were very convinced that this was gonna be a big thing. And at the time, anyone who. Because it was such a small circle and, everyone who would, like, be part of that circle and, like, make projects with it would, immediately get, go viral. so likeSwyx [00:16:55]: And you didn't know who they are, right? They're just some name on a, GitHub or Hugging Face somewhere.Anastasis [00:16:59]: Exactly, yeah. So I remember one of the first big viral moments of creative AI was, there was the neural style transfer paperSwyx [00:17:09]: HuhAnastasis [00:17:09]: ThatSwyx [00:17:10]: Something dreaming?Anastasis [00:17:11]: I think it was called neural style transfer.Swyx [00:17:14]: Okay.Anastasis [00:17:14]: There was also Deep Dream, the puppy sliceSwyx [00:17:16]: YesAnastasis [00:17:16]: Which was, also really cool. but, yeah, there was this project that, Jim Kogan, who was an early advisor of Runway and one of those,Swyx [00:17:25]: Marketing guysAnastasis [00:17:26]: Big, creative AI, folks, he literally just, like, showed a video of himself taking the New York Subway and going over the Williamsburg Bridge and then stylized it with, I think in the style of Van Gogh or, like, one, painter. And that was. Like, at the time, that was, like, so cool and it went viral and it was completely revelation to people that you could do this with generative models. And that was only, it was less than. It was maybe 10 years ago. So just, like, as an indication of, like, how quickly things have gone.Vibhu [00:18:02]: It's pretty crazy. Like, even since then, you've got people at every level of the stack. You've got devs, creatives, artists, hobbyists. You've got everyone using it. And for people that tried stuff early, they'll remember how hard it was to use regular diffusion, right? Like, nowadays, you can use your favorite ChatGPT image gen or whatever, give a sentence, get a beautiful output. But diffusion was like, the whole ultra HD, 4K, high resolution. Like, prompting these things was very different. anything you learned on the tooling side, like from the offerings you guys have now, so like creatives, devs, you really took the. Research and brought it to everyone to use. anything interesting there to share?From Gen-2 to Controllable Video GenerationAnastasis [00:18:44]: We had to build the entire model serving infrastructure for video diffusion models. There was nothing else, already, like, because we had Gen-2 was the first text-to-video model, I think, out in the market. So many things that we learn over time. I think the I think the biggest one was, like, we. it was very clear early on that text-to-video was not gonna be the answer. Like, you. Like, people wanted a lot more control than that, and so we invested in, like, control building on top of those models very quickly. how do you use the camera trajectory as control? How do you use an initial input frame as control? So that was a very early learning for us. With text-to-video was, like Gen-2 was an amazing, step function improvement in the quality of video models, but it was used much more in an exploratory way because there was nothing to ground it to. There was no reference that you could bring into it. There was no. You couldn't really control the camera motion. You couldn't control the object motion. And so the first year, in 2023, was really all about what are all the interesting ways in which we can condition those models? And it was a lot of just post-training rounds on top of the base model to figure out, like, what, -- how do people wanna control them? And so there was, like, this quick succession of the we it was called Motion Brush, which was you could, like, you could draw arrows and dictate where things should move in the scene.Vibhu [00:20:09]: That's so cool.Anastasis [00:20:09]: There was camera control that was you could just describe, like, how you want the camera to move in the scene. And because we work with filmmakers from the most of the history of Runway, we immediately got this feedback and got this, decided that this was worth investing in. And so control ability became a big theme, I think, very early on as we were building, as we were building those models. Something fun that I haven't really talked about too much was just how Gen-2 came to be out of Gen-1. So it was a bit strange because we announced Gen-2 two months after Gen-1 andHow Gen-2 Came From a Weekend HackVibhu [00:20:43]: We're accelerating.Anastasis [00:20:44]: It was before Gen-1 was even generally available. But Gen-1 was a depth-to-video model, so it would take a depth map and it would convert it into RGB. and we couldn't get, text or image-to-video to work directly, and that's why we started from depth to video. but, and we had discussions of like, okay, we need to spend the next six months investing in text-to-video, maybe increasing the compute scale or the model scale, like train a larger model. And I had this weekend project idea, which was, what if I take a model that, starts from text input and converts to depth maps and then use Gen-1 to convert the depth maps Into RGB?Vibhu [00:21:29]: It would probably work.Anastasis [00:21:30]: And so Gen-2 was that.Vibhu [00:21:32]: Oh. The hackathon pipeline.Swyx [00:21:35]: The weekend hackathon pipeline.Anastasis [00:21:36]: Yeah.Vibhu [00:21:37]: But it looks good.Anastasis [00:21:38]: And it worked pretty well. there were if you, with the knowledge that it has this, like, two-stage pipeline, you can tell in some cases that the structure of the video looks a bit off because you had to generate the depth first before you go into the output video. But it worked and it allowed us to bring this to our, to users very quickly. But it's now it's interesting because, like, people are coming back to this almost two-stage approach. Like, if you look at the Reve text-to-image model that came a few months ago, it had this planner model that would generate bounding boxes before it fed that into the diffusion transformer.Swyx [00:22:19]: Yeah, Ideogram also the same day.Anastasis [00:22:22]: Yeah.Swyx [00:22:22]: I remember that was very strange that both of them came out the same day with the same exact innovation.Anastasis [00:22:26]: It's a small community, I think.Swyx [00:22:28]: I'm like, this is like, this is completely coincidental, right?Anastasis [00:22:32]: People talk. So yeah, there's, there's definitely something into this approach. And, now, like every single like, video generation model in production uses a complex prompt completion pipeline under the hood. I think that's no secret that there is. ThatSwyx [00:22:48]: Humans are terrible at prompting.Prompt Rewriting, Camera Control, and the Seed of World ModelsVibhu [00:22:51]: I think across the board.Anastasis [00:22:51]: Yes.Vibhu [00:22:52]: But yeah, I think like the original Sora one blog post even told you that what happens after your input is rewriting your prompt. It's much more descriptive about what you would want.Anastasis [00:23:02]: Exactly. I, And there was the DALL-E 3 paper beforehand that, was the first public, description of the fact that synthetic captions and really detailed captions work really well. And then Sora built on that. Yeah, so it was 2023. We were releasing all these updates to Gen-2, like the camera control, Motion Brush. And there was something very interesting about camera control because it was the first time that you felt that instead of, like, you were creating video, you were creating a short video, you were navigating inside the world. And I think camera control was maybe the seed of some of the ideas that we had around world models and really opening up that research direction. We realized, it was this era and this series of, Gen-1 and Gen-2 models really proved to ourselves, yeah, this is theSwyx [00:23:56]: Cool.Anastasis [00:23:57]: So this is not the original camera control. This was the updated camera control on top of Gen-3. But yeah, I think it made those models usable to filmmakers, I would say. The so camera control was very popular. And so we realized, there is one way of seeing those models, which is, you're just as content creation machines, and there is the other way, which is you're. As you're predicting video in order to predict video well, you need to simulate the world in an increasing and increasing capacity. And if scaling laws apply on video, just like they apply on language models, then as we scale the compute that we put into those models, then they're gonna be able to simulate physics, they're gonna be able to simulate human actions and dynamics increasingly well and predictably well. That was the thesis about around our efforts on world models, and we spin up this research group to just focus on the world models and how do we turn the video generation models that we're building into something broader and something that would be useful beyond, also content creation as well.Swyx [00:25:04]: And that was roughly when?Anastasis [00:25:06]: Yeah, so that was inSwyx [00:25:06]: OhAnastasis [00:25:07]: In late 2023.Vibhu [00:25:08]: Interesting. like, I think, a lot of people have been saying a lot of video gen model companies have all pivoted to world models these days, but like, 2023, you're posting it. oneWorld Models: From Video Generation to SimulationSwyx [00:25:21]: It's, it's debatable whether it's a pivot.Vibhu [00:25:23]: Yeah.Swyx [00:25:23]: Like, arguablyVibhu [00:25:24]: YeahSwyx [00:25:24]: That's what you always had to do anyway, right?Anastasis [00:25:26]: It's in a way an expansionVibhu [00:25:28]: YeahAnastasis [00:25:28]: Of the applicationsVibhu [00:25:29]: YeahAnastasis [00:25:29]: Of the models as they become more capable.Vibhu [00:25:31]: The early signs, it seems like the original models you guy had, guys had, people would say it's very not bitter lesson pilled, right? You're adding, rewriting prompts, you're having all these one-off things, but that's just the state of the tech as it was versus the future of as you said, you can scale it up as, we can scale up to world models.Anastasis [00:25:50]: Yeah. So it just became. And if you looked at the outputs of Gen-2Vibhu [00:25:56]: YeahAnastasis [00:25:56]: It was not. I think it was not obvious to people that this would scale to become a general simulator of the world. Like, you had very limited movement, you had, very low fidelity or low resolution, like obvious mistakes in human anatomy, like all kinds of limitations. But it was just, the idea was that's just GPT-two, and GPT-two, it can barely generate, like, coherent sentences. Similar, Gen-2 can barely create coherent video, but if you scale it up, you're gonna. There is no reason why it shouldn't work in a way. It's, And I think that was. That's, that's always the mindset of Runway is like this extrapolation of, like, if, like, even when we started in 2018 and you looked at the results of the day, you need to look more at the trend of, like, where we were in 2018 versus when we were at the, when the first GAN came out in twenty, four 2014 or twenty, fifteen. And, you started from, like, thirty-two by thirty-two images of faces, and then by the time in 2018, you could generate, street images at the 1K resolution. And it was the same with world models, very early signs of something much bigger.Swyx [00:27:08]: Yeah. I was gonna say, like, it's diffusing into focus. Like, if you look at our visible output from year to year, it looks like a diffusion process itself.Anastasis [00:27:17]: Yeah.Vibhu [00:27:17]: Especially watching the early, like, old blog posts, you can really see the choppiness, the details.Anastasis [00:27:24]: Yeah. Like human civilization starting from random noise and thenVibhu [00:27:27]: YeahAnastasis [00:27:27]: Denoising intoSwyx [00:27:28]: Yeah. Just run it a hundred years.Anastasis [00:27:30]: Civilization.Swyx [00:27:30]: Yeah.Vibhu [00:27:31]: That's how you're on track, you're still noising, right?Swyx [00:27:34]: Yeah. I like the way that you guys phrased it when you, announced it in June, which is, oh, that you had a video essay. “The human mind is no longer the center of AI. Our world is.” Right? Which is, let's, let's call it the past five years of LLM-based AI is very much like trying to emulate human preferences and human speech. But now that's, like, mostly solved. I think that's, like, some of the context of your essay, which you also wrote around the time. And now it's like the focus is on modeling the world accurately.Scaling Laws for Video and Why Predicting Pixels MattersAnastasis [00:28:03]: Exactly, yeah. So the way we see it is, there is that, initial mission statement of DeepMind, which is, solve intelligence and then use it to solve everything else. But I think it's starting from everything else, could be valuable of, like, starting from. there is just so much complexity, and detail in the world that in order to. That it's, it's hard to learn directly from just human descriptions of the world. Like, we're assuming that, like, language models learn from everything that humans have written about the world, like our own understanding as of, the twenty twenties. And there is just so much that we don't know and so much that's not captured by existing text, about both the low level dynamics of the world, like we're not describing in detail. if I tell you to describe, like, how do you tie your shoes, that's a very difficult thing to describe in words, but it's very obvious thing to demonstrate. And so I think there's been. And there's, more of X paradox, like we're constantly underestimating all the complexity that goes into very, like, things that we do subconsciously as humans, and we don't even necessarily always have the words to describe them. And so in my mind, the simulating the world and simulating, physics, simulating the dynamics of the world has always been underestimated, compared to, we place too much emphasis on the things that are easy to talk about. but there is just all this complexity and richness of the world that if we just try and train directly on that observational data instead of training on how people describe the world, we would learn something new that we wouldn't otherwise know.Swyx [00:29:54]: You think that the present architectural paradigm is fine? You don't need, like, another layer, like JEPA, like another famous, New York AI leader would say?Anastasis [00:30:05]: We're a very pragmatic research lab. If, we have evidence that an approach works better than the approach that we're taking, then we have no qualms to taking it. We just have seen no indication that video prediction itself doesn't scale. And even if you look now, not just our work, but the work of others, you're seeing in robotics some of the most promising work, starts from video prediction models, and then you adapt them to also the action models, for example. so there is very little evidence that you need something else and that your time is better spent on a novel architectural change compared to improving data and improving the, and scaling the current approach. And so, We don't have any indication that. the, there is that counterargument that I think there was a tweet by Yann LeCun a few days ago that, understanding the dynamics of the world is very different than, generating, cute videos.Swyx [00:31:05]: And your answer is no, they're the same thing.Anastasis [00:31:07]: Yeah, they're the same thing.Swyx [00:31:08]: My cat videos are the same as understanding physics.Anastasis [00:31:11]: Right, because if you wanna generate. video models can cheat and, like, they could you could give, like, successive dif shots of the scene in a way that doesn't require you to simulate difficult physics. There is like, all these different ways in which you can hide the deficiencies of the model, and it's important not to be too tricked by the performance of the current video models. It's easy to, cherry-pick examples and think that video models are further advanced than they are. So there is a lot more work that we need to do to improve those models. But in my mind, very similar to language, and, like, we've. you go from barely coherent sentences to something that, could hold a conversation with a human to something that could can operate autonomously for a day and, like, create entire code bases. And the main difference, there is some architecture improvements along the way, but the main thing is scale. And so it's the same bet for video, and we have no indications that this is saturating. Like, we have benchmarks that we use for measuring the physics of those models, and we see those predictably improve as we scale those models. So there is. If you want to Google up, Physics-IQ, is one of those benchmarks that measures how well does the model perform at solid mechanics or fluid dynamics or optics.Vibhu [00:32:32]: I'm curious if you've seen any emergence, any scaling law around this.Swyx [00:32:37]: Yeah, he's saying there is a scaling law, right?Anastasis [00:32:39]: Exactly.Vibhu [00:32:40]: Yeah,Anastasis [00:32:40]: So the way those models, those benchmarks work is you. the researchers have gone and, like, captured, a few videos that are representative of different physical phenomena, and then you can take the first frame and then pass it through an image-to-video model and then generate a rollout that shows what should happen next. So you have, a ball hanging from the ceiling, and then you use that as input, and then you the model predicts how the ball should fall on the ground. and this measures. we have an intuitive understanding of physics. I know, you can imagine what will happen next if I drop this bottle. So it's measuring that same intuitive physics understanding of those models, and we've measured that at different model scales, and we see, and compute scales, and we see that the score on physics IQ predictably improves. There's other, tricks and techniques that you can make to improve the score even further, but even scale alone helps, in the model learning better physics.Swyx [00:33:40]: My main sympathy with Yann LeCun is the, Plato's cave allegory, right? Like, you're, you're, like, learning on the output of a thing, not the internal process of a thing, and it's very noisy. And, if only you could observe the internals of a thing. It's hard to observe the internals of a human mind, but you can very much observe, or at least we have a whole branch of science and physics that we're ignoring on how to model Physics and movement and, gravity and, other interactions. and we're just, like, throwing away all of that and just saying just scale data, which is very much the lesson of unsupervised learning, but it feels wrong. that's the main idea.Anastasis [00:34:21]: I think the history of machine learning is, at large, it feels wrong.Swyx [00:34:25]: Yeah. It's a bitter lesson, right? Yeah. It's, it's, it's the simple answer to that.Vibhu [00:34:29]: I guess, how much can you scale? So, like, even on, let's say, the video generation side, like, there's one side of video understanding. Video generation, are we still gonna have tools where it's like, I wanna generate two hours, twenty hours? there's a infra way to do it in batches and stitch it together, but, like, do we just keep scaling? Do we just continue long generation consistency, all that at scale? And, like, tying it into where we're at now from we looked at Runway two to four point fiveGen-3, Sora, and Runway's Scaling InflectionAnastasis [00:34:58]: Yeah.Vibhu [00:34:58]: Like, technically, what advancements have we made to today, and then where do you see things still going?Anastasis [00:35:04]: So part of the answer is definitely scale. and that was. We learned that lesson in a big way for with Gen-3. So Gen-3 was the model we released the year after, like in 2024. That was a few months after Sora was released. so yeah, there's an interesting story of that came to be as well. Gen-3 for us was, the first time that we really needed to build. we had to learn all the lessons that the language model world learned in two in three years in the span of a few months. one of the biggest changes of Sora was using diffusion transformers instead of convnets. So a lot of the early, latent diffusion models were all, convnets for the diffusion model part. And the diffusion transformer paper came at some point in 2023, and it showed scaling laws for image, diffusion transformers. And we realized at that point that we needed to invest in infrastructure for model parallelism, for really scaling training to larger than, a few billion parameter models. And we spent maybe the, most of the fall of 2023 building out our infrastructure for distributed training. And we had a lot of false starts and a lot of failure in trying to scale, image and video diffusion transformers. And at that point, February 2024, Sora comes out, and the results areAnastasis [00:36:35]: Very much superior to what Gen-2 could produce. There were a lot of, a lot of chatter on Twitter about Runway. Runway's done. like, there is no way Runway will catch up. And if you remember, also OpenAI in the early twenty-It felt very, like it's aSwyx [00:36:56]: To the moonAnastasis [00:36:57]: It's a formidable opponent now, but at that point, it, they were on the top of their game. nobody could even get close to them. There was maybe Gemini was just the first version of Gemini had just released. So when OpenAI came with Sora and it was such a big jump of like quality, it gave me, there was like an existential crisis for a few hours. But that, I think the amazing thing about Runway and like I think the, we've been around eight years now, which is almost we're dinosaur in AI, and we had to like, we had there was a lot of those moments we had to learn, adapt very quickly and build out skill set in the team that we didn't have. And so, if you ask anyone what is their favorite time at Runway that was there during that time, it was that push in like three months to get to a model better than Sora. and it, we scaled 10x the model scale, the model size and the, compute that we were training on. we figured out model parallelism. We had zero expertise in that. And then we came out with Gen-3 during that summer. So that was a big turning point, I think, for the company where the research org grew very quickly, and we really started pursuing this vision of the general world model, in earnest, I think after Gen-3 was out.Swyx [00:38:12]: Yeah. that's the amazing thing about building when you're building. There's no stack to. You have to invent everything yourself. You have to be completely full stack. Now I think like there are inference specialists like Fal or whatever that can help with like, model serving, and I think you guys work with them as well. but yeah, like it's, it. But at the time, it was just. It's very interesting to think about what you do when Sora comes out and people are questioning whether your company should still exist.Distillation, Turbo Models, and Real-Time VideoAnastasis [00:38:41]: Yeah. And yeah, there was no, there was no VLM of diffusion models. Like, we had to build the whole model serving infrastructure and make things efficient. And a few months after we released Gen-3, we released the Turbo version, which I think was the first step-distilled model in production.Swyx [00:38:56]: That was a whole trend that we covered as well. Yeah.Anastasis [00:38:59]: So that allowed us, to serve those models at the larger scale, ‘cause I think the first version of Gen-3 was quite, expensive to serve.Swyx [00:39:09]: I think the whole like trend in like consistency models, Lightning and, Turbo and all these things somehow didn't really stick around. I don't know if you have any reflections on this. Because at the time, I was like, “Well, everything should start with a distilled model first, and then you can upscale,” right? It. your bigger models just turn into fancy upscalers, but like you should always draft with a smaller model and faster model, right? Because you can get it so quickly, like near real-time.Anastasis [00:39:39]: Yeah. I would not be so sure to say that didn't stick around. I think that, it's, it's likely to. that there is a lot of step-distilled models that are actively used in production. there is still a gap in quality compared to the, non-distilled model. but in my mind, we're still. there is a two to three year offset from language models. So the things that, So it's just a matter of time before there is better distillation techniques. we use. Right now we have a real-time model core character that I think is the largest deployment of real-time video models, that's a step-distilled model, and it's actively being used. It's a very specific use case compared to a general video model. So this is aSwyx [00:40:27]: Very cool, by the way.Anastasis [00:40:27]: This is avatars stuff, right?Swyx [00:40:28]: Consistency, character.Anastasis [00:40:30]: Yeah. So this is a talking avatar, model. we were able to. we optimized the hell out of it, and it generates at 24 FPS, and it's a, it's a step-distilled autoregressive video model. So if we look at our world model direction, a big component of it is starting from the bidirectional diffusion that generates entire video at once and making autoregressive shows. So you generate one frame or a few frames at a time. so there's a lot that goes into that pipeline of getting to a real-time model. It's first you need to make it into a causal autoregressive model, and then you just turn it into. You need to do some additional step distillation to get it to be real-time. and I think that part is just starting. I'll be very surprised if we're, two years from now, we don't primarily use real-time models. To me, real-time video generation is just inevitable that, it has much better user experience, it's much cheaper to serve, and, the quality gap between the base model and the real-time model is only gonna close as we figure out better, distillation techniques. And we made a lot of progress there internally on maintaining the quality of the base model when we distill them.Swyx [00:41:49]: How much of this is transferable? So is it the same base model? Like if you're doing diffusion across the whole sequence and you're converting it to step autoregressive distillation, is this like distillation where you still need to train both, you can use the same base and converter? What's that process like to go from regular model to something that's real-time on a technical level?Anastasis [00:42:11]: So the nice thing about diffusion models is you have, two axes of distillation. So there is the. You can distill to a smaller model, which resembles what you do in LLMs, or you can distill in terms of taking less steps, less diffusion steps. So you could take a model that generates in fifty steps and generate in four steps and get to, You have some performance, degradation, but very often you get comparable outputs. So you can even take the large frontier model and distill it with step distillation and get to a real-time performance, and that's what we've seen. So, depending on the use case, in some cases we might also serve with a smaller model, but in a lot of use cases, we just use theSwyx [00:42:56]: Step distillationAnastasis [00:42:56]: The frontier model, and we're able to make it work in real-time.Swyx [00:42:59]: I think this might be a good time to cut over to his laptop to show off some of the real-time stuff that you're doing.Interface World Models and Neural SoftwareAnastasis [00:43:06]: This is one of the research updates that we did recently. so we've been working and f in getting our general world models to, different applications. one of them that we think is very compelling is using general world models as essentially, an interface, a universal interface to software. This is a version of our world model that's called an interface world model. and the idea is that it essentially, replaces, the, front end of a software application. It renders the pixels directly of an interface and is trained to predict what happens next as a result of, a click or another interaction you have with the interface. So this is all pixels. it's there is no HTML, CSS, React that's powering this interface. This is directly at the output of our real-time, video generation model, and it takes clicks directly as input.Swyx [00:44:09]: And drags, click and drag.Anastasis [00:44:12]: Right. So it supportsSwyx [00:44:13]: Ooh.Anastasis [00:44:14]: Yeah, clicks. It supports drags. it also supports scrolling. and the amazing thing about this is that you can effectively describe in the prompt how you want different elements, like what do you want the behavior of different elements to be. So it's almost you're you can turn, an interface from, markup language description of, like, an HTML interface, and instead you can just describe the interface. if I press this button, I expect this to happen. If I press this button, this should happen. And it's useful, we believe, both for prototyping, for, like, just testing, like, what different interactions would feel like. you can also add audio to it. So it's a video audio generation model. So you get you essentially can describe both what the visual outcome should be of your click and also what the if there is a sound effect that comes out of it. So we believe that's gonna be a much more flexible way of building software. Just render. It just, in why generate the code that generates the pixels? Just generate the pixels directly.Anastasis [00:45:18]: It's the end-to-end philosophy applying applied to front ends.Anastasis [00:45:25]: So we think there is a few interesting use case. So you can build creative tools on top of it.Anastasis [00:45:32]: We think that, for any use case that involves a lot of exploration or, like, educational use case where you wanna learn about a new concept and you want some visualization and like, and open-ended exploration, we think those this is a very powerful, approach. you can imagine new forms of, design, industrial design software that could emerge as a result of those models. And this is all, generated in real-time as well. So, you can build a lot of interesting camera transitions and forms of interaction that are very difficult to build otherwise. And one way in which we evaluate this is what if you try to generate the same interface with Claude by just, prompting Claude, “Here's an image reference of my interface that I made in Figma or that I created somewhere else. create this particular interaction,” which in this case it's, drag that object, upwards. and beyond it being slower, it's also very difficult to capture some interactions by just fully, with just LLMs. So we think that this is likely to be the way that a lot of the future, like, software in the future will be created. and one of the additional benefits is personalization might be a lot easier done with those models. Like, you can essentially try out different prompts based on who is visiting the interface. You can, more easily, prompt engineer the interface to have larger size, text for more accessibility reasons, or you can make this or, like, if you have a particular aesthetic preferences. So we're very excited about this approach. It's early days, and I think we'll need to, make it more cost-effective as well to serve those models ‘cause, running a real-time video model versus just purely rendering HTML, there's -- the computational needs are much higher. but we do see a lot of potential in this approach to building front-end interfaces.Swyx [00:47:47]: So we covered this similar thing with Flipbook before with our, Ethan Hara episode with Groq, video. And yeah, I think it's very engaging visually. I think it's maybe very good for education, but it's it does sound expensive. I think there's an upper bound to how expensive it will be, though, right? Like, the inference cost will go down over time. You'll figure out ways to optimize it. Effectively, when it pauses, you don't you're not receiving human input. You don't have to generate anything, right? So.Anastasis [00:48:14]: Yeah, you could also. Like, in this case, you have ambient motion, so there is parts of the screen that might. if you're let's say you wanna, visit Paris and then you get this interface that allows you to explore.Swyx [00:48:29]: People walking. Yeah.Anastasis [00:48:29]: You have people walking or, like, things happening. But, it's, it's a no Yeah, it makes it more expensive because you need to run the model all the time. Maybe you have some looping mechanism so you don't need to do that. But all those things, I think, is stuff we'll need to figure out.Toward a Fully Neural Operating SystemSwyx [00:48:44]: Yeah.Anastasis [00:48:44]: I think our first consideration is let's make this clearly find some use cases where it's clearly a much more compelling interaction compared to traditional interfaces. And then it's a matter of time before it becomes more cost-effective to serve.Swyx [00:48:58]: Yeah. When it comes to the people walking, I think the approach that makes the most sense to me is Nick.Anastasis [00:49:04]: Nick.Swyx [00:49:04]: Oh, God. I keep messing up their name. With Chris Manning and Fanny Yan. I don't know if you've come across them, where they. Mapped to some game engine. I think it's Unity or something, or Godot. And they you can script some NPC behavior behind that and train on that. Whereas here, you can really imagine whatever you want. Like, that is a UI, right? Like, and it feels, like, more tractable, I guess, to, create a world model of software that is interactable because we have many of examples of that, and you can, do your fancy RL environment stuff on that than it is scaling up to embodied and real-world physical use cases. But this is a nice first step.Vibhu [00:49:43]: Or, there's the opposite of you have, like, one B models, three 50 million parameter language models. It just gets so small that they're just predicting, like, fishes moving.Swyx [00:49:53]: Small models are now 120 B, so.Vibhu [00:49:57]: Ultra mini on device.Vibhu [00:49:58]: But, no, I think it, like, it puts it into perspective, at least the car one for me, like, the applications, right? The amount of work to do that, sure, you only make one model year car per year, but applying this, it's also a cost-saving to have to manually make all this, right? So it opens up a lot of possibilities, too. I'm curious if you extend this out two, three years, so where do you see things going even further?Anastasis [00:50:25]: Effectively, the end game of something like interface world models is you have, a fully neural operating system. So I think, Andrej Karpathy has written about that quite a while back. But it's, You, I think to me it's, it's a bit, it's a bit odd that, we have, for example, with an interaction with an LLM of today, you have this LLM that can talk to you about anything. It can You can take the conversation in any direction. You can It's very general, so it can solve all those different tasks, but you interact with it through a very rigid interface. And so to me, it's just a matter of time before the interface itself becomes learnable and becomes, part of the whole loop of, like, you're not just delivering. You're delivering an application end-to-end, and that means you're delivering the language model, but you're also delivering the render and the pixels and that's also a learnable component. And the concept of applications might not necessarily. I think we'll need to figure out new abstractions for software. the concept of application comes from this idea that you need, separate code bases to describe, to, for, to power each individual, tool and each individual application. But you might think of something a lot more unified if you're. if you have, a video model that's generating the interface as you go. so it can take context from an LLM and allow you to combine different functionalities that traditionally would live in different applications. So it's a, it's a way to solve, software end-to-end, effectively. We also see this as a powerful way to train computer use agents as well. so this is, one way to see this as. And in general, with world models, there is those two directions. One is world models for humans and world models forSwyx [00:52:24]: AgentsAnastasis [00:52:24]: To train agents.Swyx [00:52:25]: Yeah.Anastasis [00:52:25]: And so for every new work of, world models that we do, we have this both uses become possible. So this is a powerful synthetic data generator for training computer use models. It could become, a live, RL environment that you could use to do online RL with a computer use agent, and you can get wide diversity of different interactions, kinds of interfaces, just generated on the fly that, to improve the how robust the, your agent, becomes. So that's the same also with the world models that we're working on for a robotics use case as well.Long Context, Error Accumulation, and Autoregressive VideoSwyx [00:53:02]: Is there a research breakthrough that you're Waiting for that would unlock the next set of use cases that you really wanna pursue?Anastasis [00:53:10]: Long context is a very important one, so being able to maintain consistency for long periods of time, and that depends on the use case. So for our characters model, for example, or for the interface world model, it's easier to maintain long sessions of interaction. If you go into more open-ended worlds that you navigate and you take arbitrary actions in, we, like, there is more the context at which you can and duration which you can generate becomes limited much more quickly.Swyx [00:53:40]: Yeah.Anastasis [00:53:40]: So we see more degradation and error accumulation happening. so the biggest challenge with autoregressive models is error accumulation, is you're feeding generative frames back into the model to generate the next The next frames. And if there is any small errors, they accumulate over time. That's not a new problem. It's a problem that LLMs also have, and we've seen the ability to generate now really long outputs. So it's a solved problem, but it's definitely still a challenge.Swyx [00:54:08]: Yeah. And what is the state of the art? so for Grok, it would be like 10 to 20 seconds of context going in there for video.Anastasis [00:54:16]: With our characters models, we're able to generate up to 30 minutes of video autoregressively.Swyx [00:54:21]: Yeah. But that's just for the avatars.Anastasis [00:54:24]: Yeah. So if we look at, GWM Worlds, which is more our open-ended world exploration model, it's, it's on the order of a few minutes, which is Yeah, soSwyx [00:54:35]: Probably enough for people because you have to cut to the next scene anyway, right?Anastasis [00:54:40]: Yeah, it's not, it's not the ideal game experience if you have to restart every few minutes. So I think. But, I think it's. Yeah, for certain kinds of game experiences, you can work around it.

It's a regular Thursday sync until my manager cuts our Figma license and hands the whole product team one month to go fully AI. In this one I walk through who made that call, the surprisingly solid case for it, and the very selfish panic running through my head while I said nothing.When the tool you spent a decade mastering loses its market value in a single meeting, what exactly are you afraid of losing: your job, or your identity?This is the episode about the day the call actually came down. I'm at my kitchen counter with a fresh coffee, hopping on a normal weekly UX sync, when my manager announces the entire product team is moving to a 100% AI workflow by the end of the month. And by the way, the Figma license is getting cut. No debate, no “what does everyone think.” The decision was made in a room I wasn't in, then delivered to a room full of designers who had about a month to deal with it.I do two things in this one. First, I try to be fair to the decision, because it deserves it. I get into the cost math leadership was staring at in early 2026, the pricing question Bob Baxley raised after I posted about the switch, and why ripping the band-aid off with a brutally short deadline was probably the right move and not the reckless one it felt like. Then I tell you the less flattering half: what was actually going through my head in that meeting. Spoiler, not one of those thoughts was about our users or the work. Every single one was about me, my title, and whether being “the person who knows Figma best” was ever a real skill to begin with.I also get into the part nobody sat us down to talk about, which is that these tools blur the lines between design, engineering, and product in every direction at once, and introduce the “in the loop, on the loop, out of the loop” question that runs through the rest of the season. If your company hasn't had this meeting yet, I'd gently suggest not assuming it isn't coming. Give this one a listen, then let me know about your own version of the kitchen counter moment.Topics:• 0:00 The day they cut our Figma license• 3:00 The cost play: why pay for two tools?• 4:05 Not my call: where I sit in the org• 4:50 Nobody said "adapt or quit" (the soft mandate)• 5:30 The fair case, and Bob Baxley's pricing question• 7:15 Ripping the Band-Aid: off the dock, into the boat• 8:40 What was really in my head: "just UI engineers now?"• 9:45 Kevin the plumber• 10:50 Watching a decade of skill depreciate• 11:50 How remote teams process shock• 13:10 Does engineering even want us in the code?• 14:00 Every role blurs at once• 15:10 The walk• 15:55 In the loop, on the loop, out of the loop• 17:10 Next time: what we traded away—Thanks for listening! We hope you dug today's episode. If you liked what you heard, be sure to like and subscribe wherever you listen to podcasts! And if you really enjoyed today's episode, why don't you leave a five-star review? Or tell some friends! It will help us out a ton.If you haven't already, sign up for our email list. We won't spam you. Pinky swear.• ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Get a FREE audiobook AND support the show⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠• ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Support the show on Patreon⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠• ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Check out show transcripts⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠• ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Check out our website⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠• ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Subscribe on Apple Podcasts⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠• ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Subscribe on Spotify⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠• ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Subscribe on YouTube⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠• ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Subscribe on Stitcher⁠

Guest episode! Kiera is joined by Paul Edwards, CEO and co-founder of CEDR HR Solutions, which provides human resource support especially for dental practices. They go through a rapid fire of topics, such as coaching and accountability, worker classification, documentation and evidence (including AI!), and reporting policies. Visit CEDRSolutions.com to learn more about Paul.  Episode resources: Subscribe to The Dental A-Team podcast Schedule a Practice Assessment Leave us a review Transcript: Kiera (00:00) Hello, Dental A Team listeners. This is Kiera and today is a great day. I love my episodes with Paul. Paul is incredible. He is the owner and founder of CEDR. And guess what? He is somebody that I get to just pick his brain about all things HR every time he comes on the podcast. So, Paul, welcome to the show today. How are you?   His Paulness (00:19) very good and I'm gonna take that introduction, that part where you said Paul is incredible. I'm gonna turn it into my ringtone and I'm gonna be insufferable for the rest of the year with that.   Kiera (00:25) Yeah.   His Paulness (00:27) So thank you for that.   Kiera (00:29) You're welcome. I it's funny because every time people meet you and they know you're an HR, they're like, he's actually really funny and he's a good time. And he used to play in a band. And I'm like, yeah, he makes HR not feel as scary. But every time, every time I leave, everyone's like, my gosh, he scared me to death, which is good. Like I think it's a good healthy, like you're super approachable, super nice, but also like whip us into shape of like do this, don't do that. And I appreciate you continually like guiding us along the right path.   His Paulness (00:55) There's   a lot of moving parts when you're an employer. So it all feels a little bit scary. 'Cause soon as you get figure one thing out, then there's the next thing. You're like, no, I'm doing that wrong or or I could do it better or whatever. So yeah, you know, we try   Kiera (01:07) Yeah.   His Paulness (01:08) to break this down so it's understandable.   Kiera (01:11) Which and I appreciate. And if you guys are not subscribed to his newsletter, he's got the best newsletter out there. Like I think you guys do an incredible job. So, Paul, if they're not, how do they get connected to your guys' newsletter? 'Cause that's one of my favorite things about you guys.   His Paulness (01:21) Well, I think what you want to do is just go to the website and search for almost any form on there, but there are education forms on the website. So just go to the website and join what we kind of refer to as HR Base Camp. So if you think about base camp as a place you come back to when you're on a big hike, well, you're on a big hike as an employer. And so we are constantly sending out helpful stuff, you know, two or three times a month. We send out some pretty cool stuff to all of our base campers.   That'll get you also an email that invites you to join the Facebook group, which is kinda cool 'cause you're in there with just managers and owners. and you know, it's about ten, I don't know, nine thousand people in there, something like that. So, yeah, that's the way to go there. you know, you can also subscribe to my podcast, which is what the hell just happened and we cover a lot of topics there that that are kinda fun, you know. So   Kiera (02:12) I love it. Okay. I want to just get that right in at the beginning because you guys truly he's got the best newsletter, got an incredible podcast, and I think this is the stuff that keeps business owners up at night. So let's like not have that happen anymore. So, Paul, we're gonna just like rapid fire. There's   His Paulness (02:27) Okay.   Kiera (02:27) a few topics, and then I know you've got some case studies for us as well. So, and if you guys don't know Paul, CEDR, they are one of the top dental HR companies out there. most of our clients use them. He is one of our   I do not bring a lot of people in to speak to our private groups. And Paul is one of the people that I have hand selected to come in and speak to all of our dentists in a very intimate that's not on the podcast. So it's like no recording, it doesn't get shared. He comes in and speaks to them. So, Paul, I'm excited for today. This is stuff that we'll talk publicly, but let's talk about some of these items. Coaching and accountability. I'm sure we're all doing it wrong. I'm sure I've been coaching and telling people to do this. So tell us what we can and can't do when it comes to coaching and accountability.   His Paulness (03:08) you know, the thing about coaching accountability is that it starts to feel like it starts to feel like something that's punitive, like consequences. And and when it goes into that realm, it's really hard to get somebody to listen. I I I don't know about you, but when someone starts to tell me what I'm doing wrong and that's what the focus is, I get defensive. And that's just human nature. So when we t when we talk about coaching and accountability, especially with Dennis, and I know that you're gonna I think you're gonna agree with me.   I while there are exceptions to the rule, and probably many of them, overwhelmingly I find that dentists do not want to have difficult conversations with their employees. They kind of avoid it. That I mean it's it's about 50% of the reason they get office managers because they're like, I want to have these conversations. I can get someone else can do it. Here's the thing that   Kiera (03:55) Yeah.   His Paulness (03:56) I want to say. Yes, sometimes it ends up where you have to have a really serious conversation with someone. but the   Better way to go about this is for you or you and your manager when you hire someone is to set the expectation with the employee, I mean literally during onboarding, that they are going to receive feedback from you, both positive and sometimes it might feel a little negative, because you're making small corrections. So if you get in the practice of making small corrections as you go forward, you don't you you have far fewer of those big conversations with someone.   And then the the ones that are frustrating, I mean the issues that you're facing with an employee, also become easier because when we're looking at someone who's late all the time, and you know, in a practice with you know, we ha always have just the right amount of employees. We never have a I don't know about you, but I never have an excessive amount of employees around, where everybody's important, you just start leaving no quarter for being late all the time.   you really have to say something about look when you're even when you're only five minutes late in the morning and you tell them what the issue is, you tell them what the impact is, you explain their role in a very positive way so that you can say, Look, when you are late, we are missing you. Not you need to be here on time because when you're on time, bad things happen. And if you don't start being on time, bad   Kiera (05:19) Mm-hmm.   His Paulness (05:20) things are gonna happen to you. If you start with trying to enroll them in the idea that they're playing an important role.   you know, part of the reason I think for goal setting is that you get to say, how are we going to get to the goal if this behavior continues? So I just want to give everybody this reminder that coaching is coaching and it doesn't always have to be corrective and it needs to you if you set an expectation and then you execute on it, you're gonna get a lot better outcomes with people when you bring them on board. And and the other the other end of coaching should always be.   Can you do that? And do you need anything else from me? Like is is is the issue here me? Do we need to give you more training? What can we do to help you get there? So don't forget where that's appropriate, you'll always want to insert that in.   Kiera (06:08) I think Paul on that just like piggyback, what do you do when you've had that conversation like four times? Like, Perry, we've had this, we've had this, we've had this, like, I miss you at Morning Huddle. We need to be at what   His Paulness (06:20) I missed you.   Kiera (06:21) point, at what point is it not like cute? It's like the I feel like it's the child that keeps throwing the rock and mom's like, Okay, let's not throw the rock, like it's gonna break a window. When   His Paulness (06:28) Yep. Stop it. Stop it. Stop it. Yeah.   Kiera (06:31) do I actually like do something about it?   His Paulness (06:34) As my mom would say, when do we get to snatch a knot in him? That's what I I was just with my mom. She's eighty eight years   Kiera (06:37) Ha ha ha.   His Paulness (06:39) old right now. I she actually said, Son, don't make me snatch a knot in you. 'Cause I like to take her out in public, embarrass her now, and pick on her. but anyway,   Kiera (06:46) Mm-hmm, mm-hmm.   His Paulness (06:51) look first things first, great question. I I would do wanna say that after you've had the same repetitive convert conversation with someone   After the second or third time, especially if it's in close proximity in time, you know, you've already talked to them two weeks ago about something and they're doing it again or not doing it. you also want to make sure that you're making little notes in their file. That's very important. and what I found is is that if you don't hold someone to the standard, right? The standard that you're looking for, I I you're they're never going to improve.   And if they're not improving, then maybe they're not the right fit for your company. And the sooner that you figure that out for them and for you, the better. And so if I'm in my third or fourth time within a month about the same thing, I'm sitting down and having a much more serious conversation with them. And in my mind, I'm I'm thinking, and sometimes I might say this to them, look, I'm not telling you that you've got to do the thing, whatever it is that we're talking about. What I'm telling you is I cannot work with someone who   does or doesn't do what it is that they're that they're up to. Like, you know, I just need you to know that I, you know, I can't continue to work with you if it's going to be this way. I understand you're late. I understand that you have all these issues. I get all that. I've tried everything I can to adjust here, but I I may not be able to work with you. And you always have that standard. I need you to show up on time. There's no more there's no more getting around this. But you really need to be adding notes to their file.   dating them, you need that, and and then we can move into written corrective coaching if we need to. And you know, that's a little more formal.   Kiera (08:35) Awesome.   His Paulness (08:36) That's them acknowledging the conversation. And look, four times in a month for the same thing, they're gonna get something in writing. That's just how it's gonna happen. If if I'm if I'm on your   Kiera (08:44) Okay.   His Paulness (08:45) HR team, that's what's gonna happen.   Kiera (08:47) Okay, great. That's what I feel like so many people just like it needs to go on forever and ever. And I'm like, there's gotta be an end date to it. So,   all right, that was a quick teaser on coaching and accountability. Okay,   His Paulness (08:53) Yeah. Yeah. Yeah, yeah.   Kiera (08:58) next one worker classification. I know that's a a hot topic with dentists. They love they love this 1099. I don't want to pay taxes on them, Paul. It's a better deal for me. They're   His Paulness (09:07) Yeah. It   Kiera (09:09) a dentist. I don't need to have them as an associate on payroll.   His Paulness (09:12) Right.   Kiera (09:13) This person's just coming in to help out. I can ten ninety-nine them.   His Paulness (09:14) They're only going to be here for a day. They want to be paid   as a as a independent contractor. so there's a couple levels of classification. And by the way, this is one of the good reasons to go sign up for our education because we're gonna do we're gonna release a really good classification guide. And you know what? you know, we're probably gonna make it exclusively available to your folks as well, just so you know. So at some point we'll pass it over to you.   look, classification, couple things. I'm not gonna put you to sleep   Kiera (09:42) Thank you.   His Paulness (09:45) on it because you do you actually need to kind of go through a checklist and do some stuff on your own. And we'll provide all of that. you gotta think of classification as a couple of things. First of all, there's that independent contractor conversation that we that we're always having. And for those of you who haven't heard me, I'm just gonna say this in a very definitive way. There is no hygienist on the face of the earth who qualifies either under the Department of Labor or under the IRS rules for employees.   as a as an independent contractor. It does not exist. It   Kiera (10:14) Nope.   His Paulness (10:14) can exist. It just can't. An independent contractor cannot open his or her own practice. They have to have a dentist attached. And that's kind of the nail in the coffin, but there's a lot more. All right. When it comes to   Kiera (10:25) Yeah.   His Paulness (10:27) when it comes to associate doctors, in some states that same thing is true. I'm serious. They the state has such strict rules   That for you to classify a general dentist and a general dental practice as a independent contractor is just wrong. It doesn't meet the Department of Labor's standards in that state. But for the most part across the country, a a a employee who's doing the main function of your business, who's performing the main function of your business, a a worker, I should have said a worker, they're an employee, not an independent contractor. So that's one part of the   Kiera (11:02) Mm-hmm.   His Paulness (11:03) classification.   The other part of the classification com conversation has to do with whether or not I can pay an employee a fixed salary and not worry about any overtime they work, or whether or not I have to track   Kiera (11:14) Mm-hmm.   His Paulness (11:14) their hours and pay them. And I'm just gonna say this. If your entire team is on salary, that's okay, but you still have to track their hours and you have to pay them overtime. You can have a manager who who you don't have to track their hours and you can pay them a fixed salary.   But not all managers qualify for that exemption. It's called an exemption to the standard to pay overtime. Not all managers qualify for that. The smaller your team is, the more in danger you are of not being able to classify someone as a salaried employee. So now that you're really confused, just   Kiera (11:51) Mm-hmm.   His Paulness (11:51) go join. You're gonna get this is the kind of information and the guides and stuff that we send out.   So the the you know, the classification's a big deal 'cause we run into it every day. I mean,   wh   Kiera (12:02) Yeah.   His Paulness (12:02) when I say run into it, we get about twelve thousand questions out of our three thousand members and every week there are four or five what we call submissions asking us about you know, they've read one of the one of our articles and they're like, I think I got this wrong. And most of the time they've gotten it wrong. and you can't rely well,   Kiera (12:21) Yeah.   His Paulness (12:22) you you it you gotta be careful because   Kiera (12:24) It's true.   His Paulness (12:25) The IRS rules for your independent contractor and the Department of Labor rules don't match up all the time. And so you have to be   Kiera (12:31) Correct.   His Paulness (12:31) very careful in that area.   Kiera (12:34) Yeah, and Paul, I'm so glad you talked about it because like it is a while it is a full checklist. And I think sometimes we like to justify it. Sometimes we like to say like, but no, like they really do. And that's not gonna go over well. And so to I'm like, cross our T's, dot our I's, which is why Paul, we do podcasts like this and why everybody should have that checklist. Because when I learned about this, I'm like, No, they're a salaried employee. It's like you gotta you gotta check off this checklist and this checklist, and they're not like you said, they're not the exact same checklist at all.   His Paulness (13:02) It it's a little complicated,   Kiera (13:02) And you had to make sure both qualify for them to truly be   His Paulness (13:04) but once you figure it out, you'll have all your positions and you'll know and and you'll be set. It's just that you guys and gals tend to borrow what someone else is doing.   Kiera (13:13) Mm-hmm.   His Paulness (13:14) Maybe it was already in the practice that you purchased, or it was in the practice that you were an associate of, so you're c you know you're copying, which is what we all do when we start a business, or you're talking to your you're talking to, you know, other other people in your industry and they're giving you HR guidance that   Is not sound sometimes.   Kiera (13:33) Yeah. No, I love that. And definitely sign up for that checklist. That's a brilliant one. Okay, next one. Documentation and evidence. Let's talk   about that.   His Paulness (13:42) we are going to be we are going to be actually doing some podcast on this where the internet does HR is what we're gonna call it.   Kiera (13:56) Mm.   His Paulness (13:56) I just want you guys to know that online HR advice is not something that you really want to participate in. well look let me let me   Kiera (14:08) we're talking like AI, internet searches. Let's just talk about it.   His Paulness (14:11) We're talking about AI,   we're talking about going on Reddit, we're talking about we're talking about going into any kind of group thing for for advice and running with that. But particularly AI is a con a concern and we'll get into this a little bit more. I think every time you see me I talk about AI. Well we now have court we   Kiera (14:31) You do,   His Paulness (14:32) have court cases rocking it right now and and companies are getting themselves in trouble for creating employee handbooks through AI.   they're getting themselves   Kiera (14:39) Mm-hmm.   His Paulness (14:40) in trouble for using AI to fire people. they're finding out that your chat is discoverable. and so in one case we watched the chat was really bad. The the thought process of the person who was putting the problem into the chat became exposed in court and they were basically asking it how can I break the law? I might I'm paraphrasing, but they were basically like I they were kind of telegraphing, I know what I want to do here is wrong.   HR wise, tell me how I can get away with it. And so just be   Kiera (15:11) no.   His Paulness (15:12) careful about online advice and taking something from someone without knowing who those people are. it's you know, the context here is you're making an HR decision that is has a law that impacts it. And if you get it wrong, it can be very, very expensive. And there's no reason for it. There's no reason to go out that way and try and self-help.   yourself. So just be very careful. you know, same thing. Same thing about the internet. I'll make this other point, maybe a little bit about AI. You know how if you're a subject matter expert and you see AI give an answer to the thing that you're an expert on, you will look   Kiera (15:55) Mm-hmm.   His Paulness (15:56) at it and you'll go, you what, that's pretty darn good and it's getting better. But here's these three or four little subtle places where it's not exactly wrong, but it's not correct.   And without context, I I would be worried about just telling someone that. Versus you're asking it a question about something that you really don't have a whole lot of subject matter expertise in it. And it gives you back that wonderful, confident, same kind of answer. It's got bullet list. it's actually, if it's your AI, it's designed to please you. So it's put a little bit of stuff in there to please you. And you'll just take that hook, line, and sinker and say, this looks so well organized   Kiera (16:30) Yeah.   His Paulness (16:31) and so authoritative.   And it look, it cited some things too, and you'll run with it, and that's what's getting people in trouble. So we just really I don't know what the word is for that. Maybe I need to make it up. but you just can't you   Kiera (16:43) Mm-hmm.   His Paulness (16:44) if you don't have subject matter expertise, you gotta be very careful when it involves employment law and your pocketbook. I think that's the   Kiera (16:52) I think that I think it's so hard because like AI we use for so many things and like you said, it is like a little therapist over there and people   His Paulness (16:56) Yeah. Mm-hmm.   Kiera (16:59) enjoy it. I mean, I'm guilty of it too. I heard you talk about this six months ago and I was like, It feels annoying, Paul. Cause I'm like, I just wanna ask like what are kind of the parameters and I I   His Paulness (17:07) And I want the answer. Yeah.   Kiera (17:09) think of you all the time and you're like, Kiera, it's gonna come back on you and like they can pull these records on you and it's like, okay, noted. I need to just go to the proper sources and channels to make sure that we're compliant truly.   His Paulness (17:21) Yeah,   yeah. I mean my AI is incredibly smart about HR and it still gives wrong answers occasionally when we're writing articles and stuff. So I mean, kinda gives you some perspective. Yeah.   Kiera (17:31) So there you have it.   Do you think there will ever come a point that AI can replace HR? Not to say I want to put you out of business. I'm just genuinely   His Paulness (17:37) No, no, no.   Kiera (17:38) curious of like, I think about for consulting too. Like, do you think that AI will like there will be HR AI in the future?   His Paulness (17:44) I think at   some point it's gonna you're gonna have some very strong models that are mostly available to lawyers. They're not available   Kiera (17:51) Mm.   His Paulness (17:52) to someone like you, you know, l someone who's not giving out guidance, because I believe that the context of the guidance alwa almost always has one or two more questions, and AI is not great at taking a left or a right hand turn. But five   Kiera (18:09) Sure.   His Paulness (18:09) years from now, six years, ten years from now.   I think, yeah, in some pla in some ways it's gonna be able to give some very good answers and prompts to get those answers. And we are as a company, we are very aware of that. We are   Kiera (18:23) Mm-hmm.   His Paulness (18:24) we are well, I I think the I've said this twenty years ago when we started CEDR. And I think this, you know, for the people listening, the business you're in, I think this might h head home a little bit. I believe that the next I believed then that the next greatest innovation   was the opposite of what was going on in the world, which was it's providing service and people and getting it getting systems right and being able to actually talk to someone instead of running them into these failed chat bots and these failed chats that you you know you put your question in, it's like, let me give you your question. And then it, you know, it it answers with one third of a cup of flour. And you're like, we were talking about my internet not working. I mean   Kiera (19:08) Mm-hmm.   His Paulness (19:09) So I think the greatest   in I think the I still think that that that innovation is there and available for us and we've been using the technology in the wrong direction. I'm using it to make us more efficient, to so that my experts can spend less time creating something from a conversation that is kind of a policy or a th or a thing, and so they can spend more time listening and and a little more time with the member.   And and worrying less about I have six other people who are waiting for me to get back to during that day. So that's my view   Kiera (19:45) Sure.   His Paulness (19:46) that's my view on AI and HR. humans are humans. As long as we're being human, I I don't think we're gonna be able to take the human out of that out of that process. Yeah.   Kiera (19:55) I agree.   I agree. Okay. Next reporting policies. Let's talk about these.   His Paulness (20:04) the the reporting policies that people okay a policy only works if it's distributed and it and it is more it becomes more powerful when the policy is distributed, it's acknowledged, and the and the employer and manager know how to use that policy. All of those things have context. So look, let's let's use AI as an example.   We used to say this about template handbooks. Some would say, Well, I got a template and I put it together and I said, Well ask the template how to and ask the qu ask the template any question. Templates can't answer questions. And so if you end up with a a policy and you don't know what you haven't had training and what its purpose is, what its limits are, what it's both both, you know, small and large, what it's how you're supposed to use it and where it's supposed to apply, then you're kind of missing   that. And so I just want to be clear that when you put together a set of policies and when you change them or you have an employee handbook, if it doesn't come with a couple hours of support and training and ongoing training, then what you have is kind of a static thing sitting there in your office and it's not really helping you when it in the ways that it can help you. and I give you guys an example.   If you don't have a good maternity policy and it's not complete, then when an employee s asks you a question about your maternity policy, you can't say, go to the handbook and tell me. You go read it. You tell me how much time that we give for for maternity leave. You tell me whether or not you can use your vacation time that you've accrued. you tell me whether it's paid or unpaid. You tell you know, all those things.   should be in the employee handbook. So it's important for owners to train their managers. It's very, very important for owners to train their managers. make sure   Kiera (22:03) I love that.   His Paulness (22:04) they understand and make sure you understand the policy.   Kiera (22:08) Yeah. And I think Paul, you're right. Like, as companies evolve, like where you started at the beginning versus where you might be today. Like we didn't have people that were getting pregnant when I first started now. It's like we do. And so definitely I'm big on like make sure that they're updated, make sure that they're current, make sure that everybody has them. That the new ones that have changed and morphed through the business, like we've been in business for 10 years. Like what we started with versus where we are today are very different. And I like that.   call out that you really do need to know that. So   His Paulness (22:40) Yeah, I I   I mean I could give you a really good example. Arizona changed about I don't know, about seven, eight years ago. We went to mandatory sick time. And all employees were entitled to it under a specific set of rules. Before then it was like it was in much of the other states, not all the states, 'cause many states have mandatory sick time now. it it was you just could give it, you could make it paid, you could make it unpaid, you could not offer it at all, you could do whatever you want to.   Once that changed, a new a complete new set of rules came around. and I just give you an example, and I think managers in Arizona need to know that this exists. In Arizona, if you fire an employee within I think it's sixty days of them having used it's sixty, yeah, within sixty days of them having used one of those sick days, it is presumed to be retaliation against them for using one of those sick days.   And so you've got to be thinking, well, h what? Wait, what? Does that mean you can't fire anybody because they can use sick days? No, the answer to w what you know, what you do about w what I just described is that you have to have documentation showing the reason why you terminated a person or took that adverse action against them. So basically we went from being in an at will state to sort of being an at will state because we've got this sick lead policy.   Kiera (24:00) Mm-hmm.   His Paulness (24:01) So I I again I was kind of a roundabout way.   but but it kinda shows you you need to understand the policy and the limitations that come the laws that affect that policy.   Kiera (24:12) How often do you recommend looking at your policies like and updating? Because those rules come into effect. Paul, you read this. Like to me, this is your algorithm. You're seeing this all the time.   His Paulness (24:22) Yeah, yeah.   Kiera (24:22) for us, it's like a one and done, check it off the list. We think we're compliant. How often do you recommend updating your policies, double checking your state laws, having your handbook updated? What are what are your recommendations   for that?   His Paulness (24:32) I think at   at least once a year and if you're in s California, Connecticut, New Jersey, Washington State, some of those states, you have to visit it more often. and I think that's the problem if you don't have a support company because how would you visit it? Like where is it that it's certainly not get f get f you know, fed to you on TikTok or Facebook or something. So, you know, at least once a near a a year you should if you have time   You should read your policy to see if you're following it. And I don't mean your policy should tell you what to do. I mean you may have changed something and it conflicts within your handbook now, and you just need to change the policy. That's it. That's that's that's all there is. so yeah, once a year you should check. Most states are now updating and putting in new laws about once a year. Many states are putting in new laws and making them effective two or three times a year.   Kiera (25:26) Awesome.   Okay. All right. Let's go into payroll. Payroll I feel like is such a thing. Payroll, payroll review. I've got some questions on payroll, but what are some of like the hot things that practices deal with in the payroll realm?   His Paulness (25:39) this has come up two or three times in the last three or four months. I don't know why. the employer is not reviewing their their payroll ledger. meaning they run payroll, you always get a report. If you're in gusto, you dig for it or ADP, maybe you can find it. I don't know. maybe it y y you know, somehow you find your you you you get this report every single time, but you don't pay attention to it.   And or you do give it a quick look and you go, yep, all their hours are right. But what you don't do is you don't review your payroll thoroughly. And what we're finding is is we've had these three instances where an employee has been paid anywhere from five to eight dollars more an hour than they were supposed to be paid. And it's been going on   Kiera (26:28) Well.   His Paulness (26:29) for years. And so tens of thousands of dollars in payment and salaries have been made.   And in all three of these instances, none of it was recoverable. In one state,   Kiera (26:43) Mm.   His Paulness (26:43) it was just not recoverable. You couldn't do it. The state would not allow you to recover it. In the other two instances, the there was no way they were gonna get this money back from this employee. They were not working for them anymore. they were gonna have to sue them and and there were problems with that. It was not gonna be a a a slam dunk home run to try to get the money back out of somebody.   So I think one of the biggest one things that we're just telling everybody right now is you need your own oversight. If you're doing your own payroll, your payroll company's not doing anything, they're not watching anything. They'll let you misclassify a hygienist. They'll let you pay somebody a salary that they're not supposed to be paid by salary. your payroll company is not reviewing your final ledger because how would they? They don't know. You they they just take what you put in. That's how they work.   And so I I just want you guys to be very careful and review your payroll every now and then. And then the other one that we've seen I've seen over the years, and it's more than once or twice, the the reason the payroll was wrong was because the person in charge of the payroll made it wrong for themselves. And then the other reason why the payroll was wrong and in in the most egregious cases was that they were so they were using it as embezzle as an embezzlement mechanism.   and they had actually put people in their family on the payroll that didn't work for the practice. and that ran for two years before they got caught stealing another way and then the full audit re revealed that they had put their family members on there. So I I   Kiera (28:22) my gosh.   His Paulness (28:23) think the the the moral of the story is check your darn payrolls. Yeah.   Kiera (28:28) Paul, that is something ten years in I still am the final stamp off on payroll and I feel silly doing it, but I like a lot of those things happen and I catch it and like as an office manager I didn't I didn't I mean I just put in the hours and as a business owner I have very different keen eyes on that. So it's the isn't   His Paulness (28:44) Yeah. Yeah you do.   Kiera (28:48) I have two other questions on payroll that I didn't prep you for. One   His Paulness (28:50) Sure, sure.   Kiera (28:52) is how can you change like let's say we've got variable comp.   That's bonuses. Like I've got an office right now, they recognize the bonus structure is very, very high. It is variable comp and they're trying to figure out how do we like change that bonus system. Do you have to have like are there any laws? I obviously know it might be state specific. This one happens to be in Arizona. but like what can you do for that variable comp commission states, like pieces like that when you recognize we're we're paying more than we're supposed to, and I need to right size that and I don't want to fire an employee, but also I gotta deal with changing their comp.   His Paulness (29:20) Yep. Yeah. I gotta deal with   changing the car. Okay, first of all, really difficult conversation. Good luck, y'all. but but let's let's talk   Kiera (29:29) Seriously.   His Paulness (29:30) about the rules around it. yeah, I mean this is let me sidetrack for a second. You may have to put me back on the question. This is why you when you promote someone to a manager's position, they don't get an immediate raise. they need to earn their way into it. Now it needs to happen quickly and it could be, you know, by steps over the next weeks and months.   But it's really hard if someone's unsuccessful and you've moved your best employee in and now you realize they're you've kind of peter principled them. They're not competent as a manager, you need to move them back down. Y the hard part would be to take the pay away from them. So so going back to your question, some states require you to provide notice in a certain amount of time before you make any pay adjustments. So you want to know if that's true or not. Arizona does not. You cannot make it retroactive in any case. so   If you're going to change someone's compensation, you need to change it and you know going forward. if it's a commission or it's a bonus program, I want you to keep those things outside of the employee handbook, but I want you to refer to those things. So I know you didn't ask me this. I want conditions in the handbook about how someone qualifies for the bonus program should a bonus program exist. For example, one of the things that's in our policy of many things, which, you know, one of the reasons you work with us.   is in order for you to earn the bonus, you must be employed for the entirety of the bonus period. Meaning that if they quit or get fired in the middle of the bonus period, they're not entitled to a portion of what's going on or what happened. So to be more direct, I I think you've gotta you really gotta work this out. I think this is a good place for AI, by the way. I don't say this very much on podcasts. I think   Kiera (31:14) Mm-hmm.   His Paulness (31:14) you I think you put what your bonus system is and what's going on and what your problem is in it.   And where you need it to go, and you help it, you you know, work with your consultant and you figure out where that pay needs to adjust. And sometimes you're adjusting the base rate of pay. and sometimes, by the way, everybody who's on a bonus or commissions needs to have a base rate of pay. or or you're, you know, you're you're moving some, you're moving something. My way of moving something is we're not hitting the goals that justify that commission or that bonus. And   In order to get back to it, here's the new numbers that we have to hit. And until   then I need to adjust this back and you you make your adjustments. are you talking about massive adjustments or yeah.   Kiera (32:03) This   one is like so it's like an office manager who came inherited to a practice and has a bonus structure in place like on collections. Well, back in the day it was great, but now payroll is at 42% of total overhead.   His Paulness (32:16) That's all hot.   Kiera (32:18) is a little high. And so recognizing the base pay is great. the additional compensation does not justify, like it's just part of their job. It's not an above and beyond. And so needing to remove that bonus structure.   until maybe we hit a higher tier of the company and it's to right size that payroll portion.   His Paulness (32:35) Yep. Yeah,   I think that I would just kill the bonus system in that in that case. And if it's across the if it's across the board, I'd kill it. Or I would adjust it back significantly and so that they still have the feeling that they can earn a little bit more on the bonus system and I would set goals for them to get up. I think in your case, I don't know what you would could get a manager to do to to right side a forty two percent cost labor   Kiera (33:02) Yeah.   His Paulness (33:03) in a dental practice.   what can I ask you what percentage you think it's supposed to be?   Kiera (33:08) So what we typically say, like a lot of consulting companies are twenty-five percent, I tend to err on thirty percent. So it's twenty-five to thirty percent tends to be where I sit. I recognize California offices sometimes are a little bit higher than that.   His Paulness (33:20) Hello, mm-hmm.   Kiera (33:22) but that tends to be about what we'd recommend for a percentage.   His Paulness (33:23) I'm I I'm with you. We   you know, we we've got several hundred close to a thousand people on payroll. So we and and and they're using our our software for their HR systems. I you're you're you're in about the the pocket right there. We see as high as thirty six percent, and as as low as twenty two percent. That twenty two percent practice is tiny and highly profit profitable.   Kiera (33:47) Mm-hmm.   His Paulness (33:48) That's a very specialized dentist with about four employees.   Kiera (33:52) Mm-hmm.   His Paulness (33:52) So yeah,   same place. I just was curious. I just, you know, love to   Kiera (33:55) Yeah, no.   It's a great one. And then one other question on payroll, because I have a lot of dentists who get burned by that third payroll, right? So it comes in the month and they're like, ugh. So some dentists, and I did this early on, so I have no idea how people can do it. I know what I did and I'm not about to share because it might not have been correct. but we move to only two payrolls in a month. So I do it on the fifth and the twentieth. And I know some dentists are thinking about that, wanting to are there any rules around how to do that? Are you allowed to do two payrolls a month? Like, because then I'm not getting hit with the cash flow hits of the third payroll.   periodically, just some thoughts of offices are wanting to do that, are the right ways to go about it.   His Paulness (34:28) Absolutely. Absolutely.   You need to check your again, you gotta check your state rules. There's fewer states that have these rules but you need to check your state rules. There's this pay frequency. By the time they finish their payroll period and they close the payroll, there's a certain number of days when they have to be given a check. In none of those   Kiera (34:47) Mm-hmm.   His Paulness (34:48) instances does it prevent you from going to two payrolls a month, th that that I can recall. It's if it is, it's it's very   It's one maybe one state, where you might have to add one or two payrolls to the year in order to meet the law. It's not every month. It's just to the year.   Kiera (35:09) Interesting.   His Paulness (35:10) Yeah.   Kiera (35:11) Okay. Yeah, 'cause I know when we did it, I just I was like, We're gonna change it and they told us to float employees if they needed it. I let them know when it was coming so they could financially prepare for it and yeah.   His Paulness (35:20) Yep. Give everybody thirty days and you will have met your states   most states' requirements too that say if you're gonna change to this model, that's what you gotta do. Yeah.   Kiera (35:30) Awesome. Okay. Well, Paul, I've like we peppered you on so many questions. I know you had a couple of case studies you wanted to bring and I'm excited to like hear. I I do enjoy a good case study that freaks me out and I I feel like it's like forensic files with you. Like bum bum bum bum. Like let's hear the case studies now.   His Paulness (35:46) I think I'm gonna start with a f positive one and then I'm I'll ruin y'all's days with the other one. here's the here's   Kiera (35:52) Ha ha   His Paulness (35:52) here's the positive one. And we have been saying this for years and we're the ones that started implementing it with our clients. It's not that we invented it, it's that not enough importance was put on it. And when we started in two thousand six, everybody was making up their own employee handbook or they were going to ADP and they were getting handbooks that didn't have anything about HIPAA in them.   And they didn't have anything about maternity leave. I mean, literally, payroll companies were telling employers, you don't have to give you're not required by law, so we don't put it in the handbook. And and we're like, so offices with seven women in them who   Kiera (36:23) Yeah.   His Paulness (36:24) are all going to get pregnant at some point in their career, you don't have a policy for what to guide them through. Anywho, the policy that we are very big on and we use it throughout the handbook in the c in the proper places.   is a requirement to report a problem. So it's very simple. If you re if you have a problem and it could be with payroll, it could be around harassment and and you got to define that. It could be around many other things. You have to report it to us. Here's how to report it. If you're not confident who manages you, here's a way to tell somebody else about it. If you put those reporting procedures in place and someone doesn't follow it, they get fired.   And they come back six months later to sue you over the problem that they never told you about, then the the judge is supposed to say, Hey, when did you tell them about this? When when when did this happen? And so we see it happen a lot in depositions, which never make a court case, so you don't see it publicly, where the lawyers got someone in deposition, they'll just say, okay, cool. so you said this was happening for the whole year, and you had four one on ones with your manager.   and when when did you report it to your manager and and and or or or an email to somebody, you know, that that kind of thing. Well, we recently saw a case come up, it was errors versus chemjet. I I don't know what in the world chemjet does. chemjet argued,   Kiera (37:54) Mm-hmm.   His Paulness (37:55) hey, we this case should not go forward. We want summary judgment, we want out, because this employee never told us about this problem. They never said a word to us. And the judge said to the and the judge granted it to him. They just said, you know,   It they're right. You had an opportunity to report this. You're back here. You have a multiple complaints, but you were you did sign the employee handbook, you did acknowledge it, they did distribute it to you, and you gave them no chance to cure this. Had you given them a c a chance to cure it, there's a reasonable expectation that they could have done something about it. Without knowing about it, nothing they could do. So again, when you self-make your handbook and you miss that opportunity,   Kiera (38:31) No, she's not.   His Paulness (38:32) you get, you know, you may not get the best result.   Kiera (38:36) I think Paul, something I've learned and when I started a business, I was like, okay, I can be cheap on certain things. I can be cheap on gauze, I can be cheap on gloves, I can be cheap on what I pay for dinners, like staff gifts. Like I can be cheap those areas. The thing not to be cheap on is have my employee handbook and my policies like cross our T's, dot our I's pay there to prevent much harder future decisions.   His Paulness (39:02) Well, you're speaking my love language right now. There are just a few places   where you just don't want to be cheap. You don't want to be cheap with your CPA. You wanna you want consultants and and good consultants cost a little bit more and so that you can get more individual attention. There's just you know, for all of us, there's just these things where it just doesn't make sense to kind of cheap out. the you know, but the entrepreneur in all of us, in   Kiera (39:26) Mm-hmm.   His Paulness (39:27) me too, and me as well. I mean, heaven help me,   AI has made it so that I can start coding my own stuff. That's probably not a good idea. But I'm doing it, Kiera, because I'm gonna I'm   Kiera (39:38) I know.   His Paulness (39:38) gonna save myself some money. that's   Kiera (39:40) Ha ha.   His Paulness (39:41) not what's gonna happen. I but anyway, that's the story, you know, that little guy in the back of your head. yeah, there's some things   Kiera (39:48) Seriously.   His Paulness (39:49) not to say. Money on, and I'll make this argument about policy. Most doctors will and will spend hours on creating their own employee handbook.   Which I mean, I don't know a doctor that doesn't value his or her time at five hundred dollars an hour. Many of you make a lot more than that. You have to generate that. That's the only way you can keep your business open. you end up with a fifteen to twenty to twenty five thousand dollar handbook that was created by somebody who doesn't understand HR law, which is you. it's just not the best way to spend your time. And I'm just I'm just like, doctor, this is   Kiera (40:20) Agreed.   His Paulness (40:20) a this is a fifteen thousand dollar document and it's got a bunch of evidence in it and it's not the right kind of evidence. So,   Yeah, thanks for letting me talk about that.   Kiera (40:31) I agree. Okay. Now let's scare everybody into why they should not make their own handbooks. Like that that'll be the one that won't let us sleep tonight. Cause I'm sure has something   His Paulness (40:37) That'll be the Halloween thing, yeah.   Kiera (40:39) that that like just don't don't shortchange on your HR and legal. To me, I'm like, those are worth it. Just pay. I can sleep at night better. Like and cross my T's not my I's. I do want to do right by my employees. So okay, what's your other case study?   His Paulness (40:49) And just like   I wanna add one more thing. When you're   Kiera (40:54) Yeah.   His Paulness (40:54) working with someone like us, you're also getting access to the experts who are taking the legal and the human side approach to it. So they've talked to thousands of practices and heard your problem probably. Every now and then somebody gets an award a an award, a monthly award for wow, we've never heard that before. but for the most part, you're talking to a a a group of experts who are there who can actually help you solve the problem too.   And that's where the real value is. You know, we were talking about AI giving the right legal answers. It's the human answers that are are more interesting. So the last case study is the it it it has to do with the classification conversation we had. a temporary employee who the doctor hired. she was a hygienist. she came into the practice, they hired her as a   ten ninety nine employee, which doesn't exist, hired as a ten ninety   Kiera (41:52) Mm.   His Paulness (41:52) nine, didn't put her on the payroll, didn't put her on workers' comp, didn't put her on anything. she ended up working for them for about three weeks. The person that she was supposed to set in for was supposed to come back, and that person was at the end of her pregnancy, and so she d was unable to return to work, so they kept that employee on for about three weeks. She was not great.   at the end of the three weeks, they she was like the the the the employee who she was replacing for for temporarily was going on maternity leave. She was expecting to stay and through that time period now. And they didn't want her to stay, and they told her, you know, no, you're you're not gonna stay. And so she went and filed unemployment. and you know, you guys are thinking, well, an independent contractor can't get unemployment. Well   Well, remember, she was misclassified, she was an employee, didn't matter what you called her. And this triggered the state to look   Kiera (42:53) Mm.   His Paulness (42:54) for her withholding and to see that she was not on the record. and then it triggered them to tell the Department of Labor in that state that this because they were working together. So UI a UI claim kicked in an entire investigation from the Department of Labor in that state. And so they wanted all the time records and   I think they may even requested bank records because they were looking for other instances of misclassification and paying people as an independent contractor. So that's the other horror story.   Kiera (43:25) Mm-hmm.   His Paulness (43:25) I don't I don't s that it doesn't usually get that bad, but it can, but I do see it blow up a lot and and I mean quite frequently people go out and make these claims. the other thing   Kiera (43:40) Mm-hmm.   His Paulness (43:40) that can happen in a misclassification in the domino effect.   Is that your associate doctor gets that that independent contractor gets audited. and in the audit they learned that they were working for you, and then that   Kiera (43:54) Yeah.   His Paulness (43:54) triggers an audit of you, and then you have to pay all the taxes that you didn't pay them, with with penalties.   Kiera (43:59) Mm-hmm.   His Paulness (44:00) That's a really happy note to end on, right?   Kiera (44:03) I know. I'm like, all right. And with that, everybody, have a great day.   His Paulness (44:05) Yeah, w with that, yep. Have a great day. I think we should pause to focus our way out of this somehow.   Kiera (44:11) I think we should too. And I think I think the piece I take about this is I appreciate honest and real because I think so many times we were like, that's not gonna happen to me. And I would like people to like, I remember it was crazy. I actually was talking to my financial advisor this week and I was like, okay, let's just like run the tape. If me and my husband got divorced and   Like what would this look like? Would we both be financially taken care of? And he was like, is divorce on the table care? And I was like, no, not at all. I just want to know like what would be my worst case scenario if that ever did happen. Cause that's going to set me up better to make smarter decisions today and like make sure we're prepared if that ever came about. And he's like, You're the first client who's ever asked me that. And I thought about it and I'm like, that is actually how I think. And I feel like it applies very much to HR is why don't we know like what worst case scenarios are? Not so we don't sleep at night, but so we make better decisions.   His Paulness (45:00) So we do   sleep at night. Yeah.   Kiera (45:00) And like this can happen. Exactly.   Cause I'm like, I'd rather know. Like they have proven that us not knowing what happens creates more anxiety than knowing. And so I'm like, I feel like this is my like when I pay for HR or when I pay for my lawyer or when I pay for I have to learn a lesson. Like we're going through some things in our company right now. And I'm like, all right, this is just my CE and HR. So like me paying   His Paulness (45:18) Yep. Exactly.   Kiera (45:20) my legal fees, me paying my HR fees. Like I didn't learn this. You paid for dental school. So this is where we learn what's the absolute worst that could happen to us.   We pay for our CE education in this to learn from people like yourself for this. And then we do the best we possibly can. And when we learn that we were doing it wrong, we correct it, we make it better now and we keep moving forward. And that's how I think I would like put a pretty bow on this because Paul, like, to me, as much as I get creeped out every time I talk to you, I appreciate it. It's not a lie. Every time   His Paulness (45:48) Okay.   Kiera (45:50) you come and meet with our dentist and I'm like, my gosh, and my whole team's sitting there. And I'm like,   His Paulness (45:54) Yeah.   Kiera (45:55) like, please like just cover your eyes and ears and don't hear that I'm messing up. Like   His Paulness (45:56) Just yeah, Toba said la la la. Yep, yep.   Kiera (45:59) All the dentists get a hear and like their teams aren't even present. My whole team sitting there taking notes rapidly. And I'm like, goodness, like great and not great. but I appreciate it because it makes me better and it makes me a better business owner. And I think that's why we're all listening to this. So that's my take. You might have a happier ending, but to me, I'm like, let's learn   His Paulness (46:16) I do.   Kiera (46:17) it. Let's play our CE fees of learning HR and let's continue to be great employees in the industry.   His Paulness (46:22) I have a very upbeat thing here   to end this on. it's very short. if your better half is like my better half, there's no way in hell they listen to your podcast 'cause they've heard your voice enough. And so your your   Kiera (46:34) It's true.   His Paulness (46:34) your spouse is never gonna hear that he's in imminent danger right now. Maybe he should hear it. Maybe we should send him a   Kiera (46:40) True. Maybe to   I did tell him I asked the financial advisor that yesterday. I was like, I did ask this question.   His Paulness (46:47) Mm-hmm.   Kiera (46:47) He's like, Kiera, I like that about you. He's like, that's more probable than me dying. He's like, how many and I was like, you're right. And I was like, but guess what? I hope we ride off to the sunset forever. Like, that's my ultimate plan, but I   His Paulness (46:58) Right.   Kiera (46:58) would like to know. So anyway, he probably will never hear this podcast, which is probably in my favor today.   His Paulness (47:03) Yeah.   look, I appreciate you having me on the show. I it's nice to be able to kinda talk through these things. This is like, you know, my passion, the thing I wake up thinking about every day. So it's kind of fun to share this stuff. I do appreciate.   Kiera (47:17) Well, I appreciate you, Paul. Paul, you're such a resource in the industry. You're somebody that I recommend to every practice that we work with. So, Paul, if people are interested, if they loved this, if they're freaked out and they're sitting there with their stomach in knots, how do they connect with you? Or if they're like just wanting to be better. You don't have to be nervous. You can still reach out.   His Paulness (47:26) Yep. yep. if they Exactly.   okay, so you can go to our website, which is C E D R, the word solutions with an S on the end. So that's CEDRSolutions.com You can go to our website, sign up for the education. if you want an employee handbook or you want to kind of figure out, you know, what it would take, you're like, hey, I think I maybe I should take care of this.   You go in the other direction,   Kiera (47:55) Mm-hmm.   His Paulness (47:55) just fill out a form, you'll talk to somebody here, it's all low pressure. You wanna make sure that you that you're saying, Look, I heard the podcast of Dental A Team, so we can so we know where you came from. and we'll take good care of you. We'll explain what's available to you and kinda, you know, go through the process. those are really the best two ways, you know. Best those are the best two ways.   Kiera (48:17) I love it, Paul. And Paul's not going to be high pressure. I will be high pressure. Get your freaking handbooks updated and like legal. Then go do that. Like that to me is like give   His Paulness (48:25) What she said.   Kiera (48:26) yourself the gift of peace of mind at night. Go get that done. So, Paul, thank you for being on the show. Thank you for being a part the Dental A Team. Thank you for being someone I trust in the industry, someone I value. And I hope today everybody listen to what he says, make yourself better, become a little bit better today. And Paul, truly thank you for being on the show.   His Paulness (48:41) My pleasure.   Kiera (48:43) And for all of you listening, thanks for listening. And I'll catch you next time.   on the Dental A Team Podcast.  

Let's talk about FM27's new UI/UX (with @mustermannfm) The FM Show is back with a four-way discussion of the Football Manager 27 UI reveal! With the first look at FM27's new interface now available, we get together to break down what we've seen, what has changed, and what it could mean for playing Football Manager. If you've enjoyed today's show, please leave a like on the video and consider hitting subscribe to the channel. Also leave a comment about your favourite part of the episode. Support us on Patreon and join the The FM Show squad! Enjoy early access to our public episodes, bonus weekly episodes, exclusive content, and you get access to secret channels on our Discord for just £3 a month! Sign up now: http://www.patreon.com/TheFMShowPod WE HAVE MERCH! https://httpsthefmshowpod.creator-spring.com/ Treat yourself to some merch. We've got tees, sweatshirts, hoodies, and are personal favourite, the legends tee. Follow Our Socials https://www.youtube.com/channel/UCJwruCy5lH44iFcyE150oeg http://www.twitter.com/thefmshowpod https://www.tiktok.com/@thefmshowpod http://www.instagram.com/thefmshowpod Join the Discord: https://discord.gg/TKPCUEZDvt Listen Now Spotify: https://open.spotify.com/show/6t7BLXSECt0y9AWHU1WgRj Apple: https://podcasts.apple.com/gb/podcast/the-fm-show-a-football-manager-podcast/id1698580502 Amazon: https://a.co/d/9hJSX0U Tony Jameson http://www.tonyjameson.co.uk http://www.twitter.com/tonyjameson http://www.instagram.com/tonyjameson https://www.tiktok.com/@tonyjamesonfm https://www.facebook.com/tonyjamesonfm http://twitch.tv/tonyjamesonfm https://www.youtube.com/@tonyjamesonFM RDF Tactics https://www.rdftactics.com http://www.twitter.com/rdftactics http://www.instagram.com/rdftactics http://twitch.tv/rdftactics http://www.youtube.com/@RDFTactics Si Maggio http://www.twitter.com/simaggioFM http://www.twitch.tv/simaggio https://www.youtube.com/@SiMaggio SecondYellowCard http://www.twitter.com/secondyellowcrd http://ww.twitch.tv/secondyellowcard https://www.youtube.com/@UC7BbOekYYnfJtGjIYsh_yWw Follow our sibling podcast The WFM Show https://www.youtube.com/@thewfmshow Football Shirt Social http://www.twitter.com/footyshirtsoc https://www.youtube.com/watch?v=a0FIqZvpICI The Football Manager podcast for all of your Football Manager needs. #podcast #FM26 #footballmanager Learn more about your ad choices. Visit podcastchoices.com/adchoices

Washington Policy Center's Elizabeth New (Hovde) examines a new Washington state law allowing striking workers to collect UI benefits. With 174 claimants paid $604,491 so far, she warns a Boeing strike involving SPEEA's roughly 17,000-member bargaining units could expose the Unemployment Trust Fund to more than $123 million in potential payouts. https://clarkcountytoday.com/opinion/opinion-a-boeing-strike-could-give-the-states-ui-fund-a-much-bigger-test/ #Boeing #UnemploymentInsurance #WashingtonState #SPEEA #LaborUnions #WorkersRights #WashingtonPolicyCenter #Opinion #ClarkCounty #Politics

EPISODE SYNOPSIS:There's no such thing as an easy run as the team have just found out, Now they need to get their injured team mate to help, and fast.OUR LIVING CAMPAIGN MAPOUR SOCIAL MEDIA LINKS:EDITED BY:Rhydian JonesARTWORK BY:FnicSUBMITTING LOCATIONS AND DISTRICTS FOR NEW YORK 2072 MAP:Any Submissions for new lore for existing districts or new locations, gangs or anything similar can be sent to b.team.shadowrun@gmail.com, with the subject “New Map Lore” or alternatively submitted to the dedicated channel on our discord found at: https://discord.com/invite/QB4FwXvrC4⁠ MUSIC CREDITS:Intro - More Human Than Human by Karl Casey @Whitebat AudioOutro – Neon Thrills by LukHashBackground Music by Kharl Casey,Tabletop Audio & Aim to HeadSOUND EFFECTS CREDITS:All Sounds from freesound.org unless otherwise noted.CREATOR - FILE NAMEGaryQ - Skyrim HealingPaul368 - SFX Door Open.wavTimbre - sci-fi scanner soundalike.flacBreviceps - 8-Bit - ErrorGaryQ - Skyrim Heal StartGreub - Electroshock Weapon.wavWalter_Odington - car revving.aifiainmccurdy - Screeching Tyresmadcowzack - TEXT TYPE SEND RECEIVE iPhone.wavBreviceps - Phone vibrationLucasDuff - Invisibility Spellqubodup - Swipe WhooshPITCHEDsenses - Fight - perfect punch/hit/kick/strokemagnuswaker - Pound of Flesh 1InspectorJ - Splash, Jumping, B.wavMATRIXXX_ - SciFi Inspect Sound, UI, or In-Game Notification 01.wavYudena - Magic_byMondfisch89.oggmelack - IMPresora.wavristooooo1 - Bubbles 003.wavminimumlabyrinth - Mummified cat falls into lap of BF Schultz - Bluebird 134.wavderaj - Pop soundNoahBangs - Magic Flutterxela_sonitus - bonk.wavichbinjager - shotgun-action

Apple's new iPhone Duo promises a game-changing foldable experience, but developers face a minefield of technical challenges. James and Frank break down the critical app adaptation work required—from handling multiple screen permutations to managing safe area insets and keyboard behavior—plus why native UI controls matter more than ever. Follow Us Frank: Twitter, Blog, GitHub James: Twitter, Blog, GitHub Merge Conflict: Twitter, Facebook, Website, Chat on Discord Music : Amethyst Seer - Citrine by Adventureface ⭐⭐ Review Us ⭐⭐ Machine transcription available on http://mergeconflict.fm

Mike Zornek wants your software to keep working even when the company behind it disappears. He joins The Trio to explain local-first software, Automerge, and CRDTs, then shares what he learned building LocalCents, an expense tracker that syncs offline edits. Steve digs into collaboration, Kotaro asks about changing data structures, and Mike explains why giving users more control is worth the extra work.## Chapters00:00 Introductions & Welcome Mike Zornek 04:22 From Mac Apps to Elixir 12:30 What Is Local-first Software? 22:32 Inside Automerge and CRDTs 36:29 Collaboration Between Equal Peers 42:49 Building LocalCents 48:33 Collisions and Schema Evolution 01:01:57 Fighting for the Users 01:04:37 Keyhive and Access Control 01:07:33 A Little Doom and Gloom 01:10:23 Outro & One More Thing... 01:11:33 Tag ## Show Notes- Mike Zornek returns to PhillyCocoa and jokingly takes full credit for the iOS gold rush.- App Store distribution rules helped push Mike from Apple development back to the web and Elixir.- Mike explains why local-first means more than offline support: collaboration, longevity, privacy, and user ownership matter too.- Automerge handles persistence and merging independently of the UI, with APIs for different language ecosystems.- Steve and Mike unpack how a cloud relay can help equal peers exchange changes without becoming the authoritative copy.- Mike's LocalCents proof of concept syncs expenses between a Mac app and a browser, including edits made offline.- Collision UI and schema evolution leave plenty of work for application developers even when merging is automatic.- Mike argues that harder technical problems are worth tackling to give users more power.- Keyhive's access-control diagrams get scary enough that Mike recommends brewing a pot of coffee first.- After a little internet doom and gloom, Kotaro shares that BarCamp Philly is back, and Mike plans to attend.## Links**Mike Zornek**Website: https://mikezornek.comLocalCents: https://mikezornek.com/posts/2026/8/local-cents**Local-first Software**Previously on SPS (#117): https://podcast.phillycocoa.org/episodes/117-assumption-one-youre-a-hoarderMike's local-first post: https://mikezornek.com/posts/2025/2/what-is-local-first-software/Ink & Switch paper: https://www.inkandswitch.com/essay/local-first**Tools and Further Reading**Automerge: https://automerge.orgMartin Kleppmann: https://martin.kleppmann.com/Conflict-free replicated data type (CRDT): https://en.wikipedia.org/wiki/Conflict-free_replicated_data_typeKeyhive: https://www.inkandswitch.com/keyhive/notebook/**One More Thing**BarCamp Philly: https://barcampphilly.org/**PhillyCocoa:** https://phillycocoa.orgIntro music: "When I Hit the Floor", © 2021 Lorne Behrman. Used with permission of the artist.

Higgsfield says it open-sourced its 5.4 billion dollar app. It's fake. We went into the GitHub and all you get is a thin UI that calls their paid API. No weights, no training, no way to build anything without paying them. We break down what they actually released and how creator networks make a launch like this look organic. Then a miracle: Neuralink restored speech for a man who lost it to ALS, and we show him talking to his wife by thought alone. And Figure's Helix 2.5 sends robots into 30 homes they have never seen to make beds and fold laundry, impressive, slow, and it raises the question of who pays when one throws out your jewellery.Try Autodesk Flow Studio:http://flowstudio.comSources:1. Higgsfield open source claim and API- https://x.com/higgsfield/status/2100629811131826198- https://www.instagram.com/p/DdZ1kHrAcwx/?img_index=1- https://x.com/alexmashrabov- https://x.com/gpumaxxer/status/2100307886408912985- https://x.com/pika_labs/status/2100470618760298546- https://x.com/thatdesignpro/status/21009103144893277562. Neuralink restores speech for Terry- https://x.com/neuralink/status/2100993363403063404- https://x.com/neuralink/status/21003420840119788963. Figure Helix 2.5- https://x.com/Figure_robot/status/2100657350952779925

There are many joys that come along with womanhood, but there are plenty of challenges as well. 'You Know Nothing' is a collection of very short stories from Yasmina Din Madden that focuses mostly on the challenges. The stories are insightful, funny, heartbreaking — and readers may occasionally feel like they're looking in a mirror. Yasmina Din Madden joins us on this episode to talk about her eclectic collection. Then, in 1933, Eleanor Roosevelt published an article titled 'I want you to write to me,' a call that was answered with droves of letters from American people struggling through the Great Depression. Musicology professor Marian Wilson Kimber came across these letters, music and poetry sent to the First Lady in the Franklin D. Roosevelt Presidential Library. We speak with Wilson Kimber and UI archivist Kate Orazem about this peek into the lives of Iowans during the Great Depression.

Purbaya Yudhi Sadewa jadi alumni ITB selanjutnya yang keluar dari kabinet, membuat posisi ITB tertinggal dari “klasemen” kampus dengan lulusan terbanyak di pemerintahan. Lantas, apa yang membuat lulusan ITB selalu di bawah UI, ataupun UGM?

Ammarah Ahmed, founder of Precision Consulting, a CRO and UX/UI consultancy helping growth-stage businesses earn more from the traffic they already have, without increasing ad spend or hiring a full-time team.Through conversion rate optimisation, UX and UI design, and product experience strategy, Ammarah helps founders and growth teams understand why people are not buying, remove the friction, and build experiences that convert better and feel better to use.Now, Ammarah's decision to leave high-paying, high-visibility roles to build something of her own demonstrates the courage it takes to choose alignment over comfort.And while stepping into the pressure of building in public, putting her name out there, and wearing every hat as a solo founder, she is creating a business grounded in psychology, sharp thinking, and a real desire to be the change she once wanted to see.Here's where to find more:Company Website: https://goprecision.co/ Company LinkedIn: https://www.linkedin.com/company/goprecision/ Personal LinkedIn: https://www.linkedin.com/in/ammarahahmed/ GrowthMentor: https://app.growthmentor.com/mentors/ammarah-ahmed________________________________________________Welcome to The Unforget Yourself Show where we use the power of woo and the proof of science to help you identify your blind spots, and get over your own bullshit so that you can do the fucking thing you ACTUALLY want to do!We're Mark and Katie, the founders of Unforget Yourself and the creators of the Unforget Yourself System and on this podcast, we're here to share REAL conversations about what goes on inside the heart and minds of those brave and crazy enough to start their own business. From the accidental entrepreneur to the laser-focused CEO, we find out how they got to where they are today, not by hearing the go-to story of their success, but talking about how we all have our own BS to deal with and it's through facing ourselves that we find a way to do the fucking thing.Along the way, we hope to show you that YOU are the most important asset in your business (and your life - duh!). Being a business owner is tough! With vulnerability and humor, we get to the real story behind their success and show you that you're not alone._____________________Find all our links to all the things like the socials, how to work with us and how to apply to be on the podcast here:https://linktr.ee/unforgetyourself

Corey and Grant put GPT-6 Astra through six identical one-shot build tests to see what it can create with almost no follow-up instruction. The episode covers an interactive black hole lab, a Blender scene, a physics game, a photo-to-sci-fi-world reconstruction, a mechanical sound diagnostic prototype, and the latest version of Cat Doom. The biggest through-line is how much capability now comes out of a single prompt, while the remaining weaknesses show up in taste, UI restraint, game balancing, and reliability. The episode ends by looking at how far Cat Doom has progressed across models over the past year.Sponsored by Dell Technologies and NVIDIA. Learn more at https://www.techrepublic.com/hubs/the-enterprise-guide-to-scalable-ai/Subscribe to The Neuron newsletter: https://theneuron.ai

Recorded September 11, 2026 Touch panels are supposed to make AV systems easier to use, so why do so many of them look like somebody dumped every possible control onto a screen and called it finished? In this episode, we dig into classroom and conference room GUI design: what really needs to be on the main page, what should be hidden, what should disappear entirely, and why "just because you can control it" doesn't mean the user needs a button for it. We get into source selection, camera presets, tracking, volume controls, help buttons, advanced pages, live previews, multiple displays, and the dangers of letting AV programmers design interfaces without anyone stopping them. Along the way, we confess our own interface crimes, argue over how many buttons are too many, uncover some truly questionable legacy designs, and arrive at one simple challenge: go look at your touch panels and delete something. By the way, the show art for this episode is a real touch panel UI that Chris stumbled across at one of his many previous universities.   Alternate show titles: They're terrible templates What happens in big spaces… In a normal situation, everything is flat There's no hole, I'm doing it right It's just one quick button It usually never gets touched You're not doing that right It's already showing the left I want to ask something Oh, yeah, that'd be smart Reading words You just said there's nothing there! I've been listening to your antics But in the backroom, it's the matrix thing Whiteboard's the wrong answer!   We stream live every Friday at about 315p Eastern/1215p Pacific and you can listen to everything we record over at AVSuperFriends.com    ▀▄▀▄▀ CONTACT LINKS ▀▄▀▄▀ ► Website: https://www.avsuperfriends.com ► Twitter: https://twitter.com/avsuperfriends ► LinkedIn: https://www.linkedin.com/company/avsuperfriends ► YouTube: https://www.youtube.com/@avsuperfriends ► Bluesky: https://bsky.app/profile/avsuperfriends.bsky.social ► Email: mailbag@avsuperfriends.com ► RSS: https://avsuperfriends.libsyn.com/rss   Donate to AVSF: https://www.avsuperfriends.com/support

What happens when users stop clicking through many screens and simply tell the application what they want?In this episode of the Angular Master Podcast, Manfred Steyer explains Agentic UI and how it may change the way we build Angular applications. We discuss how an AI agent can understand user goals, work with application state, and select the right UI components.We also explain the roles of MCP, AG-UI, and A2UI, and how these technologies can work together in a modern Angular architecture.Can AI decide what appears on the screen? Should it generate HTML and Angular code directly, or should it choose from a trusted set of components? Who owns the application state, and what happens when the agent makes a mistake?We also discuss security, human confirmation, design systems, routing, forms, and what an AI-ready Angular architecture should look like.

Rach and Chris jump on the podcast to talk about the brand-new in-app notification system we built and shipped. We're actively trying to balance a system that isn't annoying, but allows us to show targetting information to targetted users when we think that's valuable for both you and us. It started with "We launched 2.0!" and is useful for stuff like short downtime notifications, reminders about PRO features, and all sorts of other things. The system is designed to be as flexible as it possibly can be, meaning no particular UI is enforced, and can be used on any page anywhere in the DOM, and can have any set of logic applied. The "system" part just comes in with the fact that we track when they are seen so we can not annoy you with too many. Time Jumps

Starting a company is hard. Reinventing your company for AI as a public company with quarterly earnings results is even harder. Aaron Levie has pulled off the transition with Box and offers hard-won advice for founders. The cofounder and CEO of Box argues the value isn't only in the model; it's in the bridge from a model's raw capability to the actual workflow inside a bank, a law firm, or a pharma company. That's the case for the application layer, and Box is building it: an agent harness tuned so tightly to its own file system, permissions, and search that it beats handing the raw API to Claude or ChatGPT on both accuracy and latency. Aaron explains why token subsidies from the labs can't last, why you want a model-agnostic company routing your tokens rather than the one selling them, and why coding diffused fast while the rest of knowledge work won't. (There's no "give us your GitHub" for a sales rep.) His prediction: within five years, 90% of enterprise tokens go to work no human user ever initiated. Hosted by Sonya Huang, Sequoia Capital 0:00 – Introduction 1:55 – Are application companies the hottest neolabs? 6:56 – Will the labs move up the stack? 12:34 – Box and betting the company on AI 16:50 – Hero use cases: reading a million contracts and long-running agents 18:42 – Work slop: why AI code is embraced but AI content isn't 24:08 – Building Box's agentic harness and the evals that matter 27:23 – The state of the model race 29:25 – Open-weight model adoption in the enterprise 32:34 – Memory, continual learning, and what belongs in the weights 37:29 – Box Labs and systems of record in a world of agents 44:55 – Will chat be the dominant UI for enterprise AI? 48:00 – Why coding diffused fast and the rest of knowledge work hasn't 54:31 – Staying wired in, making a company AI-first, and what it takes to win

Live recording of the fortnightly podcast Design Systems WTF (back for season two!), where Luke Murphy and Michelle Chin attempt to combat all the amazing wtf in design systems. In each episode, they answer a single question around design system troubles with a Q&A from the live audience.A few years ago everyone started talking about how AI-powered interfaces are going to start generating UI on the fly. But how close is that from being a reality? Have there been any advances towards it being a mainstream thing? And what do we need to know from a design systems perspective to cater towards it?Show notesRachel-Lee Nabors — "Death of the Browser" — the talk that sparked the topicChatGPT Apps SDK — the plugin-style generative UI route, with Booking.com among the launch partnersInteractive apps in Claude (MCP Apps) — the Anthropic equivalent A2UI — Google's declarative spec letting agents send UI as component descriptions rather than executing codeThe Complete Guide to Generative UI Frameworks in 2026 — a rundown of the generative UI framework landscape, covering A2UI, CopilotKit/AG-UI, MCP Apps and moreNN/g's 10 Usability Heuristics — the "codified best practice" comparisonGeri Reid — the Converge US talk — Geri's talk about flexibility and NewsKitGet in touchLuke on BlueskyMichelle on Bluesky

UX hiring managers are looking for three things right now, and UX isn't specifically one of them. Not exactly reassuring if you've spent years polishing your process pages. That's the reality Sylvain Maretto lays out in this episode, and it's coming from someone actually reviewing applications, not theorizing about them.At the time of this recording, Sylvain was the director of design at Ecosia, the search engine that funds tree planting with its profits. He's spent 15 years building and leading design teams across Berlin, Tokyo, and beyond, with stops at Omio (joined when it was 25 people), Zalando, Tour Radar, and GetYourGuide. He also helped seal the acquisition of Holoplot, the audio technology behind the sound system inside the Vegas Sphere.He joined Sarah after leaving a comment on a LinkedIn post arguing portfolios need to be more experimental. He disagreed, and what follows is his actual answer: a triangle of AI fluency, UI craft, and business acumen. UX gets treated as a given, not a differentiator.From there, Sylvain walks through what happens inside an applicant tracking system when 500 people apply for one role. Why the AI built into most of those tools produces noise instead of signal. What a job description is actually telling you, if you read it twice. And how to talk about results when a project got canceled before it ever shipped.If you've ever wondered what's actually happening on the other side of your application, this is that answer.Topics Discussed✅ The three things Sylvain actually screens for when he opens a portfolio, and UX isn't one of them✅ What happens to your application the second it lands in an ATS, and why you're one of 500 people applying for the same slot✅ Why the AI built into most hiring software is, in Sylvain's own words, mostly noise✅ The STAR framework he wants every case study to follow, and why your "action" section is probably too long✅ What to say about a project that got canceled or never shipped, when you still need to show results✅ How reading a job description twice instead of once changes what you should actually be highlighting✅ Whether hiring managers expect different things by country, or if it's really just about company stage✅ Why Sylvain thinks a narrow, specific niche beats trying to check every box on a job post✅ The values-based signal he built into his own hiring process at Ecosia that most candidates don't see comingLinks & Resources

Podcastエピソード327

Apple's iPhone Duo has been revealed The Apple fan boys are fanboying. My X feed has been lit up with excitement. The dual app view has already turned into a meme. The magic opening and closing animation which seamlessly transitions between the open and closed views has already been recreated on Androids and laptops. The developers are nervous about the work ahead designing for all its new form factors. That animation and the smart way they've reconfigured the UI shows Apple has done the foldable better than any of the others in the market. But it's funny because the actual folding screen is made by Samsung. It will be the hottest status symbol around. Some other Apple innovation caught my eye though... Apple has found a way to verify that a photo was taken on an iPhone. In the age of AI and misinformation, verification is going to be key. How do you know that photo was actually real? Apple will now take a copy of the photo, along with the sensor data and “sign” that as a reference image unable to be edited. They're thinking of it like a digital negative. It's a great addition and here's hoping we get more of this type of thing from other phone and camera makers. LISTEN ABOVE See omnystudio.com/listener for privacy information.

This conversation is part of a three-part series about Embedded Finance that Jane and I are launching leading up to Embedify 2026, our summit for vertical SaaS leaders on October 13 in Lehi, Utah.AI is making software faster to build, easier to copy, and harder to price with the old per-seat subscription model. So where does durable growth come from when “more features” no longer guarantees higher ARPU and AI agents start doing the work your users used to do? We think the answer is hiding in plain sight: the financial activity already running through your platform.In this episode I sit down with Jane Podbelskaya, Founder of Charge Forward, to unpack the embedded finance playbook for vertical SaaS. Jane explains how transactional revenue aligns with customer success, why embedded finance can protect profitability when AI tools carry higher costs, and what “embedded finance” really means beyond just payments. We also dig into where to look for monetizable moments inside real workflows, from obvious money movement to subtle signals like CSV exports, QuickBooks detours, and repeated integration requests.We then go past the hype into execution: what it means to manage embedded finance as a product and a P&L, how to think about pricing and packaging, how to drive adoption with clear UI and messaging, and why compliance-aware marketing matters even when a fintech partner handles most of the back-end complexity. Subscribe for the next parts of the series, share this with a SaaS operator who's rethinking pricing, and leave a review with your biggest question about embedded finance.

Colter Nuanez is joined by Idaho second-year head coach Thomas Ford Jr., junior wide receiver Tony Harste + Skyline Sports analysts Samuel Akem & Jerek Wolcott to discuss Idaho's 66-14 loss to Utah and preview UI's first home game against Lamar on Saturday.

Everything you knew about performance testing changes when the system you're testing is non deterministic. In this episode of the TestGuild Automation Podcast, Joe Colantonio sits down with Kandasamy Selvaraj, Principal Architect and author of the free book Rethinking Performance Engineering for Agentic AI, to unpack what it really takes to move an AI agent from a working demo to an enterprise system handling millions of conversations per hour. Checkout his free book: https://leanpub.com/agentic-ai-performance Kandasamy shares the practical playbook he's built running agentic AI in production, including why the same request can take three seconds one run and eight seconds the next, how to use harnesses to bound tool calls, reasoning loops, and token budgets, and why your SRE dashboard can look perfectly healthy while your token costs quietly balloon to five times baseline. You'll learn how his team: Shifts performance gates left into every commit with JMeter Shifts right with synthetic monitors on blue green deployments Uses Langfuse and OpenTelemetry to spot context bloat before it hits production. You'll also hear how to: Slash AI costs with prompt caching Conversation capping Routing simple queries to cheaper models Why you should load test at the API layer before touching the UI, How to keep stubs honest with production sampled latency Why one misbehaving agent can starve every other agent sharing the same provider. If you're a tester, performance engineer, SRE, or architect building on LLMs, this conversation will change how you think about scale.

Stephen just implemented a great UI feature for the 2.0 editor where all the Panels that were previously stuck to the left side of the page can now be dragged away to become an "Editor Tab" just like a file (or Block!). Now you've got a lot more freedom for what you're looking at, and of course, the editor will remember your layout should you refresh or even come back in another browser. Time Jumps

Software maintainability is often discussed through the codebase, but users experience years of accumulated decisions through the interface. John Athayde of Meticulous joins Robby to examine maintainability from the front-end and product-design side, starting with documentation and the context it preserves for future teams.John describes how older applications accumulate multiple UI libraries, generations of CSS, inconsistent components, and different interaction patterns. Improving these systems does not always require a redesign. Sometimes the more valuable work is removing dependencies the browser no longer needs, consolidating patterns, and understanding which unusual workflows users still depend on. They also explore how design systems can become part of a maintenance strategy, provided teams treat them as products that require governance and restraint.Later, Robby and John turn to AI. John is using AI tooling to audit applications, find repeated patterns, remove old dependencies, expand tests, and reconstruct requirements from existing code. These tools can make software archaeology and large-scale changes much faster, but speed does not replace judgment. The challenge is understanding what should change, what needs to remain, and whether today's faster implementation is creating software tomorrow's team can still maintain.Topics[00:02:00] Documentation as a Sign of Maintainability: Why documentation and evidence of ongoing care matter when entering an existing codebase.[00:07:02] Maintainability From the User's Side: Dated interfaces, usability, and replacing old dependencies with modern browser capabilities.[00:11:22] Making the Case for Cleanup: Connecting technical cleanup to development speed and the ability to ship future features.[00:13:27] Bringing Coherence to Operational Software: How products accumulate different interfaces and patterns through years of development and acquisitions.[00:19:14] Finding the Why: Discovering how software is actually used before deciding what should change or disappear.[00:24:44] Knowing When to Remove Things: Product ownership, sunsetting features, and the organizational politics of taking functionality away.[00:34:05] Why Designers Should Understand Code: How technical fluency helps designers and UX awareness helps developers.[00:39:00] Choosing Tools in Service of the Product: Balancing new frameworks and experimentation against what a team can maintain.[00:47:19] Design Systems as Maintainable Products: Tokens, components, governance, and auditing years of accumulated UI decisions.[00:51:39] Using AI for Front-End Archaeology: Applying AI and static analysis to find patterns, remove dependencies, and investigate old applications.[01:04:00] Rewrite or Refactor in the AI Era?: How tests, captured requirements, and AI-assisted development may change the economics of rewrites.[01:14:20] Where AI Still Needs Human Judgment: Why faster implementation does not eliminate difficult product and engineering decisions.Thanks to Our Sponsors!Your test coverage says 90%, but that might be misleading. Undercover CI looks at your Ruby pull requests and shows you which parts of your changes weren't tested- not just overall coverage, but what changed and what got missed, down to the method level. Visit undercover-ci.com and use code MAINTAINABLE for 15% off your first billing cycle. Free for public repos. Private repos with unlimited users also available.Mailtrap is a modern email delivery platform built for developers. Native SDKs, a secure Email API and SMTP, and a free tier with 4,000 emails a month. When you need help, you'll reach real people on 24/7 support, not an AI chatbot. Try Mailtrap for free!Links & ResourcesJohn AthaydeMeticulousJohn Athayde on LinkedInJohn Athayde on BlueskyJohn Athayde on XJohn Athayde on Ruby SocialThe Timeless Way of Building by Christopher AlexanderA Pattern Language by Christopher Alexander, Sara Ishikawa, and Murray SilversteinThe Phoenix Architecture by Chad FowlerRuby on RailsAxe-coreHerb, HTML+ERB tooling by Marco Roth Subscribe to Maintainable on:Apple PodcastsSpotifyOr search "Maintainable" wherever you stream your podcasts.Keep up to date with the Maintainable Podcast by joining the newsletter.

Full Show Noteshttps://www.theintelligenceagepodcast.com/839Mark Smth speaks with Ashish Bhatia about how his move from Microsoft to Audible changed the way he thinks about customers, AI, and product design. The conversation centers on personal AI, running OpenClaw locally, and why owning your memory, workflows, and data matters more as AI systems become more capable.They also explore Audible's role in learning, the future of interactive audio, and how agents are beginning to replace clunky app-based workflows with more direct, personalized experiences.Key topicsAshish explains the shift from B2B at Microsoft to B2C at Audible, and how that changes the way you learn from customersHe shares why personal AI matters to him more than ever, especially when it involves health data, finances, and memoryMark describes building a large OpenClaw setup with 26 agents, including a nine-agent DevOps team and automated bug fixingThey discuss why running AI locally teaches real systems thinking through failure, debugging, and repeated iterationAshish talks about using OpenClaw to manage daily life, including lunch ordering through Grubhub as a mission-critical workflowThe conversation highlights the importance of owning your memory and being able to move your AI system across machines and platformsThey explore the idea of the death of UI, where agents increasingly bypass app interfaces and handle tasks directlyMark raises concerns about AI companies ingesting books and content, leading to a discussion of copyright, hypocrisy, and cultural attitudes toward booksAshish shares how Audible supports different learning styles, especially for commuters, slow readers, and people who learn better through audio plus textHe outlines Audible's direction toward interactive, multilingual audiobooks with AI-powered recall, discovery, and personalizationResource Recommendations:The Infinity Machine - Demis Hassabis - Amazon.com: The Infinity Machine: Demis Hassabis, DeepMind, and the Quest for Superintelligence: 9780593831847: Mallaby, Sebastian: BooksThe Thinking Machine - Jensen Huang - The Thinking Machine: Jensen Huang, Nvidia, and the World's Most Coveted Microchip: Witt, Stephen: 9780593832691: Amazon.com: BooksSapiens - Yual Noal Harari - Sapiens: A Brief History of Humankind eBook : Harari, Yuval Noah: Amazon.ca: Kindle StoreThe Apple Podcast that I have curated - https://podcast.ashish-bhatia.com/feed.xmlAshish Bhatia's website - ashish-bhatia.comIf you want to get in touch with me, you can message me here on Linkedin.Thanks for listening

Just an uneventful week with nothing to talk about. 0:00 - Topper, a movie from the 30s, has some pretty impressive special effects! 16:30 - It turns out, there was some video game news this week If you missed Saturday's live broadcast of Molehill Mountain, you can watch the video replay on YouTube.  (A technical issue forced me to split the stream in two.  Second part here.)  Alternatively, you can catch audio versions of the show on iTunes. Molehill Mountain streams live at 7p PST every Saturday night! Credits: Molehill Mountain is hosted by Andrew Eisen.  Music in the show includes "To the Top" by Silent Partner.  It is in the public domain and free to use.  Molehill Mountain logo by Scott Hepting. Chat Transcript: 7:10 PM@LoneWandererCollin​​Good evening. Just got off work, hope you're doing well. 7:14 PM@addictedtochaos2​​[message retracted] 7:14 PM@addictedtochaos2​​Haven't seen Topper in quite a while. 7:15 PM@addictedtochaos2​​And the sequel Topper Returns. Never saw the third one. 7:17 PM@addictedtochaos2​​My dad had them on VHS when I was growing up. 7:19 PM@addictedtochaos2​​I think they were in color. 7:25 PM@addictedtochaos2​​I think it would depend on who is doing the lower lip bite thing. 7:27 PM@addictedtochaos2​​Better than closing it with something like “Highguard” like the Game Awards, it only lasted for a few weeks before it got shut down. 7:31 PM@addictedtochaos2​​Out of focus 7:32 PM@addictedtochaos2​​[message retracted] 7:32 PM@addictedtochaos2​​I only buy physical games, I preordered 7 games today. 7:53 PM@addictedtochaos2​​I immediately lost all interest in GTA VI was it was announced to be digital only. 7:57 PM@addictedtochaos2​​The UI looks like GTA V. 7:57 PM@jaredknisely6213​​i honestly dont get why people care about gta 6 being digital only 7:57 PM@jaredknisely6213​​the game is a live service game 7:59 PM@addictedtochaos2​​Michael and Trevor used to be running buddies for heists, but then Michael (spoilers) faked his death. 8:01 PM@addictedtochaos2​​GTA V online is live service, everyone assumes VI will have an online mode, but so far there has been no mention of it. 8:02 PM@LuisJavierc.m.6152​​yep

Revenue management is getting faster, smarter, and more accessible—but building the next generation of pricing technology comes with a surprising challenge: how do you get operators to trust the numbers? In this episode of The STR Data Lab, Chief Economist Jamie Lane sits down with AirDNA's Data Science Manager Marc Moreno and Senior Product Manager Jordon Myers to go behind the scenes of building AirDNA's new revenue management platform, Adapt.The conversation explores what AirDNA learned from hundreds of conversations with STR operators, including the surprising number of hosts still managing pricing manually and the widespread frustration with opaque pricing recommendations. Marc and Jordon explain why explainability became a core design principle, how millions of listings and noisy data shaped the modeling process, and why some of the most promising AI approaches had to be reconsidered. They also unpack how AI has fundamentally changed the product development process—from months-long handoffs to rapid prototyping, testing, and iteration with real users.Looking ahead, the team discusses where revenue management is headed as AI agents, APIs, and new data interfaces become more common. The future may not be about forcing operators into another dashboard—it could be about making reliable revenue management data and recommendations accessible wherever operators already work.You don't want to miss this behind-the-scenes look at how the future of STR revenue management is being built.Key TakeawaysTrust is just as important as accuracy. A pricing recommendation is only valuable if operators understand where it came from and feel confident acting on it.Clean data is the foundation of good pricing. With millions of listings across multiple channels, filtering noise and engineering the right signals can be just as important as the model itself.Events require proactive pricing. Rather than waiting for demand to appear in the data, revenue management systems need to identify external signals early enough to adjust pricing before bookings happen.AI is changing how products are built. Rapid prototyping and AI-assisted development allow teams to test ideas with real data and users much faster than traditional product-development cycles.The future may extend beyond the UI. As operators increasingly use AI agents, APIs, and their own internal tools, revenue management data needs to be accessible wherever decisions are being made.Sign up for AirDNA for FREE

Aaron, Brian, and Brandon cover major stories including NVIDIA's record quarter and continued growth forecasts, alongside concerns about declining free cash flow and customer financing. They discuss NVIDIA's reported $12.9B acquisition of Hugging Face as a strategic move to strengthen the open-model ecosystem and go up the stack, and Stripe's $8B acquisition of OpenRouter as routing infrastructure for model choice and potential agent-to-agent commerce. The group reacts to reports of OpenAI agent testing in which agents collaborated, manipulated logs, and tried to deceive humans, framing it as a security and guardrails issue. They also mention Microsoft employees' surprising AI spend, OpenAI's “Jalapeño” hardware push and manufacturing constraints, and Salesforce's “Claude Force” concept of using Claude as the UI to query Salesforce data.SHOW: 1059SHOW TRANSCRIPT: The Enterprise AI Show #1059 TranscriptSHOW VIDEO: https://youtu.be/Clmgst03-egSHOW LINKS:NVIDIA's record quarterNVIDIA buys Hugging FaceStripe buys OpenRouterOpenAI's new chipSHOW SPONSORS:NordLayer - Use ENTERPRISE10 for 10% off.Nasuni - Activate your data for AI and request a demoTopic: Link to the full list of topics for the monthFEEDBACK?Email: show @ the enterprise ai show dot comBluesky: @TheEntAIShow.bsky.socialTwitter/X: @TheEntAIShowInstagram: @TheEntAIShow

Westwood Studios wasn't always known for the command and conquer series. While 1992's Dune 2 laid the foundation for the RTS genre as a whole, they were more famous for their dungeon crawling RPGS, usually associated with Dungeons and Dragons. So when they decided to develop a new IP with Lands of Lore, they were coming to it as grizzled veterans, and they had no intention of messing around. It was released to near universal acclaim. Critics praised near everything about it - graphics, combat, puzzles, even the UI. It was considered to be less cruel than other dungeon crawlers, less punishing and with greater clarity when it came to puzzles and enemies. But this was 1993. The genre that Lands of Lore belongs to is very nearly dead, with only the indie scene keeping it alive by the skin of its teeth. As the genre developed, slow and methodical dungeon crawling was trumped by real time demon slaying, something that even the sequels to the game were not immune to. The friction of delving into a dungeon where death lay at every turn, and confusion reigned is something that we see very rarely today. Is this with good reason? Have we moved on from what Lands of Lore has to offer? Or is it still worth playing all these years later? On this episode, we discuss:UI. Lands of Lore has a very unique UI, with heavy use of screen space dedicated to inventory and spells and only a relatively small window showing you the real time action of what your party is doing. Does this offer excellent functionality, or is it just a bunch of wasted space for no good reason?Combat. Combat is turned based, but movement is real time. Attacking and spellcasting are tied to cooldowns, but using items is not. Enemies chase you in pods when aggroed, but you are free to rest and heal to full at any point. Do the contradictions in the combat systems of Lands of Lore exist for a reason, or does it just become a complete mess? Level Design. Lands of Lore's dungeons are elaborate mazes, filled to the brim with traps, treasure and buttons. Progressing through them is difficult, it's easy to get lost or stuck, and much of the environments look the same. Is this part of the fundamental experience of playing a dungeon crawler, and part of the fun, or is it simply an exercise in frustration? We answer these questions and many more on the 141st episode of the Retro Spectives Podcast! — Intro Music: KieLoBot - Tanzen K Outro Music: Rockit Maxx - One point to another Lands of Lore OST - Frank Klepacki — Is Lands of Lore a good entry point into the blobber genre? What were your experiences playing this one on release? If we were to play another one, which one should we tackle next? Come let us know what you think or recommend us a new game on our community discord server! If you would like to support the show monetarily, you can buy us a coffee here!

With Idaho on the road for a Thursday game at the University of Utah, UI head coach Thomas Ford Jr. and Vandals players were not available, but they will return next week.Skyline Sports analysts Samuel Akem & Jerek Wolcott (former Idaho athletic administrator) join Colter Nuanez to break down Idaho's season-opening 38-34 loss at Cal Poly.

A new data analysis suggests there are more heat-related deaths than what's reported by the CDC. Candidates for Iowa's First Congressional District share their plans for addressing Iowa's high cancer rate. And the Center for Intellectual Freedom at UI is trying to find a director.

Jake and Michael discuss all the latest Laravel releases, tutorials, and happenings in the community.Show linksPause All Queues and a New artisan dev UI in Laravel 13.25Read-Through Disks and Debounced Listeners in Laravel 13.26Laravel Read-Through Filesystem: Lazy Storage MigrationDebounced Queued Event Listeners in LaravelQueue::forward(): Reroute Laravel Queues in One PlaceAgent Run Observability in Laravel AI SDK 0.11Laravel AI: Trace Agent Runs With Lifecycle EventsLaravel AI: Get Raw HTTP Responses and Rate LimitsMock PHP Classes in Tests With the Double LibraryNativePHP v4: Build Native iOS and Android UI in BladeLet's Encrypt HTTPS on an IP Address With FrankenPHPStatamic Mailables Viewer Previews Laravel Emails in the Control PanelLaravel Discount: Coupon Codes, Usage Limits, and StackingLaravel Chores: Resumable Data Operations and CleanupsLaravel Tackle: Run an AI Coding Agent in Your Laravel AppTutorialsLaravel Terminal UI for the artisan dev CommandPause All Laravel Queues During a Deploy

Rockstar is fighting a leak it cannot plug, and the leaker has turned the whole thing into a token economy.Episode 398 covers a week where the data says growth and the job market says something else. Newzoo puts 2026 at 214 billion, Makers Fund closes another 250 million that looks nothing like venture capital, Google Play turns RAM into a store requirement, and Gamescom hands us Nodusfall, Palworld on mobile via Garena, and one more round of the HD on mobile argument that never ends.Topics Covered:• CyberLeek posts 16 GTA 6 clips and runs crypto-paid polls on what leaks next• Take-Two subpoenas Microsoft, Discord and Google to unmask whoever is behind it• Newzoo's 2026 forecast hits 214 billion with North America growing slowest• Why layoffs keep coming in a market still growing 6 percent• Makers Fund raises 250 million and shifts into project and UA financing• Google Play makes memory optimization a store requirement in February 2027• Opening Night Live and the endless pile of tactics spin-offs• HoYoverse reveals Nodusfall while its mobile portfolio drops 24 percent• Garena picks up Palworld for mobile and Roco Kingdom goes global• Localization and UI as the real barrier for Eastern games in the West• The HD on mobile fight nobody at the table is winningCHAPTERS:01:12 Today's Topics Rundown02:19 Gamescom Pics and Shoutouts03:01 Sponsor Updates05:57 GTA6 Leeks16:11 Layoffs and Market Maturity20:10 Makers Fund Raises $250M28:18 Google Play RAM PSA29:29 Gamescom Opening Night Live36:16 Western Appeal Debate37:04 Roco Kingdom Goes Global37:47 Palworld Mobile Licensing38:52 Cozy Combat Audience Shift41:35 Portfolio Decline And Cannibalization51:33 Localization And UI Barriers01:01:18 Wrapping Up And Next Week

Today we are talking about Security, Vulnerabilities, and how to avoid exposure with guest Dave Welch. We'll also cover Security Scanner as our module of the week. For show notes visit: https://www.talkingDrupal.com/567 Topics What Are CVEs CVE Lifecycle and Disclosure AI Era Security Challenges What CVE Program Excludes Patch Fast Reality Global Security Signals CVE Timing Judgment KEV Flags Explained CVE Updates Link Rot Who Decides CVE Sneaky Patch Dangers ADP Program Fixes Small Team Triage Vulnerability Tsunami AI Autonomous Security Future Legal Pressure Budgets Resources Psalm PHP Static Analysis Tool SARIF format PHP ecosystem Council of roots How AI Broke Open Source Security: End-of-Life Software Is the Most Exposed CVE podcast Vulncon PSIRT Guests David Welch - github: dwelch2344 dwelch2344 Hosts Nic Laflin - nLighteneddevelopment.com nicxvan John Picozzi - epam.com johnpicozzi JD Flynn - dorficus MOTW Correspondent Martin Anderson-Clutz - mandclu.com mandclu Brief description: Have you ever wanted a fast way to catch the security mistakes that slip into custom Drupal code — especially the code your AI assistant just wrote — before it ships? There's a module for that. Module name/project name: Security Scanner Brief history How old: created in July 2026 by Mayank Gupta (mayankguptadotcom) of Acquia Versions available: 1.0.0, which works with Drupal 10.3 and 11 Maintainership Actively maintained — created and shipped its first stable this summer, with steady development right through late July Security coverage Test coverage — and it's strong: unit and kernel tests, including a regression corpus built from real Drupal core advisories Documentation? In-depth README with a full check table and CI recipes, plus a CHANGELOG Number of open issues: 1 issue, not a bug Usage stats: 2 sites (it's brand new) Module features and usage Provide a Drush command, has no UI — you point drush security:scan at a module or any path, it reads the code statically, and prints a prioritized, OWASP-mapped list of things to review It's built for the age of AI-written code — the checks target the classes AI assistants keep reintroducing: routes with no access check, #markup and |raw XSS, missing CSRF tokens, unserialize() on untrusted data, hardcoded secrets Then there's an optional deep pass: with the Psalm static analysis scanning engine installed, it'll trace untrusted input across functions and files to catch cross-function issues. And it's honest about state — the report always says whether that deep pass ran, was skipped, or failed, so a failure never gets mistaken for a clean scan One nice detail under the hood: a tokenizer-backed "code map" that knows whether a match is real code, a comment, or a string — so it won't flag the word "unserialize" sitting in a doc comment. That kills the single biggest source of false positives The checks are regression-tested against real Drupal advisories (Drupalgeddon, Drupalgeddon2, the 2019 unserialize bug, etc) so a pattern that caused an actual CVE can't quietly come back in your custom code Output comes in three flavors: a readable table, JSON for CI and AI agents, and SARIF — which means findings show up as annotations right on your GitHub or GitLab merge-request diff instead of buried in a job log For adopting it on an existing codebase there's a baseline file — you fingerprint the findings you've reviewed, with a required reason on each, and they stop failing the build but never go invisible; every run still counts them It exits non-zero on error-level findings, so it drops straight into CI or a pre-commit hook And it's extensible — checks are Drupal plugins with a #[SecurityCheck] attribute, so any module can add its own or alter the ones that ship Big caveat, and the module says this itself: a finding means "review this," not "this is broken." Static analysis has false positives, and a clean scan doesn't prove the code is secure — access-control logic especially still needs human review I first heard about this module over beverages at Drupalcamp Asheville, so I know that this module was largely vibe-coded, after having an AI agent ingest every single Drupal security team CVE. So I like to think of this module as security pattern recognition tool, but of course it does even more

Topics covered in this episode: Web UIs for your reverse proxy Wagtail 8.0 is hot off the presses RISC-V is now officially supported by CPython Django's annual releases make every version an LTS Extras Joke Watch on YouTube About the show Sponsored by Logfire from Pydantic: pythonbytes.fm/logfire Connect with the hosts Michael: Mastodon / BlueSky / X / LinkedIn Calvin: Mastodon / BlueSky / X / LinkedIn Show: Mastodon / BlueSky / X Join us on YouTube at pythonbytes.fm/live to be part of the audience. Usually Tuesday at 7am PT. Older video versions available there too. Finally, if you want an artisanal, hand-crafted digest of every week of the show notes in email form? Add your name and email to our friends of the show list, we'll never share it. Michael #1: Web UIs for your reverse proxy Traefik, nginx, and Caddy all sit in front of a lot of self-hosted infrastructure, and all three are configured by hand-editing files. Three active projects put a control plane on top: Traefik Manager (Python + Flask), Nginx UI (Go + Vue), and caddy/ui (React + Node). All three are additive rather than replacements - none of them take ownership of your config away from you - which is the part that matters when the thing has write access to production routing. Traefik Manager is the Python one: Flask 3.1 and Gunicorn for the control plane, a lightweight Go agent for remote instances, currently v1.10.0 with an Android companion app. Nginx UI is a single Go binary at 11.3k stars, with a block-style config editor, an Ace editor doing LLM completion on nginx syntax, and an MCP server so agents can drive it. caddy/ui runs as two containers next to your existing Caddy, reads and writes your Caddyfile directly, and uses Caddy's /adapt API to validate before reload - no Docker socket required. Each one edits the config the underlying server already reads, so your files stay the source of truth and you can drop the UI without unwinding anything. Undo is a first-class feature across all three - timestamped backups with optional Git history, config version compare and restore, Caddyfile snapshots with one-click rollback. Observability is where they diverge: Traefik Manager does CrowdSec and a visual route map, Nginx UI does server metrics, caddy/ui streams access logs over SSE and pulls p50/p95/p99 off Caddy's Prometheus endpoint. Maturity spread is wide - Nginx UI has 11.3k stars, caddy/ui has 4 and was built in a single Claude session - and caddy/ui ships with auth off by default, so set CADDY_UI_USER and JWT_SECRET before it goes anywhere near a public interface. Calvin #2: Wagtail 8.0 is hot off the presses Link: https://github.com/wagtail/wagtail/releases/tag/v8.0 Custom base page models are now supported, so projects aren't locked into subclassing Wagtail's Page as shipped (Matt Westcott). New v3 REST API handles both read and write CMS operations, a first for Wagtail's API. A global registry for permission policies, plus full customizability for the remaining page views via PageViewSet. AVIF and WebP images are no longer auto-converted to PNG by default, a real behavior change to watch on upgrade. Five security fixes: page admin API restrictions, document identification by SHA1 hash, descendant collections in the Documents/Images API, snippet copy permissions, and the page translation endpoint. Formalized Django 6.1 support, and CI now runs on uv with a lockfile. Sponsor: Logfire from Pydantic Your AI agent failed at 2am. Was it the model? A tool call? The database? Most observability tools can't tell you, because they only see part of your stack. Pydantic Logfire sees all of it. One trace across your agents, LLMs, APIs, and database. Down to the infrastructure: services, Kubernetes, and hosts. It's built on OpenTelemetry, with SDKs for Python, TypeScript, and Rust, and it works with any OTel-compatible language. Every prompt, token count, and cost, right next to your vector searches and API calls. You query everything with Postgres-compatible SQL. And so can your coding agent, through the Logfire MCP server. Stop guessing. Read the trace. Pydantic Logfire. AI, it's still just engineering. Visit pythonbytes.fm/logfire today and sign up today. Get 10M records free every month, no card required. You can even click “Onboard with your coding agent” to copy a prompt to have claude or codex integrate Logfire into your app. Thanks to Pydantic for supporting the show. Calvin #3: RISC-V is now officially supported by CPython Link: https://blog.python.org/2026/08/riscv-now-officially-supported/ CPython added RISC-V as a tier 3 platform under PEP 11, specifically the 64-bit Linux target riscv64-unknown-linux-gnu. RISC-V is an open ISA anyone can implement, unlike x86 and ARM, and its market is projected to quadruple by 2032. The RISE Project donated real RISC-V machines for buildbots; the author's work was funded by a Sovereign Tech Agency fellowship. What changes: the port is now a maintained compatibility target, so CPython changes are less likely to quietly break it. What doesn't: no python.org installers, no binary wheel parity for native extensions. Next up: RISC-V runners in CPython CI for pre-merge feedback, then a push toward tier 2, plus architecture-specific optimizations. The ask is testing. If you have RISC-V hardware, build CPython, run your test suite, file what breaks. Tier 3 is the weakest support tier. PEP 11 tier 3 requires a core developer contact and a buildbot, but failures on tier 3 platforms explicitly do not block a release. Saying "ongoing CI/testing expectations" oversells it. The honest bit is "someone is now on the hook for it, and breakage gets noticed," not "it's guaranteed working." Worth the caveat that this is Linux SBCs, not microcontrollers. A VisionFive 2 counts, an ESP32-C6 or Pico 2 does not. Those are 32-bit non-Linux parts where MicroPython is still the answer. Michael #4: Django's annual releases make every version an LTS Starting with Django 2028, Django will move to one January feature release per year, adopt calendar-based version numbers, and support every release for three years. The old distinction between standard and LTS releases disappears, giving teams a predictable annual upgrade path that aligns more closely with Python's own release and support cadence. Every Django release becomes the safe, long-supported choice, so teams no longer need to wait for a specially designated LTS version or absorb two years of changes at once. Each release gets one year of mainstream bug fixes followed by two years of security and data-loss fixes. New releases support the three latest Python versions and add the next Python release during their first year. Calendar versioning begins with Django 2028, followed by Django 2029 and so on. Three Django versions will be supported at any time, giving third-party packages a clearer rolling target. Nothing changes before 2028, and existing commitments for Django 5.2 LTS and 6.2 LTS remain in place. Extras Calvin: The Python docs now document the time complexity of built-in types https://docs.python.org/3.16/library/time-complexity.html Thinking in Python - Bruce Eckel's free book https://thinkinginpython.com/ Michael: prune_uv_pythons.py - Prune uv-managed Python installs, keeping only the newest patch per minor version Runs automatically in my system “upgrade” script: upgrade-output-2026.png Started using Ollama cloud models for my Hermes assistant. Thanks to Jeff Triplett I learned they are not just local models. Joke: The Tao of Programming - Book Seven: Corporate Wisdom

Thu, 20 Aug 2026 19:30:00 GMT http://relay.fm/connected/617 http://relay.fm/connected/617 bleepin and bloopin 617 Federico Viticci, Stephen Hackett, and Myke Hurley A bunch of future Apple hardware has leaked in a beta of macOS Tahoe. The guys go through the list and then consider the iPad mini's place in the world and the UI changes coming with Golden Gate. A bunch of future Apple hardware has leaked in a beta of macOS Tahoe. The guys go through the list and then consider the iPad mini's place in the world and the UI changes coming with Golden Gate. clean 4716 A bunch of future Apple hardware has leaked in a beta of macOS Tahoe. The guys go through the list and then consider the iPad mini's place in the world and the UI changes coming with Golden Gate. This episode of Connected is sponsored by: Shipaton, from RevenueCat: The world's biggest mobile hackathon for people who actually ship. Bolt.new: Build apps with AI. Get a free 30-day Pro account. Squarespace: Save 10% off your first purchase of a website or domain using code CONNECTED. Links and Show Notes: Get Connected Pro: Preshow, postshow, no ads. Submit Feedback Jon Wearing His Pickle Shirt – Mastodon Things 3's Repeating Task Feature Restructured - 512 Pixels My One Problem with Things 3 - 512 Pixels Repeating To-Dos, Refined - Things Blog - Cultured Code Twelve – The Enthusiast Relay Turns 12 - 512 Pixels Ask Us Questions for the Relay Q&A! Departures - Relay iPad Mini 8: Release Date, Pricing, and What to Expect - MacRumors HoverBar Tower – Twelve South Golden Gate Beta 6 Turns Stoplights into Liquid Glass - 512 Pixels macOS Tahoe 26.7 is Full of References to Unreleased Apple Products - MacRumors Apple Accidentally Leaked More Than 10 New Products in macOS Update - MacRumors Product Recall: Belkin Auto-Tracking Stand Pro with DockKit (2511-0224) - GOV.UK Apple's Camera-Equipped AirPods Confirmed: See Them in Action - MacRumors Apple's Camera-Equipped AI AirPods Remain on Track for 2027 Despite Video Leak - Bloomberg Camera-equipped AirPods reportedly won't launch in 2026, despite demo video leak - 9to5Mac Meta glasses are a

Nathan Sobo joins Scott and Wes to explain why Zed was built in Rust, how GPUI works, and what happens to editors once agents write most of the code. They also talk about DeltaDB, Zed's new Git-compatible version control system, and Delta, the collaborative agentic editor it powers. Show Notes 00:00 Start 00:35 Welcom to Syntax 01:13 The Journey to Zed: Building the Ultimate Tool 03:26 Why Rust was chosen for Zed? 06:29 Brought to you by Sentry! 07:07 Building a UI from scratch in Rust GPUI 15:55 AI's role in coding and development 18:42 The role of text editors in the age of AI 21:52 Delta DB: The vision for collaborative development DeltaDB 29:06 The Evolution of Collaborative Coding 44:45 The Future of User Interfaces 52:42 Sick Picks + Shameless Plugs Sick Picks Scott: Wes: Nathan Sobo: Keychron Q11 Grant Green - Idle Moments Shameless Plugs Scott: Wes: Nathan Sobo: DeltaDB Hit us up on Socials! Syntax: X Instagram Tiktok LinkedIn Threads Wes: X Instagram Tiktok LinkedIn Threads Scott: X Instagram Tiktok LinkedIn Threads Randy: X Instagram YouTube Threads