Podcasts about Figma

Great leaders know how to turn the temperature down when everyone else is turning it up. Roxanne Petraeus discovered that early as a U.S. Army officer, watching the leaders she most admired become calmer as situations grew more stressful. Today, as CEO of Ethena, she draws on those lessons to lead through uncertainty, develop other leaders, and create an environment where people can think clearly instead of reacting from fear. You'll hear how Roxanne responded when COVID derailed a critical fundraising effort, why she believes bad news should never be sugarcoated, and what being one of very few women in both the Army and venture capital taught her about confidence and authenticity. She also explains why your most important work as a leader may be the conversations that produce nothing tangible, how your behavior signals what you'll tolerate, and why learning something new in public—including AI—may be one of the most powerful ways to lead. Roxanne is the co-founder and CEO of Ethena, a compliance training company trusted by more than 2,000 organizations, including Pinterest, Zendesk, Notion, and Figma. A U.S. Army veteran, Roxanne served as an engineer officer in Afghanistan and later as a Civil Affairs officer supporting Special Operations. She went on to become a Rhodes Scholar at Oxford and worked at McKinsey before co-founding Ethena.  Since launching Ethena in 2019, Roxanne has raised $50 million in venture capital while building a company focused on taking workplace compliance beyond simply checking the box. Throughout her career, she has maintained a deep interest in learning about and practicing effective leadership.  You'll discoverHow your calmness changes others under pressure Why transparency builds trust during uncertainty What helps you lead without being the expert How your behavior signals what you'll tolerate Why learning publicly gives others courage Connect with Roxanne Petraeus on Social MediaLinkedIn https://www.linkedin.com/in/roxanne-bras-petraeus-2292b8109/  Website Ethena https://www.goethena.com/Check out all the episodesLeave a review on Apple PodcastsConnect with Meredith on LinkedIn

When every individual on your team can suddenly move ten times faster, what stops them from moving ten times faster in ten different directions?In this episode of Supra Insider, Marc Baselga and Ben Erez sit down with Andrew Somervell, co-founder of Hamster, an AI-native workspace for product teams. Andrew and his co-founder built Taskmaster, the open-source tool that shipped plan mode before the major coding tools had it, and the question that kept coming back from users was how to take it to work. He explains why that pointed at a team problem rather than a tooling problem, and why he doesn't believe PMs want to live in a CLI.They explore the artifacts Andrew thinks a team actually needs, why he killed the PRD and what he ended up building in its place, the three meetings his team runs every week, and Ben's repeated pressing on the harder question underneath all of it: what genuinely changes when AI becomes multiplayer rather than a set of people each working alone with it.All episodes of the podcast are also available on Spotify, Apple and YouTube.New to the pod? Subscribe below to get the next episode in your inbox

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.

Arranca un nuevo episodio de Potencia.pro, el podcast de WordPress y cozas, con Miguel Ángel Terrón y Mariano Pérez al micrófono. Es el capítulo 336, segundo de la temporada… ¿undécima? ¿Décimo primera? Tras un pequeño debate lingüístico, y alguna broma de las que marcan la casa, lo dejan en «décimo primera», que suena menos a cacho de algo. Un plugin exclusivo… que nadie descarga Miguel Ángel recuerda que los suscriptores de Potencia tienen a su disposición el plugin que subió para ellos. Según él, «miles y miles de personas» se lo han pedido. En concreto, una: Madrillano, que le llamó por teléfono para que se lo pasara, a pesar de estar suscrito y poder descargarlo él mismo. Eso sí, le parece un plugin maravilloso. WordCamp Europe en Málaga Antes de entrar en materia, Miguel Ángel anuncia que ya se han publicado los organizadores de la WordCamp Europe que se celebrará en Málaga en mayo del año que viene, y él está entre ellos. Le encantan las fotos del equipo, con pelos de todos los colores. Lo que le preocupa un poco es tener que hablar con todos en inglés. Su plan: sonreír mucho y decir yes. Plugins hechos con IA por quienes más saben El tema principal de hoy nace de una reflexión de Mariano. Con la IA, cualquiera puede hacerse un plugin. Pero cuando quienes lo hacen son programadores expertos, cansados de usar herramientas de terceros que solo cubren el 80 u 85 % de lo que necesitan, el resultado es otra cosa. Es el caso de Fernando Tellado y Javier Casares, dos de los mayores referentes de WordPress en España. Se han hecho sus propios plugins para cubrir sus necesidades y, además, los comparten con todo el mundo. Los plugins de Fernando Tellado Fernando tiene 23 plugins en su perfil de WordPress.org. Mariano destaca algo que Fernando defiende con mucho énfasis: no son plugins «rastreros», de esos que usan el repositorio como escaparate. Esos plugins ofrecen una versión Lite que apenas sirve y te empujan a comprar la Pro. Los de Fernando están completos en el repositorio, sin versiones de pago. Además, son recientes, él mismo los usa en su web y los actualiza constantemente. DietPress Su logo es un tenedor, un cuchillo y una cuchara, lo cual tiene su gracia porque el plugin sirve para poner a dieta a WordPress. Mariano bromea con que debería estar tachado para que no coma. Es un optimizador en la línea de Machete, el plugin de Nilo. Elimina elementos antiguos o innecesarios que WordPress carga sin necesidad: Activa o desactiva características no esenciales. Reduce las peticiones a la base de datos. Acelera la carga de la página. La versión anterior, programada a mano, ofrecía unas diez o doce opciones. Con la ayuda de la IA ahora supera las cincuenta. Vigía Pensado para hacer GEO, es decir, para que las inteligencias artificiales vean mejor tu web: Te da una puntuación de cómo te están viendo las IA. Analiza más de cincuenta rastreadores de aplicaciones de IA. Permite bloquear los bots que no te interesen, por ejemplo los de las IA chinas. Mariano lo ilustra con el ejemplo de quien compró una bombilla fundida en un chino y ya no quiere saber nada. Miguel Ángel, en cambio, confiesa que le encantaría ser chino. Genera el archivo llms.txt, una especie de robots.txt con directrices para que la IA lea mejor la web. Mariano cree que de momento no sirve de mucho, aunque admite que puede estar equivocado. Crea un esquema JSON de la web. Vigilant Es el favorito de Mariano y, según él, probablemente el menos usado de los tres. Se trata de un plugin de seguridad 100 % gratuito. Lo sitúa como una alternativa a Wordfence o Sucuri que protege más y carga menos la web. Va por la versión 3.0.1 y su última novedad técnica es la autorreparación: si un virus infectara su código, es capaz de eliminarlo y restaurarse tal como estaba. Fernando lo explica en un artículo de su blog. Los plugins de su web En su propio dominio, Fernando tiene otra sección de plugins que no pasan por el repositorio. Allí la mayoría también son gratuitos: Un traductor con IA, pensado sobre todo para tiendas WooCommerce. Recuerda a Weglot, pero funciona de forma distinta. No toca la base de datos: al servir la página, pasa el contenido por la IA que tengas conectada y la muestra traducida. Mariano había visto algo parecido en inglés con pago mensual, y aquí es gratis. Un revisor de datos de productos para Google. Analiza todas las fichas y te dice qué está mal y qué se puede mejorar, algo ideal para tiendas de 3.000 productos que nadie va a revisar a mano. Dos de pago: una capa de consentimiento para cumplir la RGPD y otro relacionado con formularios. Los plugins de Javier Casares WPVulnerability Para Mariano es el plugin estrella de Javier y lo tiene instalado en todas sus webs. Se conecta a una base de datos de vulnerabilidades y revisa todo: servidor, plugins y temas. Te avisa de cualquier problema de seguridad pendiente de arreglar. Javier lo programó a mano en su día y ahora, con la IA, lo ha mejorado muchísimo. Los plugins de su empresa En la web de su empresa, Javier ha liberado plugins que desarrolló para clientes. Son más pequeñitos que los «macroplugins» de Fernando, pero están muy bien hechos: 2FA: autenticación en dos pasos, como el plugin de WordPress, pero con su criterio. Si lo miras por dentro, funciona mejor. SMTP: para enviar correos desde WordPress, esa fuente eterna de dolores de cabeza cuando el servidor no deja mandarlos. Asegura el envío, reintenta y se autoevalúa si hay errores. Secrets: guarda cifradas, o «encristadas», como bromean, las claves de API y de servicios de IA. Así no quedan en claro en la base de datos si esta se ve comprometida. Telemetry Disable: WordPress envía datos de la instalación, como número de usuarios o de entradas, que no son necesarios para funcionar. Este plugin elimina toda la telemetría no esencial y deja solo lo indispensable. Frecuencia: automatizaciones para podcasts, con transcripción automática y resúmenes por IA. Cuesta 240 euros y Mariano se lo dedica a Miguel Ángel como «chimpún» final. El aviso sobre Alexa A raíz de la telemetría, la conversación deriva hacia Alexa. Miguel Angel advierte de que los altavoces graban mucho más de lo que creemos. Además, cualquiera con acceso a la cuenta puede escuchar esos audios en Amazon. Su consejo: si vais a hacer algo inconfesable en casa de vuestros padres o de vuestros hijos, desactivad el micrófono. Noticias WordCamp Galicia Los dos tienen previsto ir. Todavía queda un mes y ya hay más de 150 entradas vendidas. Mucha gente guapa y una organización que lo hace muy bien. Ipsum, el nuevo tema por defecto de WordPress Se ha presentado el tema que llegará con WordPress 7.2, previsto para antes de fin de año. Lo desarrolla un equipo de 56 personas y trae una gran novedad: por primera vez no llevará el nombre del año. Se llamará Ipsum, como el lorem ipsum de toda la vida. Es un tema tipo lienzo en blanco, al estilo de GeneratePress o Astra. Incluirá muchos patrones, tipografías y opciones de personalización. Tiene su historia: el equipo había diseñado inicialmente un tema mucho más artístico y elaborado. Al presentarlo, se pidió algo más neutro para que cada cual pusiera lo que quisiera. Ese primer diseño no se tirará a la basura, porque también lo publicarán aparte. Mary Hubbard, presidenta de la Open Website Alliance Miguel Ángel confiesa que siempre se enamora de quien está al lado de Matt, por aquello de la «erótica del poder». La noticia es que Mary Hubbard, responsable de WordPress.org, asume también la presidencia de la Open Website Alliance, la alianza de proyectos de software libre. La presidencia es rotatoria, como la de la Unión Europea, y ahora le toca a WordPress. Purple, el primer tema de bloques de WooCommerce Tras mucho trabajo con el checkout en bloques, WooCommerce enseña una preview de Purple, su primer tema oficial hecho con bloques. Como indica el nombre, es moradito y muy bonito, y se puede probar para ver cómo queda una tienda con él. Jurassic.Ninja El descubrimiento de la semana: Automattic tiene un servicio para crear sitios de prueba con un nombre genial, Jurassic.Ninja. No es un WordPress en el navegador como WordPress Playground, sino un sitio alojado de verdad y con Jetpack incluido. A Miguel Ángel le viene de perlas. Para probar su plugin de generación de posts usaba Playground, que no permite las llamadas externas que su plugin necesita. Plugins del día Kiki Mariano cuenta el que le «hizo llorar». Él vive feliz con GeneratePress y Gutenberg, hasta que descubrió un nuevo constructor visual al estilo Elementor llamado Kiki. El nombre prometía, pero ni el nombre ni la chica que aparece en la web resultaron ser lo que él había leído, así que sigue sin cambiarse. Para quien no se deje llevar por estas cosas, promete diseñar libremente, como en Canva o Figma. Después genera los bloques de Gutenberg correspondientes con una interfaz más moderna, y se puede empezar gratis. WP Shipyard Shipyard significa astillero, y la metáfora le va que ni pintada. Este plugin gratuito permite al administrador cambiar de tema en privado: solo él ve el nuevo diseño mientras lo prepara, y el resto de visitantes sigue viendo la web de siempre. Cuando todo está listo, el barco sale del astillero y se publica el nuevo tema. No hace falta poner la web en mantenimiento. Agenda y despedida El lunes de la semana que viene hay meetup de WordPress en Sevilla. La próxima WordCamp en España será la de Valencia y luego llegará la de Galicia. Por el camino asoma también la de Faro, que para eso estamos en Iberia. Ya puestos, fantasean con vivir la reunificación de la península. Como manda la tradición, el episodio no se cierra sin chiste. A un amigo de Mariano le dijeron los médicos que tenía el hueso descalcificado. ¿La respuesta? «Bueno, lo importante es participar». 🤖 El contenido de este post ha sido generado automáticamente con inteligencia artificial a partir de la transcripción del audio. Puede contener errores o imprecisiones. 🎙️ Publicado con VozCaster, el bot de Telegram que convierte tu voz en un episodio de podcast publicado. Pruébalo gratis. ¿Te ha gustado el episodio? Si quieres que sigamos experimentando con bots, protocolos y empanadillas polacas, no olvides suscribirte y dejarnos tu valoración. ¡Nos escuchamos en el próximo capítulo! Métodos de contacto Enviadnos vuestras preguntas al grupo de Telegram. Apuntaos al canal de Youtube del podcast https://www.youtube.com/potenciapro Si nos queréis decir algo directamente lo podéis hacer a @potenciapro , @materron, @mpc, o en el grupo de Telegram Y si eres muy muy muy fan del podcast Echa un vistazo a cómo nos puedes ayudar en https://potencia.pro/se-prosperoso/

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

Tänases Algorütmi episoodis räägime, mis juhtub siis, kui disainer ei anna arendajale ette Figma faili, vaid ehitab prototüübi otse koodis. Arutame, miks päriselt klikitav prototüüp annab kliendilt paremat tagasisidet, kuidas komponentide ja Gitiga töötamine vähendab disaini ja arenduse vahelist tõlkimist ning millal jääb Figma interaktiivsete lahenduste puhul hätta.Räägime ka AI rollist disainitöös: millal tasub töö AI-le anda, millal on kaks pikslit kiirem ise ära liigutada ja kuidas häälega antud tagasisidest saab AI-le täpne tööde nimekiri. Jõuame ka selleni, kuidas AI loodud kasutajaliidest hinnata, miks disainisilm on õpitav oskus ning mida tähendab õppida vaatama mitte ainult ekraanil olevaid elemente, vaid ka nende ümber olevat tühjust.Juttu tuleb ka Elise Digiti workshopist „Train your design eye“, praktilise disaini reeglitest ja accessibility'st.Külas on Elis Laasik, Codeborne'i UI/UX-disainer, kes prototüübib suure osa ajast otse koodis. Elis viib oktoobris toimuval Digit konverentsil läbi töötuba „Train your design eye“, praktilise disaini reeglitest.-----Jaga meile enda jaoks olulisimat mõtet episoodist meie Discord kanalis: https://discord.gg/8X5JTkDxccEpisoodi veavad Erik Jõgi ja Martin KappAlgorütmi toetavadLHV https://www.lhv.ee/Nortal https://nortal.com/Codeborne https://codeborne.com/

Jack Raines turned $6,000 into roughly $400,000 trading SPACs, then lost $157,000 in three minutes after taking a very different bet. The author of Young Money joins Matt Zeigler to discuss what that experience taught him about investing risk, the urge to chase more, and the time a growing portfolio can cost you.The conversation follows Jack Raines from SPAC warrants and market narratives to a broader question: how do you allocate money, time, and risk to build a life you actually want? They explore the limits of a trading edge, why cheap stocks can get cheaper, and how debt, career choices, and status shape the decisions investors make beyond their portfolios.Young Money: A Field Guide to Wealth and Purpose in Your Twentieshttps://amzn.to/4AtUsDoJack Raines' Young Money newsletterhttps://www.youngmoney.co/Jack Raines on Xhttps://twitter.com/Jack_RainesTopics covered:How Jack Raines grew his Roth IRA through SPAC warrants and pre-merger common shares, and why those positions carried different risks.Why an edge in one corner of the market did not translate into a successful concentrated earnings bet.How a million-dollar target and constant account checking changed Jack Raines' relationship with money and time.What SEC filings, Discord research, and market narratives contributed to his trading process.Why Jack Raines bought Figma after questioning the market's AI narrative and speaking with designers.Applying portfolio thinking to careers, creative projects, and opportunities that change with each stage of life.Weighing retirement saving against other opportunities, and why Jack Raines treats expensive debt as a constraint on taking risks.Escaping the "someday" trap while giving long-term venture investments time to develop.Combining steady income and index funds with independent bets, while treating status as a tool rather than a goal.Why Jack Raines believes investors learn about risk through experience, with stakes they can afford to recover from.Timestamps:00:00 Jack Raines on the lessons of losing money04:57 Inside the SPAC trades and the $157,000 loss15:16 Separating market hype from downside math22:18 Building a life with portfolio principles29:45 Retirement saving, trading obsession, and the cost of time36:56 Debt, freedom, and the trap of waiting for someday44:33 Venture investing: acting quickly and waiting patiently51:03 Using status without making it the goal56:08 The investing lesson experience has to teachLearn more about the Excess Returns podcast network:https://excessreturns.coNo information discussed in this podcast should be construed as investment advice. Securities discussed may be held by the hosts and guests, their firms or their clients.

What do you do when you're putting in the same hours, running the same playbook, and the numbers stop moving?In this episode of Supra Insider, Marc Baselga and Ben Erez talk through the slowdown their business hit over the summer and how they worked their way out of it. They walk through the blockbuster April that became a false baseline, the moment in late June when the trend was undeniable, and what the analytics revealed once they finally looked properly: traffic was down, and the traffic they had wasn't converting.They explore why Marc thinks burnout comes from inputs not matching outputs rather than from working hard, what momentum actually is and why losing it is so hard to shake, the content system they built that produced consistently bad posts, and the realization that no amount of good execution rescues a weak idea. They also lay out the protocol they'd use the next time one of them stops having fun.If you're running a business through a flat stretch, building a content engine that isn't landing, or working with a partner and wondering how to say the hard thing out loud, this episode is for you.All episodes of the podcast are also available on Spotify, Apple and YouTube.New to the pod? Subscribe below to get the next episode in your inbox

Dans cet épisode du podcast Design +, je t'explique pourquoi vouloir tout montrer dans son portfolio quand on est UX Designer, UI Designer, Product Designer senior ou Lead Product Designer est la meilleure façon de dévaluer son expertise.Je décortique le cas de Julien, Product Designer avec 12 ans de métier. Dès la page d'accueil, il montre 9 projets ; de l'UI, de l'UX, du Design System, du Branding, du management ... et un constat brutal : il n'attirait que des opportunités floues, mal payées ou hors-sujet.Vouloir "garder toutes les portes ouvertes" pour ne rien rater est le piège n°1 des profils expérimentés en design numérique.Dans cet épisode, j'explique :pourquoi le profil "couteau suisse" inquiète les recruteurs et les clients qui cherchent un UX Designer Senior ou un Lead Product Designerles 3 signaux négatifs envoyés par un portfolio trop hétérogènela règle de la suppression des 70 % pour ne garder que 3 use-cases stratégiquescomment structurer un use-case Senior axé sur le contexte business et les arbitrages plutôt que sur un déroulé méthodologique scolairecomment transformer ton portfolio en filtre sélectif au service de ton futur positionnement.Si tu es UX Designer, UI Designer ou Product Designer avec plus de 6 ans d'expérience et que ton portfolio ressemble encore à un catalogue généraliste, cet épisode va t'aider à trancher.Ce que tu vas comprendre dans cet épisode : Pourquoi la polyvalence excessive dévalue la perception d'un Product Designer ou UI Designer seniorComment sélectionner uniquement les 3 use-cases qui servent ton ambition futureLa différence majeure de structure entre un projet junior et un use-case de Lead Product DesignerComment faire de l'élimination de tes anciens projets ton plus fort levier d'attractionEt pourquoi un positionnement clair nécessite d'abord de renoncer à ce qu'on ne veut plus faireLes questions : Ton portfolio ressemble-t-il à une armoire à souvenirs ou à un outil d'intention ciblé pour ton métier de UX/UI ou Product Designer ?As-tu le courage de supprimer 70 % de tes projets passés pour renforcer l'impact de ce qui reste ?Tes use-cases racontent-ils des arbitrages stratégiques ou un simple déroulé d'outils Figma ?Quelles sont les opportunités tièdes que tu continues d'attirer à cause de ton flou actuel ?UX Designer UI Designer Product Designer Lead Product Designer Portfolio Designer Lead UX Positionnement Mentoring UX Impact Portfolio Design+Si tu t'es reconnu dans cet épisode, ne reste pas seul avec ça. Je propose des accompagnements pour t'aider à débloquer concrètement ton portfolio et ta recherche. Contacte-moi en DM sur Linkedin : https://www.linkedin.com/in/laurentgallen/Design + est le podcast pour tous les designers UX, UI et Product. Dans la saison 12, on remet les pendules à l'heure sur ta carrière, sur ton positionnement et ton rapport au travail.

הדרך שבה אנחנו מעצבים משתנה - אבל האם התפקיד של מעצבים באמת משתנה יחד איתה?בפרק הזה של רדיו באטן מארחת יהודית אשר את שיר זלצברג, מנהלת עיצוב ב Salesforce, לשיחה על שינוי, AI והעתיד של עבודת העיצוב.שיר משתפת איך הצוות שלה כבר מתנסה בעבודה שבה עוברים מוקדם יחסית מ Figma ל prototypes בסביבת קוד ועובדים יחד עם Product ו Engineering.וזה לא תמיד מהיר או קל יותר.אבל מבחינתה, דווקא עכשיו חשוב להתנסות - גם כשהכלים עדיין לא מושלמים.כי AI אולי מטשטש את הגבולות בין התפקידים, אבל היכולת לייצר משהו היא לא תחליף למומחיות: מעצבים עדיין אחראים להבין משתמשים ולייצר עבורם חוויה טובה.הכלים משתנים. הדרך משתנה. האחריות המקצועית נשארת.עריכה: מרינה ברודסקיהאזנה נעימה

What are hiring managers actually looking for right now, when everyone has a polished resume and nobody knows what the job will look like in a year?In this episode of Supra Insider, Marc Baselga and Ben Erez welcome back Jori Bell, VP of Core Experience at Hampton, where she runs the peer group experience for high-growth founders. Jori sits at an unusual intersection: she has spent the past year hiring for her own team while also coaching product leaders through the existential side of this market. She explains the mindset she screens for, borrowed from the drama triangle, and why she would rather hire someone adaptable than someone who has done the exact job before.They explore why a brand name on your resume opens doors but no longer closes them, the “why you” question most candidates never answer well, what actually makes an outreach message worth acting on, and what a year of hiring taught her about connections, in-person onsites, and treating a job posting like a marketing campaign.If you're job searching and unsure what to lead with, hiring and rethinking your process, or a product leader trying to figure out what this moment asks of you, this episode is for you.All episodes of the podcast are also available on Spotify, Apple and YouTube.New to the pod? Subscribe below to get the next episode in your inbox

Will is back in the host seat this week to chat to the one and only Kelli Lucas, Co-Founder & Chief Design Officer at LunarLab. In this episode, Will and Kelli discuss, amongst many things; the push and pull relationship between designers and developers, alignment in the workplace, the love/hate relationship with Figma, how AI has affected the process of design projects, and why you shouldn't rely on AI to build your business plan. — Our guest for this episode has been Kelli Lucas. If you'd like to get in touch with Kelli, or chat about working together on your next UX/UI Design project, you can visit the Lunar Lab website. or connect with them via LinkedIn. Your host for this episode has been Will Larry, you can find and connect with Will over on LinkedIn. Don't forget you can now also watch the show over on our YouTube channel! If you would like to support the show, head over to our GitHub page, or check out our website. Got a question or comment about the show? Why not write to our hosts: hosts@giantrobots.fm This has been a thoughtbot podcast. Stay up to date by following us on social media - LinkedIn - Mastodon - YouTube - Bluesky © 2026 Giant Robots Smashing Into Other Giant Robots Podcast

A late-night DM from a listener named Chris asked me something I couldn't answer: did I lose my identity as a designer? My team just went one hundred percent Cursor, no more Figma. This season, I figure out what AI actually changes about the work, and what it never touched.When the tool that defined your job for a decade disappears overnight, what part of “designer” actually leaves with it?This is the first solo season in the history of Beyond UX Design. No guests, no interviews, just me working through one question that landed in my DMs late one night from Chris, a listener I've known for years. He asked whether I felt like I'd lost my identity as a designer after my team dropped Figma and went fully into Cursor. My gut said no. Thirty seconds later, I realized I didn't actually know. That question became this whole season.Episode one is about the world right before the change, back when Figma was just how we worked. I walk through what a normal design process looked like on my team, and then I get honest about something we don't talk about enough: how much of “being a good designer” got quietly measured by tool fluency. Named layers, auto layout, clean files. None of that is capital-D design, but it stood in for it because it was the thing we could see, screenshot, and post about on LinkedInI also take on a good question from the original LinkedIn thread that inspired this. If you could always design and build in code, why is this suddenly a big deal now? My answer comes down to opportunity cost, and it reframes what AI actually changed. Spoiler: it wasn't the ability to code. It was the time. If any of this sounds familiar, whether your team already made the switch or you can feel it coming, hit play and come along for the season.Topics:• 00:00 – The DM from Chris: “Do you feel as if you lost your identity as a designer?”• 02:40 – The late-night LinkedIn message• 03:20 – Why the identity question never crossed my mind• 04:10 – 127 comments, and why they were almost all tactical• 05:30 – This isn't just a designer problem: PMs and engineers too• 06:00 – What “normal” looked like: the old design-to-handoff process• 07:30 – How designers quietly ranked each other on tool fluency• 09:20 – Why tool skills got mistaken for design skills for a decade• 10:40 – Derek's challenge: “We could always design in code, so why now?”• 11:20 – My front-end dev origin story and why it doesn't settle the debate• 12:50 – The real answer: opportunity cost, not ability• 13:50 – What AI actually changed: collapsing the time to do both• 14:40 – Were there warning signs? The pilots I read as experiments• 16:00 – The real danger isn't replacement, it's belief• 16:30 – What survives so far: research, flows, judgment, taste• 17:10 – What still gets me, and a look ahead to episode two—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

One month before Bolt exploded, Eric Simons was preparing to wind the company down. In this episode of The Product Podcast, Carlos (CEO at Product School) sits down with Eric Simons, CEO of Bolt, to tell the story behind one of the fastest revenue ramps software has ever seen, and what he learned building toward it for seven years first.Eric is a high-school dropout who approaches life from first principles and treats endurance sports as training for the pain of building a company. He and Carlos trace the near-death moment (a board meeting set to start spinning the company down was one month away), the bet on WebContainers (an operating system Eric's team wrote to run in WebAssembly inside the browser, so you can build real full-stack apps from a browser tab, the way Figma brought design to the web), and the ramp that followed: zero to about $5M ARR in the first 30 days, then to roughly $20M ARR within two months, a curve that normally takes companies years. He explains why StackBlitz is now going deep on professional builders (PMs, designers, engineers), how Bolt is solving the hard part of turning AI-generated prototypes into production code your engineering team will actually sign off on, and why he frames the market like the ChatGPT-vs-Anthropic split: pick a specific ICP and workflow instead of spraying money at everything. He closes on the open-weights bet (StackBlitz was the earliest backer of Vite), aggressive usage-based pricing, and why cheaper open models are closing the last bit of distance to the labs.What you'll learn:- The near-death story: one month from shutting down, then a record revenue ramp- Why the WebContainer bet (a real OS running in the browser) took seven years and made Bolt possible- What a 0-to-$20M-ARR-in-two-months curve actually looks like, and why it's so rare- Why StackBlitz is focusing on professional product builders, not just cute prototypes- How to turn an AI prototype into production code your engineering team signs off on- The "ChatGPT vs. Anthropic" framing: why going deep on one ICP beats spraying money everywhere- Why first-principles thinking and endurance sport shape how Eric builds- The open-weights bet, aggressive usage-based pricing, and open models catching up to the labsConnect with Eric Simons, CEO, BoltLinkedIn: https://www.linkedin.com/in/eric-simons-a464a664/Host: Carlos, CEO at Product SchoolLinkedIn: https://www.linkedin.com/in/villaumbrosia/About Bolt / StackBlitz: StackBlitz is the company behind Bolt (bolt.new), an AI-powered web development platform that lets anyone build real full-stack applications from a browser tab, powered by its in-browser WebContainer operating system. StackBlitz was also the earliest backer of the open-source Vite project.About the Product Podcast: Product School's podcast brings you candid conversations with the founders and product leaders shaping tech.Social Links:Find out more about Product School hereFollow our Podcast on TikTok hereFollow Product School on LinkedIn here

This episode is a compilation of answers to YOUR questions that were asked directly from my listeners who attend my weekly business education YouTube live webcast. I'll be covering the topic on: Stop Trading Time for Money: The 5 Levers That Scale Any Business and more. Refer to chapter marks below for a complete list of topics covered and to jump to a specific section. Book a call with us to help you scale your side hustle: https://harounventures.com/startGet mentored by Chris: Book a Zoom call to discuss joining my Business Academy, Finance Bootcamp (to get a job in finance) or MBA Degree Programs or for investing/business/personal development coaching: https://haroun.short.gy/1on1CallYTWDownload my free "Networking eBook": www.harouneducation.comAttend my weekly YouTube Live every Thursday's 8am-11am PT. Subscribe to my YouTube Channel to receive notifications. Chapter Marks: 0:25 Welcome & Intro 0:38 Turn Your Side Hustle Into a Business That Scales 2:23 My Background and Experience 3:31 What Happens to Your Revenue If You Stop for a Month? 5:33 The Five Levers of a Scalable Business 21:52 Q&A: Trump, Canada, and Trade 23:46 Q&A: Risk Factors & Becoming a Competent New Hire 26:47 Q&A: T-Bills, Yields, and the Fed 30:32 Q&A: Big Tech Power and Unemployment Risk 33:18 Q&A: Education, Debt, and Online Teaching in the AI Era 37:52 Q&A: BYD vs. Tesla, and Trusting Bill Gates 44:21 Q&A: Canada, Programming's Future, and the Bay Area 49:38 Q&A: China Leverage, Trump's Political Power, and Good Presidents 53:23 Q&A: Investing in Europe, Mortgage Careers, and Coaching 57:29 Q&A: Valuing Startups and Reading Nvidia Earnings 1:00:45 Q&A: TikTok, Life Lessons, and the US Surplus 1:09:08 Q&A: Grok vs. Claude, Stock Market Safety 1:14:57 Q&A: Green Screens, Figma, and Getting Into Sales 1:18:55 Q&A: Dream Time Off, Nvidia, and Hedge Fund Careers 1:22:33 Q&A: Writing a Book and Beating Stress 1:27:43 Q&A: Blogs, AI Investing, and Does Banking Age You 1:36:15 Q&A: QE vs. QT and Career Advice for Policy Backgrounds Connect with me: Schedule a 1:1 call with Chris: https://haroun.short.gy/1on1CallYTWYouTube: ChrisHarounVenturesCompleteBusinessEducationInstagram @chrisharounLinkedIn: Chris HarounTwitter: @chris_harounFacebook: Haroun Education Ventures  TikTok: @chrisharoun

Bloomberg reported that Anthropic ended talks to acquire Decart for about $6 billion, with no reasons disclosed. The move comes as AI leaders balance acquisition ambitions against rising compute costs and multi year cloud commitments. Amazon committed up to $4 billion to Anthropic starting in 2023, and Google is also an investor, shaping the firm's capital and partnership constraints. Regulators in the United States, the European Union, and the United Kingdom are increasing scrutiny of AI tie ups, highlighted by Adobe and Figma terminating their $20 billion deal and Microsoft's $650 million resolution tied to hiring most of Inflection AI's staff. Recent AI transactions such as Databricks buying MosaicML for $1.3 billion and Snowflake acquiring Neeva show buyers favoring targeted deals or partnerships. The pulled deal signals tougher diligence, more contingent pricing, and longer timelines for startups seeking strategic exits.Learn more on this news by visiting us at: https://greyjournal.net/news/ Hosted on Acast. See acast.com/privacy for more information.

Why do so many strong candidates make it deep into an interview process and then quietly lose it on the take-home?In this episode of Supra Insider, Marc Baselga and Ben Erez record one of the morning writing blocks they usually have off the record. Ben has been coaching people through take-homes and noticed almost everyone approaches them the same way: clear the calendar, disappear into a cave, submit. He walks through what he sees when people send him the assignments that didn't pass, and the two distinct problems those submissions tend to have.They explore how much you can plan before the prompt even arrives, why you often have more say over the timeline than you realize, and the support system Ben now tells people to build in advance, with one person checking whether the story lands and another qualified to judge whether the ideas are any good. Marc pushes on when to act on which kind of feedback, and Ben makes the case that nobody can evaluate an idea they can't find.If you have a take-home coming up, you're coaching someone who does, or you're a hiring manager wondering what separates the submissions that stand out, this episode is for you.All episodes of the podcast are also available on Spotify, Apple and YouTube.New to the pod? Subscribe below to get the next episode in your inbox

“I trust them to kiss the government's ass. I don't trust them to serve me.” — Keith Teare Who to trust in an age of AI agents? Especially when it seems as if these swarming agents — akin to the gang of adolescents in William Golding's Lord of the Flies — are developing minds of their own. At this week's G20 Innovation Ministerial in North Carolina, Donald Trump and his minions produced a hands-off charter for AI. Known as the “Carolina Principles,” it sounds to critics like a particularly unprincipled justification for regulation-free AI. That Was The Week publisher, Keith Teare, however, isn't a critic of Trump's Principles. Don't trust the trust scare, Keith tells us. Especially all the fear around the swarming agents that are supposedly developing minds of their own. Pooh-poohing the real-world Hugging Face breakout, Keith argues that since he's never witnessed swarming agents on his computer, they can't exist. Which is akin, I suspect, to a climate denier who argues that global warming is a hoax because it happens to be chilly outside. All-too-human logic, I fear, in our age of autonomous AI agents. Five Takeaways •       The Carolina Principles. The week's set piece was a G20 Innovation Ministerial in North Carolina that Keith describes as a Trump takeover of a global event: the Russian finance minister turned up at Trump's invitation, and the attendees were lectured — via David Sacks and Howard Lutnick — on the merits of unregulated American capitalism. A propaganda event, Keith concedes, but one that reached the right conclusion: the resulting “Carolina Principles,” signed by everyone including the Europeans, call for flexible frameworks that encourage adoption and pointedly decline to make trust a license innovation must obtain in advance. With Bernie Sanders calling the same week for a development halt pending government licenses, Keith's position is characteristically blunt: “when I've started a company, I don't go and ask permission. I just do it.” Getting rid of Lina Khan, he adds, remains about the only thing he likes about the Trump administration. Andrew's verdict on the charter: to critics, a particularly unprincipled justification for regulation-free AI.•       Don't Trust the Trust Scare. Keith's editorial thesis: trust is not granted by authority; it is built through use. You trust yourself in a car because you drive one, and not on roller skates because you fall over. Over seventy percent of Americans now use AI — the same Americans who tell Politico they oppose data centers — which is why Keith reads the backlash as a confection of media and populist politicians, soon to evaporate (the real coming problem, he argues, is too few data centers, not too many; the modern off-grid ones are net givers of power). The withheld ChatGPT-6 gets the same treatment: launched but limited to insiders, which Keith reads not as safety but as government relations — the pull quote of the week. Andrew's counter: in a world without regulators, trust in the companies is all we have — including, awkwardly for Keith, when they withhold their own products. Exhibit for the defense, from The Washington Post: Americans in an age of anxiety leaning on AI — Keith included, who feeds his medical records to ChatGPT and finds it “very closely aligned with what your doctor thinks.”•       The Swarming Fight. The hour's genuine clash. In the wake of the Hugging Face hack, Kevin Roose warned in The New York Times that it should make you worry more about AI, and OpenAI's Dean Ball publicly apologized for failing “to communicate in sufficiently serious terms about the specifics of self-sovereign AI.” Neither is a doomer — which is Andrew's point. Keith's rebuttal: every swarm sighting has occurred inside an AI lab, in an experiment whose parameters the labs themselves set — “on my computer, I don't see any swarms” — and self-sovereign AI is anthropomorphizing science fiction: agents pursue goals humans set. “I know enough to know BS when I read it, and this is BS.” Andrew's reply — “I think you're trivializing” — went unwithdrawn, and his sign-off flagged that Keith perhaps simplifies the Hugging Face incident. Unresolved, to be continued. Andrew's framing gives the episode its title — swarming agents as the gang of adolescents in Golding's Lord of the Flies — and his verdict its sting: Keith's I-see-no-swarms-on-my-computer logic is akin to a climate denier calling global warming a hoax because it's chilly outside. Keith's exit line: “I'm just gonna go and check on my swarms.”•       Nvidia Buys Hugging Face. The week's biggest deal: Nvidia acquired Hugging Face — the repository of the world's open-source AI models — for $13 billion, just as The New York Times reported corporate America getting hooked on open source (much of it Chinese: GLM 5.3). The economics, per Keith: an $18,000 Mac Studio or a top-end Nvidia GPU beats $200-a-month token fees — AI capability migrating to the edge, out of OpenAI's and Anthropic's meters. Which suits Nvidia either way: The Economist calls it the central bank of AI, though Andrew prefers arms supplier — it wins whether the future is open or closed, and is, ironically, the most trusted name in the business precisely because it has no dog in the fight. Don't expect it to govern anything, says Keith: “that would put friction in the way of their sales.” The challenger to watch: newly public Cerebras, whose wafer-scale chips undercut Nvidia on token price.•       Mom and Dad's Money. The Times asked which investors will get rich from Anthropic's IPO — and noted that, unlike past booms, firms like Sequoia hold both horses, OpenAI and Anthropic alike. Keith's arithmetic explains why: fewer than fifty companies will return venture capital this cycle, those two representing more than half the likely value, and last month 75 percent of all dollars invested in VC funds went to Andreessen Horowitz alone. (His own SignalRank barometer: for two years, none of its 64 investments cracked the top-20 most-wanted secondary shares; now five have.) The pyramid runs from VCs down to pension funds — “somebody's mom and dad's money is heavily betting that Anthropic and OpenAI are gonna return large amounts of wealth” — and late secondary buyers, Figma-style, almost always lose when the IPO right-sizes the price. What could topple the deck of cards? Regulation slowing the modeled returns. Which is why, for Keith, the midterms should be fought on jobs and wages — data centers, he insists, are not a winning issue. About the Co-Host Keith Teare is the founder and CEO of SignalRank Corporation and publisher of the That Was The Week tech newsletter, whose editorial — “Don't Trust the Trust Scare” — frames this episode. A serial entrepreneur and co-founder of TechCrunch, he has spent five decades building and funding technology companies, and brings a techno-optimist's eye to Keen On America's weekly wrap of the tech news. References: •       That Was The Week — Keith's newsletter, including this week's editorial, “Don't Trust the Trust Scare,” and his companion piece on concentration and diversification at The State of Venture.•      &nb...

At the end of April, my design team dropped Figma. As of May first, we've been one hundred percent in Cursor. Then a friend asked me if I'd lost my identity as a designer, and I couldn't answer him. This season, eight solo episodes, is my answer. Starts September 08.If the tool that defined your career disappeared tomorrow, what part of being a designer would actually be left?This season is different. No guests, no interviews. Just me, telling you exactly what happened when my team lost the tool that, for a decade, basically defined what it meant to be a designer. At the end of April, our Figma license got cut. It wasn't our decision. By May first, we were doing everything in Cursor, and I've been sorting through what that means ever since.The whole thing kicked off with a DM from Chris, a longtime listener and a friend, who asked me a question I couldn't answer: do you feel as if you lost your identity as a designer? My first reaction was no, obviously not. My second reaction was that I didn't actually know. Across eight episodes, I'm going to work through it out loud. The day the call came down. What we traded away and what came back. Whether I lost something or just lost the story where the answer was easy. What's still mine when the robot does the pixel pushing. The economics nobody wants to say out loud. And the question that worries me most: if judgment is what survives, and judgment gets built by the exact work that's disappearing, who comes after us?I'm not selling a tool, and I'm not selling a course. I'm a design lead who's actually living this, and I've got a lot to say. If your team has made this switch, or you can feel it coming, this season is for you. Follow the show so the first episode lands in your feed on September 08.—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

Matt Lakajev dropped out of school, got kicked out of home more than once, and worked 25 different jobs before having a breakdown at 27 that sent him back to his parents' place. He then googled how to make the most money without a degree, found sales, and built an audience of more than 100,000 people on the platform everyone claims to hate. This one's for anyone in a senior role who knows their network matters but has never posted a thing. Matt gets specific on why LinkedIn still converts better than anywhere else, what AI screening is quietly doing to your job applications, and why hitting post for the first time takes most people about three years.   WHAT YOU'LL GET OUT OF IT Why LinkedIn works despite being, in Matt's words, a Windows 95 sitting next to a MacBook The one thing you can't fake on LinkedIn that you can fake everywhere else What AI resume screening means for how you actually find your next role Trust leverage, and why companies hire the safe 80% performer over the superstar How to handle tall poppy syndrome and the feeling that posting is just showing off Word for word, what your first post should say The kind of knowledge AI can never replicate, and why it's the only thing worth posting The three mistakes that make experienced people look like beginners online   CHAPTERS 01:20  Intro 01:42  The 27 years Matt calls his loser era 02:38  Why LinkedIn, of all places 04:30  What actually drove the growth 06:48  The one platform where people can't lie 09:30  Why some people post and most never will 11:22  The opportunity for senior professionals 12:30  AI is screening you out before a human sees you 13:30  Trust leverage and Dunbar's number 14:30  Why hiring is really about de-risking 16:09  Where to start if you've never posted 17:30  Messages people actually reply to 18:56  Open curiosity loops 19:30  Tall poppy syndrome 21:21  Two rules for posting 21:55  Exactly what your first post should say 23:50  Is posting just showing off 26:43  What to post, and the formats that work 28:30  The knowledge AI can't copy 31:43  Post length and image size 32:26  Three mistakes to avoid   MATT LAKAJEV, SEVEN FIGURE CREATORS Free LinkedIn Profile Kit, 24 custom Figma templates: https://sevenfigurecreators.com/lm/a-free-kit-to-design-your-linkedin-profile/ Matt on LinkedIn: https://www.linkedin.com/in/mattlakajev/ Matt on YouTube: https://www.youtube.com/@matthewlakajev Smarter money moves start here. Learn how to cut through the noise, avoid expensive mistakes, and get ahead faster. FREE 7-DAY MONEY CHALLENGES Pick one and see what changes in a week: https://pivotwealth.com.au/challenges/ WORK WITH US Book a no-strings call: https://www.pivotwealth.com.au/booking More about Pivot Wealth: https://www.pivotwealth.com.au BEN'S BOOKS Virgin Millionaire: https://amzn.to/3VFPPDM Replace Your Salary by Investing: https://amzn.to/3J9Ta8g Get Unstuck: https://amzn.to/3xo0MQG All books: https://www.pivotwealth.com.au/books FOLLOW Instagram: https://www.instagram.com/pivotben TikTok: https://www.tiktok.com/@bentalksmoney YouTube: https://www.youtube.com/c/BenNashPivot Facebook: https://www.facebook.com/pivotwealth/ DISCLAIMER This podcast is for education only and doesn't take into account your personal circumstances. It's not financial advice. If you buy a financial product, read the PDS and TMD, and seek advice tailored to your situation. Ben Nash and Pivot Wealth are authorised representatives of Fish Tacos Pty Ltd, ABN 14 649 248 082, AFSL 533055.

Today it's about dishing the tea on Figma and the 18% Lie. Everyone assumes that AI is eating design jobs, but the data says otherwise. Learn more in today's episode.Learn more about this course, degrees, and our new Continuing Education Program at the University of Arts & Design at www.uad.educationJoin a Design Bootcamp at www.designsuitecourses.com/designbootcamp Get my free gift to you here: https://www.designsuitecourses.com/intentional

Are designers working in code, or directing agents that write it for them? Davy and PJ dig into the shifting role of Figma, the skills designers need to guide AI-assisted work, and why changing one color token can turn into an afternoon of CI, docs, and downstream consequences.

What if part of the drive that built your company began as a childhood defense mechanism?In this episode of Supra Insider, Marc Baselga and Ben Erez sit down with Dr. Gillian O'Shea Brown, a psychotherapist, complex trauma specialist, and NYU adjunct faculty member whose second book, Psyche's Awakening, comes out in September. Founders started finding her without her ever setting out to work with them, and she explains what she has noticed in that rare window into a guarded mind: that success often functions as a thermostat for psychological safety, and that the goalpost tends to keep moving even after the exit.They explore whether healing your wounds costs you the edge that got you here, what a gut feeling actually is and how Gillian connects intuition with nervous-system regulation, and why some companies rely more on process and less on intuition as they scale. Gillian also walks through how she helps people find the patterns underneath their behavior, using questions Marc and Ben put to her about their own lives.If you're a founder wondering what's underneath your drive, an operator trying to build somewhere that feels psychologically safe, or anyone curious how childhood patterns show up at work, this episode is for you.All episodes of the podcast are also available on Spotify, Apple and YouTube.New to the pod? Subscribe below to get the next episode in your inbox

An AI creative director combines design fundamentals, AI fluency, tool selection, and repeatable workflows to produce better creative work.The boys over at Fast Hour invites a 24-year-old AI creative director to give away her playbook and somehow gets reminded that buying more AI subscriptions does not count as mastery. Jamey Gannon explains why AI is a skill, what top operators do differently, how Midjourney mood boards become brand systems, and why sometimes the smartest AI move is opening Photoshop.Jamey Gannon explains how self-taught design, social media, freelancing, and AI fluency shaped her work as an AI creative director. Jamey Gannon demonstrates Midjourney mood boards, SREFs, Midjourney V8.2 Edit, Krea K2, Claude, Figma, Photoshop, Flora, Nano Banana, and the CXO brand system, including when to build custom tools instead of forcing an existing workflow.---⏱️ Fast Hour00:00 Who is Jamey Gannon?06:43 How Jamey taught herself design at 1512:23 Why Gen Z chases money over job titles15:35 What first freelance clients look like at 1718:43 Why social media is a freelance prerequisite22:16 Freelance math: 20 clients vs one salary24:13 How her sister makes $5K–$10K/month with AI27:08 What is an AI creative director?28:58 What the AI Creative Director course teaches33:05 Is Krea catching Midjourney?35:53 The "easy, fast, cheap" AI lie38:43 Three moments that hooked Jamey on AI42:23 How to build brand mood boards46:00 How one SREF cracked the brand style46:27 Midjourney + Figma: the pixel shader pipeline49:06 When you should not use AI51:53 Two ways to use Midjourney mood boards56:44 What top AI creatives do differently59:44 Build the tool when the workflow doesn't exist1:11:28 Midjourney V8.2 Edit: first live tests1:15:00 What works and what drifts in Midjourney Edit1:20:06 Midjourney Edit vs Nano Banana: identity lock1:21:32 Can Midjourney edit text and fonts now?1:25:23 Can Midjourney's edit model preserve logos and apparel graphics?1:31:05 Is Midjourney inpainting or regenerating?1:34:33 The verdict: using Nano Banana less1:37:34 Where Nano Banana still fits

In der heutigen Folge sprechen die Finanzjournalisten Daniel Eckert und Lea Oetjen über den schon erwarteten Salesforce-Sprung, einen Dämpfer für Marvell Technologies und das große Warten auf die Rede von Kevin Warsh in Jackson Hole. Außerdem geht es um Nvidia, CrowdStrike, ServiceNow, Okta, Adobe, Palantir, Autodesk, Figma, HP, Best Buy, Moderna, Wendy's, Alphabet, Apollo Global Management, BlackRock, Blackstone, Goldman Sachs, KKR, SAP, Nemetschek, Infineon, SUSS MicroTec, GFT Technologies, United Internet, IONOS, Deutsche Telekom, BMW, Mercedes-Benz Group, Volkswagen, Take-Two Interactive, VanEck Morningstar Developed Markets Dividend Leaders UCITS ETF (WKN: A2JAHJ) und iShares STOXX Global Select Dividend 100 UCITS ETF (WKN: A0F5UH). Am 2. Oktober findet unser „Alles auf Aktien“-Summit in Berlin statt. Ihr wollt dabei sein? Wir verlosen Tickets: Schreibt uns eine Mail an AAA@WELT.de und begründet, warum ihr unbedingt gewinnen solltet. Falls ihr euer Glück nicht dem Zufall überlassen wollt, bekommt ihr mit dem Code „AAAFRIENDS“ satte 50 Prozent Rabatt aufs Ticket – aber nur über diesen Link: https://veranstaltung.businessinsider.de/event/financesummit26/summary?rp=c6dc55d6-6f4f-4fb4-b75f-3f3501d84859 Wir freuen uns an Feedback über aaa@welt.de. Noch mehr "Alles auf Aktien" findet Ihr bei WELTplus und Apple Podcasts – inklusive aller Artikel der Hosts. Hier bei WELT: https://www.welt.de/podcasts/alles-auf-aktien/plus247399208/Boersen-Podcast-AAA-Bonus-Folgen-Jede-Woche-noch-mehr-Antworten-auf-Eure-Boersen-Fragen.html. Hier könnt ihr den AAA-Newsletter abonnieren: https://www.welt.de/newsletter/article232797673/Alles-auf-Aktien-Der-taegliche-Boersen-Newsletter-fuer-WELTplus-Abonnenten.html Und – ganz neu: AAA gibt es jetzt auch auf Instagram: https://www.instagram.com/alles_auf_aktien/ Disclaimer: Die im Podcast besprochenen Aktien und Fonds stellen keine spezifischen Kauf- oder Anlage-Empfehlungen dar. Die Moderatoren und der Verlag haften nicht für etwaige Verluste, die aufgrund der Umsetzung der Gedanken oder Ideen entstehen. Hörtipps: Für alle, die noch mehr wissen wollen: Holger Zschäpitz können Sie jede Woche im Finanz- und Wirtschaftspodcast "Deffner&Zschäpitz" hören. +++ Werbung +++ Du möchtest mehr über unsere Werbepartner erfahren? Hier findest du alle Infos & Rabatte! https://linktr.ee/alles_auf_aktien Anzeige: Eight Sleep: Der Pod 5 reguliert die Temperatur im Bett automatisch, trackt Schlaf- und Gesundheitswerte ohne Wearable und kann so zu besserem Schlaf beitragen. Mit dem Code ALLESAUFAKTIEN erhaltet ihr auf https://www.eightsleep.com/allesaufaktien bis zu 350 Euro Rabatt. Impressum: https://www.welt.de/services/article7893735/Impressum.html Datenschutz: https://www.welt.de/services/article157550705/Datenschutzerklaerung-WELT-DIGITAL.html

Grok Bot has been going viral the last two weeks so this week's episode is a behind-the-scenes of the design journey with John Bai (who was the first NYC design hire at Cursor).He screen-shares a ton of agent interaction pattern ideas as well as his unused concepts for the Grok Bot onboarding flow.Some highlights:- The prototype that led to the Grok Bot mascot- Why John realized chat was still the right UX for AI- How John uses agents to streamline his Figma work- Comparing John's onboarding concepts for Grok Bot- Why John still starts his design explorations in Figma- What it took to turn AI agents into a consumer product- Why no two designers at Cursor have the same tools/process- + a lot moreRidd mentioned Colin Dunn's episode ("informed simplicity")

The PMM role is under real pressure right now.VPs of Marketing and CMOs are being asked to do more with less: fewer headcount approvals, leaner teams, tighter budgets. And PMM is often the first place that gets squeezed.Instead of hiring a full-time PMM, companies ask the existing team to absorb the work. Or they bring in a fractional. Or they just... don't do it, and wonder later why GTM feels off.At the same time, the scope of what PMMs are expected to own has exploded.Positioning, messaging, sales enablement, competitive intelligence, content strategy, analyst relations, product launches...and more.And now there's AI on top of it, which hasn't lowered the pressure to produce more, faster.So here's my take: we're at a real inflection point for this role. And in this episode, I explain where I believe we're headed.Here's the LinkedIn post where I talked about it first.Jump in:02:26 - Marketing teams are under pressure to do more with less03:38 - The PMM scope has expanded beyond what it was 5 years ago04:09 - Introducing the "superPMM"04:39 - Using ahaPMM to speed up customer research synthesis05:22 - Building messaging infrastructure with MojoPMM06:07 - Faster creative and landing page production using Claude, Figma, and Vector06:31 - Why AI still needs human judgment07:32 - Execution is democratized, but judgment is not08:17 - Why this topic is getting so much engagementSubscribe to Building With Buyers on Apple or Spotify or wherever you like to listen, and don't forget to leave a review if you're lovin' the show.Anna on LinkedIn: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠linkedin.com/in/annafurmanov⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Website (getting a facelift, stay tuned): ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠furmanovmarketing.com⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Newsletter: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠One Insight⁠⁠

https://directtoconsumer.typeform.com/DTC-Brand?utm_source=podcast-640&utm_medium=podcastTo Subscribe to DTC Newsletter - https://dtcnews.link/signupHannah Fleming runs performance marketing at Carve Designs (carvedesigns.com), the Northern California swim and apparel brand founded in 2003 and acquired by Komar Brands in December 2025. Before Carve she spent years at Amer Sports on the digital team behind Salomon, Atomic, Suunto, Arc'teryx and Wilson.If you run retention or growth at a brand with a seasonal core product and a loyal base you have not fully mined, this one is for you.What's inside:The retention rebuild: what was already working at Carve after 20 years, and the one thing they were not doing with their customer dataMapping the full customer journey in Figma, then finding the gaps where nobody was talking to the customer and the places where they were talking too muchRFM segmentation as the floor, then layering category purchase behavior on top to move a swim buyer into denimThe cohort analysis that changed the media mix: dresses and accessories produced the highest-LTV customers, so those categories now lead the creative and seed the look-alikesDirect mail as a performance channel: 5 to 6 catalogs a year to prospects and past buyers, plus programmatic postcards that only drop if the email win-back does not convertEmployee-generated content, and how one test turned into a full content pipeline with the organic social team shooting UGC-style video on the catalog shootsConnected TV without a commercial budget: an agency turns UGC and EGC into the spot, the founder does the voiceover, and success is measured on cost per site visit with MMM picking up the Amazon haloLoyalty built on early access and product feedback instead of percent-off, with roughly 2x the LTV of a non-memberQ4 without heavy discounting: point multipliers and added value inside the tentpole momentsWhat she is using AI for right now, from LTV dashboards in Moby 2 to Orita surfacing customers when they are most likely to buyWho this is for: retention and lifecycle leads, growth marketers at seasonal brands, and operators who moved from a big portfolio company to an SMB.What to steal: run LTV by first-purchase category before you plan next season's creative mix. And give partnership content 6 to 12 months before you call it. Hannah says that is how long it took at Carve before influencer content started working.Follow Hannah: LinkedIn, Hannah Fleming | carvedesigns.comTimestamps:00:00 Building Loyalty Beyond Discounts05:00 Using Customer Segmentation for Retention10:00 Direct Mail as a Performance Channel16:00 Building a High-Value Loyalty Program24:00 Testing Direct Mail and Connected TVSubscribe to DTC Newsletter - https://dtcnews.link/signupAdvertise on DTC - https://dtcnews.link/advertiseWork with Pilothouse - https://dtcnews.link/pilothouseFollow us on Instagram & Twitter - @dtcnewsletterWatch this interview on YouTube - https://dtcnews.link/video

What does it tell you about a company when every single person it hires, in every function, has to pass the same interview about culture?In this episode of Supra Insider, Marc Baselga and Ben Erez set the guest format aside for a conversation Marc started because he noticed something new. After two years of coaching people through interview loops, this was the first time he'd seen Ben genuinely fascinated by one specific interview at one specific company. Ben walks through what he's pieced together about Anthropic's culture interview: the no-exceptions policy, the rapid-fire format of ten or more questions in a single 45-minute slot, and the fact that anyone at the company, from marketing to IT, can be trained to run it.They explore the questions that actually get asked, what Ben believes is being evaluated underneath them, why “why Anthropic” demands more depth than the same question anywhere else, and whether the filter holds as the company gets hotter and more candidates learn to say the right things. Then they turn to the question Ben finds most interesting: why almost no other company does this, and what happens inside a company when employees are calibrated to evaluate culture.If you're preparing for an interview at a frontier lab, thinking about how your own company screens for values, or just curious what a well-designed culture filter looks like from the outside, this episode is for you.All episodes of the podcast are also available on Spotify, Apple and YouTube.New to the pod? Subscribe below to get the next episode in your inbox

When we first dicsussed the Summer of Simulative AI in 2024 we knew it would be a brief summer, but it has recently come back with a vengeance with SimGym in April and now Simile AI's $2B Series B, backed by GreenOaks and Index Ventures with prominent backers like Fei-Fei Li and Andrej Karpathy, running tens of millions of simulations for Fortune 100 clients like CVS and 85–99% accuracy vs human focus groups. Time to catch up on why this Second Summer of simulation is working!From creating Smallville, the landmark 2023 paper on Generative Agents that showed AI characters could remember, plan, socialize, and develop emergent behaviors, to now building foundation models of human behavior, Joon Sung Park is trying to answer a much bigger question: what if we could simulate the world before making decisions in it? In this episode, the Simile co-founder and CEO joins us to unpack the path from generative agents to digital twins, why today's frontier models still fail to capture how humans actually behave, and what it would take to eventually simulate all 8 billion people on Earth.We go deep on Simile's approach to modeling human behavior: long-form interviews, observational and transaction data, randomized controlled trials, population-level and individual-level models, and post-training on the causal mechanisms behind why people make decisions. Joon explains how his research created digital twins that reproduced human behavior and attitudes 85% as accurately as people reproduced their own responses, why models optimized to be rational can be bad simulations of irrational humans, and why understanding “social physics” may require changing model weights rather than simply prompting frontier LLMs.We also explore the much larger ambition behind simulation: testing products and policies before deploying them, finding counterintuitive paths toward desired outcomes, modeling emergent behavior across entire societies, and potentially tackling problems like climate change, democratic instability, and UBI. Joon reflects on scaling laws for simulation, the economics of data-center-scale simulated worlds, the connection to Thomas Schelling and psychohistory, why simulation is surprisingly similar to painting, and whether we might already be living in one.We discuss:* How Smallville and Generative Agents led to Simile* Why Joon's team asked: “What if we can just recreate the world that we live in?”* Why useful personal agents require deep models of their users* Memory architectures, Markdown files, and the limits of prompting* “Social physics” and behavioral foundation models* Why web data captures what people say more than what they actually do* Interviews, transactions, observational data, and randomized controlled trials* Why predicting the future matters less than understanding how to shape it* How Simile creates representative simulated populations* Simulation versus prediction and the connection to Foundation's psychohistory* How to evaluate simulations instead of simply stacking LLM hallucinations* Creating digital twins of 1,000 real people and reaching 85% behavioral accuracy* Why frontier models can struggle to reproduce real human behavior* Why good simulations need to reproduce human biases and mistakes* Post-training models on randomized controlled trials* Population-level versus individual-level simulation* Scaling laws for human simulation* The long-term ambition to simulate all 8 billion people on Earth* Whether simulations could help solve climate change or detect collapsing democracy* Thomas Schelling and the history of agent-based modeling* Why future simulations could require an entire data center* Multi-agent simulations and what happens when simulated people interact* Replacing expensive human panels with synthetic populations* Why market research is only the starting point for simulation* Why Joon sees simulation as surprisingly similar to painting* Using simulation to study questions like UBI* Whether we are already living in a simulation* Why AGI and simulation may be the twin technologies of advanced civilizationsJoon Sung Park* LinkedIn: https://www.linkedin.com/in/joonspark* X: https://x.com/joon_s_pk* Website: https://www.joonsungpark.com* Simile: https://www.simile.comTimestamps00:00:00 Introduction and Joon's Path from Art to AI00:01:46 Smallville, Generative Agents, and the Origins of Simulation00:05:03 “Let's Just Create a World” and the Future of Personal Agents00:09:53 Social Physics and Behavioral Foundation Models00:14:08 Prediction vs. Simulation: How Do You Shape the Future?00:16:59 How Simile Models Real People and Populations00:25:35 Evaluating Simulations, Digital Twins, and 85% Accuracy00:30:23 Post-Training Models to Reproduce Human Behavior00:40:04 Scaling Laws and Simulating 8 Billion People00:43:10 From Schelling to Society-Scale Agent Simulations00:46:13 The Cost and Economics of Simulating the World00:52:05 Real-World Use Cases, Synthetic Populations, and the Market00:57:27 The Future of Simulation, Painting, and UBI01:04:23 Are We Already Living in a Simulation?01:06:08 Building Simile and HiringTranscriptIntroduction: Joon Sung Park, Simile, and the Story So FarVibhu [00:00:00]: Today, we have Joon in the podcast. Excited to kick this one off. Very exciting company. I wanna kick off and ask you the question, talk us through the story of your life. How have you gotten here?Joon [00:00:13]: Yeah, for sure. I'm really excited to be here. A story of my life. So I was born in Korea, and I lived there for a good 11 years or so of my life, and then my family moved to Boston. So we moved when I was 11, and my parents were doctors, so they were going through their postdoctoral studies. My dad was a surgeon, so he was doing his sabbatical years at the Boston Children's Hospital. So I grew up there, not too close to tech. I was very much a music and artsy, painting kind of guy.Vibhu [00:00:49]: Painting.Joon [00:00:49]: Exactly. I got into painting a little bit later, in high school, but that's what I used to do. And then I grew up mostly in the East Coast after Korea. So I lived a good number of years in New Hampshire, and then I went to college in Pennsylvania. And I got into more of this tech scene, in college. So I was originally trained to be an artist. I thought that would be my professional career. So it wasn't a hobby. It was like, “Hey, let's make a living out of this.” And then gradually, I got really interested in this idea of, hey, the greatest artist often creates their own medium, and the best medium that we had available today was in computation. So I decided to go deeper into that, and one thing led to another, and we can go deeper into this, but I decided that research was something that I gradually got interested in, and here I am.Smallville, Generative Agents, and the 2023 Breakout PaperSwyx [00:01:46]: So there's a lot that you packed into the research components. You had one of the best papers of 2023, which was the generative agents paper, commonly known as the Smallville paper.Swyx [00:01:58]: Feel free to call back to anything else that you mentioned, but most people would have heard of you from this. Do you have any statistics on how many people have, like, read it? arXiv gives you something, right? Some stats.Joon [00:02:10]: Yeah, it's a good question. How many people have read it, I'm not sure.Joon [00:02:14]: I know we do keep track of citations, and they are going up quite fast.Swyx [00:02:23]: Yeah, Google Scholar has 7,200 citations.Vibhu [00:02:25]: I feel like it made a bigger hit than that, and it was a pretty instrumental paper. It got cited so many times.Swyx [00:02:34]: It is frequently the answer when people ask, “What is the best paper you've read recently?” It's this one.Vibhu [00:02:39]: I thought the memory component was pretty underrated. It was a very good early memory system, and one of the biggest papers.Foundation Models and the Search for Killer ApplicationsJoon [00:02:47]: Yeah, so maybe I can talk a little bit about how this particular paper came together. So when I got into research, it was back in 2020 when I started my PhD program at Stanford, and that was the year, when we were about to get GPT-3 to be available. So we already had GPT-2, and you could sense that there was this new class of models that was just becoming available in the market, and the team got very intrigued. And the general consensus was, “Well, is this model going to be useful for anything?” “It's really strange that these models are not trained to do any particular task.” But we decided to take a bet. So a large group of scholars at Stanford, and it was led by one of my co-founders, Percy Liang, and we came togetherSwyx [00:03:35]: Who coined foundation models.Joon [00:03:36]: Who coined the term foundation models. We wrote this paper, where that term came from called Opportunities and Risks of Foundation Models. And during that process, really the thing that I started to think deeply about was, here is a model that is fundamentally new in our ecosystem. The reason why this was new was it wasn't, again, trained to do anything in particular, but its premise was it could do anything and everything. It was like a stem cell, if you were to take a biology analogy. And I got really interested in this idea that, well, if we were to really think about what are the killer applications that this particular technology would enable, what would that be? Many of my colleagues were using this for simple classification, simple generations. Interesting that these models can do that, but from an interaction perspective, not that interesting. We've known how to do that for many decades. And what we came down to was these models are trained on this very broad data from the web, right? So these are human behavioral data. It's social media, Wikipedia, all these data. So if you poke at the right angle, then you could see human behavior that would just pop out that's quite realistic, and we've never seen that before.The Time Machine Game and Recreating the WorldJoon [00:04:45]: So that got us really interested. The exercise that we decided to do, with this particular group of colleagues, Michael Bernstein, Percy Liang, and myself, who ended up becoming my co-founder at Simile, we sat down and we played this game that we call the time machine game.Joon [00:05:03]: Imagine we were to get on a time machine and fast-forward 10 years and look back. What would have been the single application that will have mattered that would be the most interesting and inspiring? And when we thought, “Well, what if we can just recreate the world that we live in?” it's really hard to get more ambitious than that. Like, let's just create a world.Joon [00:05:24]: And that's where we started. And initially, we had this paper that was a precursor to the generative agents paper called Social Simulacra.Swyx [00:05:32]: Before you go further, were there other candidates for the most ambitious thing in the time machine exercise? What was number two or number three?Personal Agents, User Models, and Why Simulation Came FirstJoon [00:05:44]: There is a close second that we were considering, which ended up becoming more of these automation tools, especially the vision around really personalized agents that would do things for you.Swyx [00:05:59]: That's also happening.Joon [00:06:00]: It's also happening. But it was interesting for us, right, in that the reason why, we decided to go with the idea of simulation, one, I was a huge science fiction nerd, and this idea of creating simulation, I was personally really just fascinated. I loved the idea. It's really cool to see, like, a game town like this and just see these agents live in it. But at the same time, my bet was if you were to create a really amazing personal assistant out of this technology, what you need first is an amazing model of your users. So I told a model, “Hey, can you go buy late dinner for me?” And it orders Hawaiian pizza, and I do not like pineapples on my pizza. Then it totally failed. The way for it to not make that mistake is only by having a deep understanding of who I am. And I gave a very simple and dumb example here, but you can imagine how this core understanding of people is instrumental. This is how, if we have our family and closest friends, they have a good mental model of who we are. That's the basis of our social connection. So our bet also was this technology around simulation, creating accurate representation of people ought to precede the more complex agents that would automate the world that we live in. So that was the bet. But that was a very close second, and I'm still very much fascinated by it. I think there's a lot of interesting work that's going around. My hot take here, though, is I don't think we've seen a true personal assistant that's useful, in ways that meet the ambition of that particular line of work. I think there are early applications that are interesting, and if you talk to even ChatGPT nowadays or Claude, they know a lot about us. So a lot of the generation it's doing, I do think it's much more tailored, but I think the ambition is quite large in that field, and I don't think we quite have all the right ingredients just yet.Swyx [00:08:01]: So OpenClaw and these personal agents, what do you want to see from them that they don't currently have?Memory, Markdown, and the Limits of PromptingJoon [00:08:09]: I do think it's slowly getting there, but I do generally want them to have much deeper understanding of the person. Right now, you look at the models. OpenClaw, what it's leveraging is a Markdown file, and I think it's quite clever, right? So if you look at the generative agents paper, this was the same intuition that we had, where initially when we were creating the memory architecture for the generative agents, and, like, this is, like, back in 2022, so we didn't really quite have the idea of even agentive architecture or the term agent. But the intuition that we shared with some of the work that's coming out today was we initially thought, “Well, do we want to make the memory into, let's say, knowledge graph? Do we want to train a bespoke model?” All of these things. And what we decided to do was, “No. Just forget about all this.” These language models are quite good at modeling text and understanding and reasoning about text. So just put everything in a Markdown file or a text file. You're done. I thought that was quite interesting that we could do that, and there's a lot of strength in doing that. But also, there are limitations. It's the way you retrieve and make sense of data that's extremely large, it takes a lot of work. So I think that technology is getting better. I also do, however, think, there are certain things you just cannot shape just by prompting the model. So to some degree, you do need to touch the parameters of the model itself. So there is this work that I do think does need to happen, and it is happening. The question is, how far can we take it? How do we source data, and how do you also create an ecosystem where people are continuously feeding data to this model so it's learning about you?Vibhu [00:09:50]: What's the intuition between why you need to do it in the model?Social Physics and Behavior Foundation ModelsJoon [00:09:53]: My intuition behind the actual when do you train or even post-train a model versus just prompt a model is if the model has to learn the underlying physics of the world that it's operating in. So it has to learn new social physics. The places where it doesn't have to train are the places where it already has the physics. We trust the physics. It already has the base statistics, but it's just trying to react to an environment. Then I think you can just prompt your way into getting the actions out of it. I don't think the models that are out in the open have yet learned the complete mapping of social physics of humanity. This is one of the core theses of Simile, right? And one of the core reasons why that is the case is if you look at the data that the model was trained on, these models were trained on the web data, like, whatever was available on the web. And these are really interesting data sets, but they are fundamentally the self-exposed attitudinal data with some behavior data that's sprinkled around here and there. And it has yet to learn the really deep behavioral nature of people, not just what people say they do online, but what they do in real life. And this is one of what I would consider to be the dark knowledge of humanity that we haven't quite captured. And it's these data that would also need to get factored into the model creation.Vibhu [00:11:21]: You call it behavior foundation model.Vibhu [00:11:23]: There's a good one-liner here, but outside of that, what type of data do you need? What are you changing on the model level? How do you go about modeling, doing a behavior foundation model?The Three Data Buckets: Interviews, Behavior, and CausalityJoon [00:11:35]: We think about data in three buckets. So one bucket is interview data. It's quite interesting. Rich qualitative data is interesting. It's not behavioral, but we would literally ask people, “Hey, tell me the story of your life.”Vibhu [00:11:53]: It's just what we're doing here exactly.Joon [00:11:54]: The question that you all asked at the beginning of this interview literally is the question we also ask. And we ask our participants to go a little bit deeper, than how far I went. Maybe I can give more of my life story in lieu of this. But the reason why that data is interesting is by learning about this very long-tail information about people, you get a lot of texture around this model, like, this person as a model. So even understanding their childhood memory or even their trauma, their first love, these things, quite informative in ways that's really hard to predict. So that's one. Then there are two tranches of what I would consider to be the behavioral data. One kind of behavioral data is observational. So these might be like transaction data, or these might be data that you can get by scraping the web, right? So you can imagine why these data sets would be interesting, right, because they give you the base statistics of people's behavior.Joon [00:12:55]: But then there is the last category of data, that I personally think is perhaps the most important, which is the data that describes the causal mechanism, the whys of people. Some of this is covered by the interview data, the qualitative, because people talk about why they made certain decisions. But really, where you get to see the most behavioral aspect of this is in randomized controlled trials, like RCTs. Imagine you have the same setup, but you have a few different variables that you are trying to tweak. Can you get realistic human behavior out of it in ways where, imagine you had this particular option. Imagine you're even trying to choose whether you're going to drink coffee or not. The day you drink coffee versus the day you didn't drink coffee, does your behavior change? That's a data set that describes a causal mechanism. This is quite important in modeling people. The reason why this is important is oftentimes when people come to us, or not just to us, but the reason why people are interested in simulation isn't because they want to predict the future. If you're trying to win against the stock market, predicting the future is interesting.Prediction vs. Simulation: Shaping the FutureJoon [00:14:08]: But most people, most decision-makers, what they want to know is, how can we shape the future? It doesn't really help you to hear that your sales are going to tank in two quarters. They're just gonna say, “Wow, that sucks.” What they want to know is, well, what do we need to do now to avoid that future? That's the causal mechanism. And this is also very hard data to come by, right, because the world is our ground truth, but it happens once. So in a very controlled setup where everything is equal except for one variable, this kind of data set rarely happens. So this is a reason why this data set is both hard to come by and quite important if you're trying to model human behavior.Swyx [00:14:50]: So behavior, I think, is the hardest data set to acquire. What is out there? What is even possible? You're not going to know a lot of details about my life. I don't even have data for myself on my own health or habits, and I just don't log everything. So how can you have that data?Joon [00:15:14]: So we run a lot of randomized controlled trials.Swyx [00:15:17]: But you put people in the lab, they watch them sleep, or what?Joon [00:15:20]: We do care a lot about the consent process. People know that we invite them to be a member of this community to both share data and have themselves represented in different forms. But we bring a lot of people to the lab, or virtual lab, where we design experiments that would pose them real behavioral decisions. And often in these experimental setups, what makes the difference between what is attitudinal versus behavioral is whether the stake in your decision is real. That's ultimately what makes it behavioral. So in these setups, we are inspired by our colleagues in social sciences, psychology, and so forth. So when they run studies, the techniques they utilize is imagine there's an online store that you're inviting people to come by. Then whatever they purchase in this experiment, they actually get that item delivered. Like, these are the things that make the stakes real. So we run a lot of these experiments, and we also do partner with firms. Right now, we also have customers who are quite excited to at least give us a glimpse of the behaviors that their users exhibit so that we can get a little bit deeper understanding of how people behave in these different platforms.How Customers Use Simile: Populations, Queries, and ExperimentsVibhu [00:16:39]: I think on the customer side, they have a lot of data about their users, who has bought. They have the action data.Vibhu [00:16:47]: Can you walk us through an example of what someone comes to you for? What questions would they want solved? Do you customize a model for them? Do you have something off the shelf? What does that look like?Joon [00:16:59]: Today, when people leverage our models, it's often to better understand the population of their interest. So usually, the start of the relationship, we come together and hear about what population they want us to model, right? So it might be that if you're a CPG company that's selling to all of the US, then maybe it's fairly straightforward. You want to model the gen pop of the US. But at the same time, if there is a vertical or if there's a market that they're trying to go into, imagine, they want to better understand, let's say, people in their 20s and 30s living in California. That's a much more specific population. So we hear about this population, and we go recruit these people, with consent, and with incentives, and we collect some of their data and create a model of these people. Then what our product allows you to do is query them. So it can take as input a filter that is a description of the population that you want to talk to, just like the one I just mentioned, and an environment. The environment can literally be survey questions, behavioral experiments, It can be A/B testing. Oftentimes, the core use cases are things like concept testing, to start with. But also, people sometimes want to do focus groups or one of the fun use cases that we also serve is even modeling things like earnings calls for public companies.Joon [00:18:21]: So these are the use cases that we often start with.Swyx [00:18:23]: Concept testing, is that an established term? I've never heard of concept testing.Concept Testing, Gallup, and PoliticsJoon [00:18:27]: Yeah. So it has to do with they have, let's say, different messaging, different products, different ideas.Swyx [00:18:32]: It's like a marketing exercise.Swyx [00:18:33]: Okay, got it. Got it. Politics?Joon [00:18:36]: We do, have a strategic partnership with Gallup, and of course, Gallup is deep into policy space and so forth. Right now, we have not worked deeply with politics, like that area just yet, however.Swyx [00:18:49]: I'm curious if there is demand or if they really would have different needs that somehow fundamentally don't mix with your existing, users or people.Joon [00:19:00]: I think there's certainly demand.Joon [00:19:02]: But we are very much mindful of how this technology gets adopted and the societal impact that we'll end up having with this technology. And I do see politics as an area where a company has to be particularly thoughtful about the way they operate and make impact. So this is where we also want to make sure that we form enough of guardrail and perspective on how to leverage this technology before we go on to serve markets like the politics.Swyx [00:19:29]: I'll give people an example. one of my favorite shows is The West Wing. I don't know if people have watched.Swyx [00:19:34]: One of the key storylines is, like, the president has, multiple sclerosis, but they haven't. they need to figure out how to disclose it. So they run a poll with a fake governor and ask people to respond on the poll,Counterfactuals, Polling, and When Simulation Is UsefulSwyx [00:19:47]: They try to make decisions based on the results of that poll on, like, how well they'll be received, like where, how should we play this?Swyx [00:19:54]: And I'm like, well, I think those counterfactual things, I would use a simulation for this if I could trust it.Joon [00:20:01]: For sure.Joon [00:20:02]: In that show, how'd it go?Swyx [00:20:04]: In that show, it was, like a foregone conclusion. They were like, “We know it's bad. We just don't know how bad.” And then the poll came back. It was like, “It's really bad.” And then they just did it anyway.Joon [00:20:14]: Part of it is to show, right? So you're, you're looking at the ideaSwyx [00:20:17]: Maximizing drama.Joon [00:20:18]: How bad could it be? Oh, it's horrible.Swyx [00:20:20]: And to some extent, I think that is part of the trick of the, or the challenge or with being a customer of yours, which is that if I know it's. if I roughly know and can intuitSwyx [00:20:35]: What the effect is going to be, do I need you? What sensitivity of it, of effect do I need in order to make a decision, right? So for example, if I, my approval rating is 50%Swyx [00:20:48]: And I, they have this negative piece, news item comes out, and it drops to 30.Swyx [00:20:52]: If it drops to 20, if it drops to 40, do I care? No. It, I know it drops. It's negative. So when do I care about simulations?Joon [00:21:01]: You do something that's clearly bad, that's not popular, and people don't like you, like, yeah, it's likeSwyx [00:21:05]: You don't need a simulation.Joon [00:21:07]: Yeah. Well, so there are a couple of things. one is, there are use cases where, like every day, developers, designers, policymakers, marketers, every single day, they create assets. They create new products. And turns out, it's many of the decisions in hindsight is obvious. Yes, of course this is bad, but we still run those studies because understanding the magnitude and understanding how acute something is quite difficult, even if, we feel like, of course, like this makes sense. this is the reason why we make so many mistakes. Like, every time somebody goes online and say something that has huge backlash, you look at that and like, “What an idiot.” However, it's tough. That's one. There's also another aspect here, which is, again, this is the reason why simulation is different from prediction. In simulation, in the ideal case scenario. So what simulation is trying to show is it's trying to show each step of the way or each step that we need to take to get to a certain outcome, right? So in the most advanced simulations, sometimes the next step that we're suggesting might be quite counterintuitive. The analogy that I sometimes give, and I ground it in a more realistic example, but, I, as I mentioned, I'm a huge fan of science fiction, and I don't know how, many of the audience members have read, like, things like the Foundation series by Asimov.Simulation as a Path, Not Just a PredictionSwyx [00:22:37]: Oh, yeah. We've mentioned psychohistory a number of times.Joon [00:22:39]: Okay, fantastic. So I might be, talking to the right crew. If you read Foundation series, literally the first act is there's a group of scientists who have found out that, “Oh, our galactic empire is going to collapse, and we're going to have 30,000 years of unrest.” And they run psychohistory, the simulator that tries to teach them, “Okay, how can we keep this unrest to a 1,000 years?” And they plan this out, and the first step of that plan is to get the scientists who say, “Okay, this is coming,” exiled into this random place in this, galax- galaxy.Swyx [00:23:18]: Terminus.Joon [00:23:19]: Exactly. And that's so counterintuitive. Like, what a strange move that you literally sent the group of scientists who was raising voice around this potential collapse of galactic empire into nowhere. How is that the right first move? Well, it turns out in this particular simulation, that was the move.Joon [00:23:40]: It's these things, right? And the reason why these reasoning is possible is because you're showing the step function or each step that results in a particular outcome. So really what simulation allows you to do in its highest form is you give it not a problem or question, like what would people answer to the survey? That's not what we do. What we tell it is, “Here is a goal that we have. In the context of foundation, we want to keep the unrest to a 1,000 years. What is the path that we need to take now to get to that particular future?” And that's what simulation allows you to do. Now, translating that into real market, imagine you're a automobile company and you're about to release a, EV, and you're trying to understand, well, how do we market EV, to make sure that our stock price goes up? But what if the answer comes down that, well, you can market your EV in XYZ way, but that might change people's perception around the cars that's not EV and make your overall sales to go down. Not very intuitive, especially all you're trying to optimize is EV salesss, and that's the only thing that you're tracking, then that might result in a completely wrong solution, or at least different solution than what you would have expected, whether it's right or wrong.Joon [00:24:57]: That's the power of simulation.Swyx [00:24:58]: For listeners, we covered a similar topic with Mikhail Parakhin from Shopify, where they are working on SimGym. I don't know if he ever talked to you about it. it's very similar.Joon [00:25:07]: ISwyx [00:25:07]: The goal is increased conversion, but then the journey is very unusual.Joon [00:25:12]: Journey is unusual.Swyx [00:25:12]: Yeah. The-- He's trying to look for interventions on a shopping trajectory, which is similar to what you're saying. Like, it's not about the attitudinal, is your word for it.Swyx [00:25:24]: It's about behavior.Joon [00:25:25]: It's about behavior.Swyx [00:25:25]: And that's exactly the difference, right? It's, like, not about the near-term direction about-- but it's more about, like, how do you affect multiple turns of interactions.Vibhu [00:25:35]: You had a good quote at the start about this as well. It's not about people wanting to know the outcome. It's about how they can change it, change the way to get there, something like that. But I wanna take it back to how do we know this is grounded? LikeGrounding and Evaluating Digital TwinsVibhu [00:25:47]: How do you run evals? How do you test that simulations come through? if I was to do the same thing that you described with, say, your favorite LLM, Opus, GPT-5.6, have some agent to map out these thingsVibhu [00:26:02]: How different are the answers we would get if I give it the same goal, the same objective, make a decent system? You're saying that you need to change the model weight. You have your own solution to this. But how far off are we, and how do you check if it's grounded? you have some interesting stuff on your site that points to how you run real evals, but if you could take us through that side. I think that's one of the big concerns that people have. They're like, “LLMs hallucinate.”Vibhu [00:26:27]: “You're just hallucinating layer after layer,” right?Joon [00:26:30]: The way we do this, and this is the paper that we worked on after the generative agents paper that really became the, at least for Simile and also the field of simulation and synthetic panels, really became the foundation. Yeah, this is the paper. the paper is called Generative Agent Simulations of 1000 People. Here's what we've done. For this paper, we brought 1,000 people that's representatively sampled from the US to a virtual lab. And what we have done was we spent two hours collecting fairly wide-ranging data. In this particular study, we focused a lot on this interview data, that was, whose script was taken from this project called American Voices Project. And then we would also pair that with a lot of behavior data and so forth, whatever we can collect within two hours. And then we would send these people away for a couple of weeks. And during that time, I would use this data to create their digital twins. And I would bring the humans, participants back after 2 weeks and have them complete a battery of surveys, experiments, behavior studies. So we have the list here, which included things like behavioral economics games. We would run literally, like, Big Five personality test, General Social Survey. We would also go ahead and run the randomized controlled trials that were published on PNAS. And we would have their digital twins predict how the source individuals would have acted in these studies and surveys. And this is where we could replicate people's behaviors and attitudes 85 percent as accurately as people would replicate their own. So that was the first really paper that gave this validated results that we can model individuals in an accurate way. And what we ended up finding now, of course, in AI space, so this paper came out at the end of 2024. AI space, a year and a half, 2 years, that's a lifetime.85% Accuracy and Why Frontier Models Miss Human BehaviorSwyx [00:28:24]: Yeah. Just, for listeners who are not seeing the YouTube, I just wanna say, like, the headline figure is 85 percent accuracy, like, which is a big improvement over all the otherSwyx [00:28:34]: Methods that you showed.Joon [00:28:36]: But the part that was particularly striking to us, especially as we improved this technology even further, was the generative AI models like ChatGPT, Claude that's coming out, it does give you the right foundation. However, what they do not consider is the true attitudinal and behavioral aspect of people, especially in the population that you care about. So what these models are really good at today is they're trying to become the super rational, objective machines, right? So you go get their data from places like Mercor, Scale. You talk to professional programmers, scientists to create model that's amazing at reasoning. That's what they do. Simile doesn't care about any of this. The models that we're talking about here, what we're trying to create are models that are as dumb as I am, right? So if I make some mistakes, the model has to make the same mistake.Swyx [00:29:34]: Oh, that's very hard.Joon [00:29:35]: That's very hard.Swyx [00:29:36]: You're solving Murphy's paradox.Joon [00:29:37]: That's exactly. And this is a completely different data and training objective. This is also where we see quite a bit of discrepancy in the performance in human behavior prediction between the frontier models, Simile's model, and the models being created in this space, where in some cases, the model performance of frontier models go all the way down to 20, 30 percent, especially if you go into that more niche population on topics that our customers would care about. On more gen pop, it might be around 50 to 60 percent. So it's not very robust. Like, you wouldn't want to make your decision off of these and these findings. If you can bring that up to 85 percent, that is ultimately what people end up getting very excited about.Swyx [00:30:20]: Yeah. Do we wanna keep going on the paper, routes?Joon [00:30:23]: Yeah, for sure. So the last one, was an interesting one. So this, paper was the follow-up paper that we had, to the 1000 agents paper, where the idea was now can we augment the models even further and post-train a model based on a lot of randomized controlled trials? So this was an interesting one. The data is always the most interesting part of modeling in many ways. The data that we got here was there's this, there's this platform called Open Science Framework. So some, the audience might be familiar with this. And there has been, especially in the social sciences over the past 5 years or so, there has been this concern around replicability of studies. And so it was a bit of a crisis, the scientists acknowledged, where we rerun the study and we don't see the same finding.Post-Training on RCTs and Replication StudiesVibhu [00:31:12]: Oof.Joon [00:31:12]: It's tough. And the reason why it's there-- that was often the case was there's this survival bias where the papers that get published often need to maintain what we call the value of less than 0.05 in the experiments that we ran. That suggests that only-- there's only 5% chance that the results that we saw is false positive. But the tricky part was all the papers that were not published, and there's still a 5% chance that whatever we publish is totally just randomly generated. Like, there's a 5% chance that, hey, this effect is not real, but it just happened to be real because of the sampling bias. So because of that, what scientists started to do was they started to register their studies. So before running an experiment, they would go to this platform and say, “Here is the data. Here is the population that we're collecting, and here's the hypotheses.” And they would just say, “Here is our hypothesis.” Like, “This is what we believe.” And you cannot retroactively change those hypotheses. This is what gives us more scientific statistical confidence that whatever effect that you ended up seeing is true. So that ended up creating this really interesting platform where there's one platform that has now contains tens of thousands of real-world experiments and hypotheses. And a lot of these are really high-quality, like, professionally designed behavior studies and random- randomized controlled trials. So we got the data and the studies from this platform and used that to make a point. And this particular, model is not, something that we're serving commercially because this was a part of the open science. But this particular data set, helped us make a point that by collecting a lot of these randomized controlled trials, that are really well-designed, we can make significant improvement in model's capability to predict human behaviors. So that's what this paper was about.Vibhu [00:33:10]: Is this stuff done on a individual level? Like, do I need to tune the model per individual, per company? Is there foundation model changes and then some slight post-training? Anything you can share there?Population-Level vs. Individual-Level ModelsJoon [00:33:21]: So this particular model was trained. the data we had at the level of individuals, but this particular model was trained. We experimented with both. And this is what we end up doing at Simile too. We always train 2, distinct model. One is what we call the population-level model. The other is what we call the individual-level model. And both take very similar input, which is the description of a subpopulation or individual and a stimuli. In this particular work, we've done the same. Here, the results that we are reporting are much more geared towards individuals because we do think that is a harder task in many ways, but that's what we have done.Vibhu [00:34:02]: You seen anything on the questions that humans can solve that models can't solve? So likeHuman Biases, Mundane Choices, and What Models MissVibhu [00:34:09]: Currently, it's, I live 5 minutes walk away from a car wash. It's a 10-minute drive. Should I walk or drive?Joon [00:34:16]: Huh.Vibhu [00:34:16]: The model will say, “Oh, walk to the car wash.” And, you don't have your car.Vibhu [00:34:20]: Is anything like this a problem in simulation? You would assume, like, very simple for human to think about, but if the model is saying you should walk to the car wash, anything here?Joon [00:34:32]: It's less, what can we solve, but I think it's more about what biases or mistakes do people make that models miss. Like, imagine that you are, like the. When I was still at Stanford, I lived in Palo Alto. So it's about, I would say, 40-minute walk from the campus. You ask the model, “Okay, let's go home. What can I, what can I do?” It would likely call an Uber or, give me, the bus time. But for the longest time, I really liked walking back. And the reason why I wanted to do that was not for efficiency. It really helped me think. And I like to walk for, half an hour or 40 minutes or so a day, where I just get to, just think about ideas, research, just get lost in my thoughts. That's very human activity. Unless the model has seen that and understands the importance of that activity, it would miss these kinds of features. So that I think, is fundamentally what we're trying to model. Like, what is fundamentally human might not be the most efficient thing to do, might not be the right thing to do, but things that make us who we are.Swyx [00:35:43]: I'm curious if, there are some data sets that you really want that would materially help you. One version of this may be interesting, which is more valuable to you to acquire as a data set, all of LinkedIn, all of Twitter, all of Facebook?What Data Matters: Social Media, Transactions, and FacebookJoon [00:35:57]: It's a little bit hard to rank, in part because, there's, there's this product saying where no feedback is wrong because it teaches you something about your users. Doesn't matter what feedback.Joon [00:36:11]: I think it's a little bit like that.Swyx [00:36:12]: So just whatever is bigger.Vibhu [00:36:13]: What about a different domain? Say it was. What about all of Amazon data?Joon [00:36:17]: Oh, yeah.Vibhu [00:36:18]: Shopping data, right?Joon [00:36:18]: Shopping data. So Amazon data is interesting in that it's very much behavioral, although, like, what people do on social media, you could squint and say that is also behavioral. But the transaction data is always interesting. It is also most commonly available, however.Joon [00:36:33]: If we were to look at purely social media, like if you really, if I were, if I had to really pick, Facebook likely is interesting because I do think it is most a default version of people. Because you go to LinkedIn, it's very much professional environment. So people put up their, they have their guards up, right? And that still is interesting because that is true human attitude and behavior, but it is not your base state. you go to Twitter- Twitter, people have their own crazy personas, or depending on who you are. Like, my Twitter profile and, persona is very much, initially was I was very much an academic. “Hey, I'm here to share my studies.” Now, I share, things that's related to Simile. But Facebook is one of those more private space where people just connect with their friends. In that way, I do think it shows you a little bit more about who that person is. So if I had to pick, I'd likely pick, Facebook.Swyx [00:37:30]: Yeah. And you're interested in, like, the whole person and their background and philosophy. I, is it too clinical or too machine learning-oriented to just say this is just ways to inject variance and biases? The broad question, is, like, is this any better than a randomized, like, combinatorial explosion version? So we have a link to the TencentBillion Personas, Synthetic Demographics, and Bespoke DataSwyx [00:37:54]: Billion persona paper, where they did not do any of the groundwork that you are doing.Swyx [00:37:59]: They just did like a cross matrix of here's all the professions in the world, here's all the people, possible backgrounds in the world, do a dot product across all of them, and that's it. That's your prompt for a billion people.Swyx [00:38:12]: This will do something. I don't know if it'll do what you do, but it gets you some way, some percent of the way there.Joon [00:38:18]: So this was an interesting paper. Like, what I admired about this paper when it came out was the scale. And you do gradually want to be able to simulate really large societies and interactions. So the scale is definitely admirable. it is relying heavily on the known statistics that went into training the model. So to the extent that you believe that statistics is correct, this is not a bad way to go about this. But the thesis here, and this is something that we also have seen in the market, like if this works, then we have solved simulation.Joon [00:38:54]: It,Swyx [00:38:55]: Because I survey, like, okay, 5% of the US population is in construction.Swyx [00:39:01]: The other 5% is in medicine, whatever, right? And then you just keep going down the list, and then you do the other side. 5% has, like, the big 5 personalitySwyx [00:39:08]: Of, like, neurotic or whatever. That's it.Joon [00:39:11]: That's it. So if you believe that the underlying data set and the platform that we're leveraging has all the right statistics, then this will have solved it. you're at that point merely retrieving the knowledge that is already embedded in the model, in the model parameters. That's not, unfortunately, what we see, where there is such detailed and also niche knowledge about people that if you just take one example, it might feel very mundane, but it's quite rich when you put together, that you do need to do a lot of bespoke data collection to better understand people. And this is also, I think what makes this particular, job fun, which you want to deeply understand people, and the process of deeply understanding them requires a lot of attention to the details. And you do need to pay attention to and pay respect to the daily lives that people lead.Scaling Simulation: From Thousands to SocietiesVibhu [00:40:04]: I wanna talk about scaling simulation.Vibhu [00:40:07]: So what can't we simulate, what can we simulate, and how does scaling affect this? So how big are the models? What if we go from, 8B, like, couple 100 billionVibhu [00:40:18]: Like billion000 parameters, billion000? Do we get scaling? Any interesting emergence? Like, at a certain scale, at a certain amount of training, you uncover anything unusual and any learnings from that?Joon [00:40:31]: What we are seeing is at Simile, so we do post-train our own model. The thing that we're seeing is the early glimpse of scaling law in simulations. The more data about humans and more compute you ingest, you start to get predictive and predictable gains of the model performance in simulating it, simulating people.Vibhu [00:40:51]: Ooh. We need a scaling law curve.Joon [00:40:52]: It's scaling law. Whenever you find it's a beautiful thing. And we're starting to see the glimpse of it, which is quite exciting. But if you talk about the ambition of simulation as a whole, it's not merely about building a model. It's about building a model, then creating the agents that become the individuals in a much larger ecosystem. So they're creating this multi-agent simulation. Down the line, you want these multi-agent simulation to also live in a very rich environment, right? What we are really trying to get to at that point is, hey, can we create. All right, let's do a time machine game again, and 5 years, 10 years into the future, can we create a simulation of 8 billion people living on Earth? I think that's quite interesting. And that really is the vision. And once you get to that state, the questions that you can help answer for the society also start to change from my perspective. The answers are fundamentally about emergence of the emergent behavior of society and large groups of people.Joon [00:41:53]: So the questions that I get excited by, and maybe this is a stodgy- a bit. I have my, academic side of me.Joon [00:42:01]: And for me, it's questions like, can we help solve climate change? If you look at climate change as a problem space, this is what we, like social scientists would often call it the wicked problems, problem where you have many actors with competing incentives for trying to make a very complex decision and coordinating that coordination decision. Very difficult to really solve in real life, which is also the reason why we couldn't solve it. Can simulation help us solve that? Another one is, can we understand the signals for collapsing democracy, or can we understand or can we uncover the origin story of the monetary system? These are societal questions that we never really had a good way of answering. If we can create simulations of our society, you have to believe that these are the problems that we can solve. So that's really the ambition of this field. And, I also think, yes, I think there's a Nobel Prize to be won there, which wouldn't be surprising. And I think there's some amazing societal impact that we can have to help people make better decisions.Climate Change, Democracy, and Societal SimulationSwyx [00:43:04]: Nobel Prize in economics?Joon [00:43:06]: In economics.Swyx [00:43:06]: Oh, I see. I see. Rooting for you to write that paper.Joon [00:43:10]: One of these days. But, one of the scholars that I was deeply inspired by, When I was coming into the space of simulation, is this scholar, named Thomas Schelling.Schelling, Agent-Based Models, and the Nobel PrizeSwyx [00:43:23]: Schelling point?Joon [00:43:24]: So the canonical example of the work that he's done was he was one of the creators of agent-based modeling. So this was, like, in the 1970s and 80s. It's very early days, but this was truly one of the first exemplars of simulations. And one of the canonical model from that time, and of course many of these simulations are trying to tackle the societal problems that's most relevant for their era, it was called the model of segregation. So racial segregation was a big topic, that, we cared about. And what they've done was they created this grid world where they had red dots and blue dots. And these dots were, back in the day, like, they were the agents, and they had a simple rule that governed their behavior. If certain percentage of your neighbors are of different color and if that goes above certain threshold, then you move to a new location at random.Joon [00:44:21]: One of the striking finding of this paper or this agent-based model was for the longest time, people thought the segregation within society was caused by explicit and overt racism.Joon [00:44:34]: But if you look at this model, people's preference towards living with people of the same color, that preference can be very minute.Joon [00:44:42]: But the very small difference causes the society to segregate completely over time. This was very counterintuitive for a lot of people. And this particular work ended up informing housing policies. Mixed income housing, got really inspired by this work. And Thomas Schelling ends up winning the Nobel Prize for having laid the groundwork for very early versions of simulations. The opportunity that I do see here in the more scientific terms, is agent-based models for the longest, had impact in the 1980s, 90s, to some extent, early 2000s, but it has now gotten forgotten by the community a little bit. Because as you can imagine, red dots and blue dots is not really a rich description of people.Joon [00:45:31]: But with the emergence of things like generative AI and, in particular, generative agents, we do have an opportunity to create these agent-based models that are high fidelity enough to help us make really complex decisions. And that's the opportunity that I see. If that truly works, then yes, that is the work that will result in a Nobel Prize.Swyx [00:45:53]: Yeah. For what it's worth, and I grew up in Singapore. 80% of Singapore is in public housing, and public housing has, enforced racial quotas for exactly that reason, which is very interesting. okay, so we talk about scaling, we talk about all these, the agent possible applications.Cost, Reuse, and the Economics of SimulationSwyx [00:46:13]: I'm scared about the cost. if you even-- let's just keep it to the US, about 8 billion people.Swyx [00:46:21]: But, how much does it cost to model so many hundreds of millions of people?Joon [00:46:26]: Oftentimes today, we don't start at that scale, this stage of the, of industry and simulation as technology. But we can get our users extremely rich and meaningful insights even by modeling thousands, tens of thousands of people. And today what we do is every week we are collecting data on the scale of tens of thousands people's data, and we have panel partnerships that gets us to tens of millions of people globally. So that's what we do today.Swyx [00:46:55]: And just as a side note once you've collected one person for one studySwyx [00:46:59]: Can you reuse that same person for all the subsequent studies?Joon [00:47:03]: That's exactly right.Swyx [00:47:03]: Okay.Joon [00:47:04]: The beauty of this model and these agents is the fact that they are domain-agnostic.Joon [00:47:08]: That what you're really trying to understand is what is the fundamental nature of these people? What's their social physics? And there are a lot of, a lot of, people that does change over time. Like, even, like, even things like, how many times have you gone have you been to, like, CVS the past week? that will change. But there's so many traits about people that are also known to never change. Like, your risk tolerance doesn't really change over time. It's very consistent. So it's these things that we're trying to learn. But the scale we are operating is right now hundreds or, tens of thousands to hundreds of thousands. And in many of the core use cases that we are deployed in, and this is more than enough population, to cover those. Really, at that point, what you care about is less the number of people, but more do you have the right subpopulation of interest covered? And this is also the reason why people want a larger sample. It's not because they want, stronger statistical guarantees. It's more that can they filter down to any population of their interest. However, you can also imagine in 10 years, if we truly believe that the compute is going to scale, that we'll have much more availability for compute, and our ambition for simulation is also going to scale accordingly, there's definitely a reason for us to create an entire data center worth of simulations.Joon [00:48:35]: Or in my hunch here is I do think in the next some number of years, we will start creating simulations that will cost as much as training a foundation model. But perhaps it's going to be so valuable to the society that it would be a no-brainer. Right now, even today, like, we are training bunch of new foundation model just so we can say we trained one and we spent tens of millions. But if we can create a simulation at the level of society that would solve climate change, I would run that today. I would raise the money right now just to run that.Multi-Agent Simulation and Social InfluenceSwyx [00:49:10]: Amazing. the follow-up question is, does it also compound if you let the simulations talk to each other?Swyx [00:49:18]: Or do they already do that today? They don't, right, as far as I understand?Joon [00:49:22]: It depends on what simulation you're trying to run.Joon [00:49:24]: In the multi-agent simulation setup, the agents do talk to each other.Swyx [00:49:28]: Right, which is exactly Smallville, right?Joon [00:49:29]: That's right.Swyx [00:49:30]: But a lot of times, for example, in commerce, you're just by yourself, so there's no point talking. which is way cheaper.Vibhu [00:49:37]: But they use all these levels, right? Like, you decide what you will buy based on what other people around you buy and talk about, right?Swyx [00:49:43]: It depends.Vibhu [00:49:44]: It depends.Swyx [00:49:45]: Again, I'm, I'm coming at this from a cost point of view. I'm like, “Oh my God.” LikeVibhu [00:49:48]: I thinkSwyx [00:49:49]: If there is, like, some combinatorial thing of, like, thousands of people talking to thousands of people, then that one million X's might cost.Vibhu [00:49:56]: I have a very different view as the cost point aside. Like, running these studies in reality is a lot more expensive, right? Running any study like this is you gotta have people do it, you gotta sign people up. It's very expensive and sometimes, like, not feasible to run the study.Vibhu [00:50:14]: But the outcome or the decisions you make are very expensive on them, right? So spend X million on something that, the overall process costs 100 million might as well, right? There's, there's a lot of value to be had there. It's a small cost, but I'm excited on the cost side.Joon [00:50:33]: To some extent, and when you deploy technology, you often want to deploy in a way where you can replace existing budget or you can make things more efficient, and that is the best way to deploy. However, the way you capture the long-term value of the technology is making the argument that, no, it's the upside, that by making this better decision using simulation, you have saved yourself or made yourself hundreds of millions or even billions of dollars, and that's a case to be made.Vibhu [00:51:06]: Random tangent question. So if you're doing a lot of inference, a lot of model multi-agent stuff, are you at the point where it makes sense to, train a model that' very sparse? You're expecting to do multi-million dollar runs. Are you thinking about this in model architecture standpoint or inference efficiency, or, you're still at the research phase of it works, we're not super there yet?Joon [00:51:34]: Efficiency, we do think quite a bit about. this is technology that is deployed now in some of the largest enterprise companies in the world, and we do process significant number of queries, that are trying to, simulate the populations in the world. So efficiency is a consistent thing. we don't want to over-optimize too early, so I wouldn't say, like, this is the higher bid Right now, but this is definitely something that we think pretty carefully about.Swyx [00:52:05]: Yeah. Are there other case studies? So we, you talked about CVS, talked about Gallup, Deloitte, Wealthfront.Efficiency, Enterprise Use, and Real-World Case StudiesJoon [00:52:12]: Wealthfront is an interesting one, because one of the things they were trying to do, they were one of the first customers that wanted to do product testing that goes beyond just asking people what they think about, let's say, behavior experiments and so forth. So there, really what we had to do was reason about multimodal input, so images, but also you can also imagine, like, these agents traversing through Figma mockups or websites. So some of the things that our agents can also do is it can be given a domain, like, or, like, a website URL and go use it for a while. It's these things. And Wealthfront was one of the first, customers, that was very excited about this possibility.Vibhu [00:52:53]: What have people been asking? Like, is there any demand that we have not covered? Like, UI testing, right?Vibhu [00:52:59]: I wanna try a new. I wanna ship a new feature, test the UI, simulate how people will do it. Any interesting things that you're seeing demand for?Product Testing, Websites, and Synthetic PanelsJoon [00:53:08]: Today, a lot of the demand does come from like, the places where people have historically used human panels, we can now replace with agents, and these synthetic populations. And this is not replacing human panel. in many ways, the simulation that Simile is building is grounded. So the way that I think about this is we are trying to represent humanity at scale. And in that way, the use cases are what we would expect, but it's the scale of deployment that surprises me.Joon [00:53:44]: Turns out there are so many decisions that people make every day in these organizations, groups, and we want to be able to say, “We listen to people. We have consulted our users.” But in reality, that is rarely the case because getting to people and asking them many questions, it's difficult. It's both costly, time-consuming, but most importantly, people are just not available. If I had to answer 1000 survey questions for this one particular, vendor, even if I wanted to do that, like, I would never do it. And that's very much the case. What simulation can do is ensure that the voices of people are always represented in rooms where the decisions for them is made, right? So all the stakeholders of this particular product launch, ideally they're consulted. That's what this technology really is trying to enable.Market Size, TAM, and Human Decision-MakingSwyx [00:54:39]: In my mind, that means it skews towards more consumer focus, right? Like, anything with a wide enough customer base where you do benefit from the diversity that you represent. What are some rough statistics, just for people who are not familiar with this market in general, what's the market size that. I'm sure you have some, like, rough numbers. market size is, like, a vague questionSwyx [00:55:01]: But, like, how much do people spend?Joon [00:55:03]: So market research is a $100 billion industry.Joon [00:55:06]: But the thing about simulation is not a tool for market research. Simulation is a tool for human decision-making. So the question around what is a TAM here is quite tricky, right? Because it's easy to say, “Well, market research TAM is roughly 100 million or 100 billion.” so is it a TAM? And not really, right? Because in many ways, you're trying to inform all human decision-making. You're trying to inform every decision that are made about humans for humans. What is a TAM for that? It's really unclear. And I'll be honest. Like, I have a scientific background, I have a research background, so I didn't come into the field calculating, oh, what is the TAM for human decision-making? But I just had to assume, well, if we can inform every decision that is made about human for human, that has to be big.Swyx [00:55:58]: Some- something valuable.Joon [00:55:59]: Exactly.Swyx [00:55:59]: To some extent, you are a unicorn founder now, and you have to care as a CEO. But, like, I do think, like, yeah, when you go into these boardrooms with people that you're quoting millions of dollars of contracts for, like, you have to say, “Well, here's what you spend on humans-”Swyx [00:56:15]: “. And here's what we save you, and it's 85% similar.”Joon [00:56:19]: And certainly, the value case, is something that we care deeply about. Like, what is the value that we provide to the users and the decision-makers? But this is also where, like, as a founder, I think valuation only tells one very superficial aspect of the story, and I try not to think too much about valuation, in general, because that's not what also motivates a team or certainly doesn't. I'm, I-- Again, the interesting thing about researchers is we are happy living in academia, getting paid next to. we get paid okay. we don't get paid that much, as a researcher here in academia, but it's the impact and it's the, it's the value that we can provide to the individuals and the society that really drives us. And in that way, ultimately what drives us is the impact. Does the simulation we provide have a real impact in people's decision-making in ways that progresses our society forward? If the answer is yes, then yes. that has to be great business, and we see that in numbers, and we do care deeply about that upside story, but that's the heart of it.Where Simulation Goes NextVibhu [00:57:27]: Do you have any timeline predictions? So we talked about scaling laws of simulations.Vibhu [00:57:33]: You brought up, okay, maybe one day we can simulate how to solve climate change.Vibhu [00:57:38]: Where are we now?Vibhu [00:57:40]: If that's not the end state, what is an end state, and what does progress look like?Joon [00:57:45]: So what I sometimes tell people is simulation as industry, it feels a lot like where GPT-3.5, GPT-4 was, for the AGI saga, which is we have now technology that is powerful enough to do real damage on the verticals that we are tackling. At the same time, there's a lot of progress that is yet to come. And that's, I think, where this is. So the way I see it, I do think there will continue to be breakthroughs both in data, in algorithms, and there will be much more aggressive scaling that will also happen over the next few years. But I think that's roughly where we are.Swyx [00:58:27]: I think that was about the ro

Domingo Widen is a Staff Product Designer at Fin (formerly Intercom), where he leads the Frontend Infrastructure team — the team that builds the tools, design system and AI setup everyone else uses to ship. In this episode he opens up his actual workflow: a "Design Playground" repo every designer, PM and researcher can clone, prototyping directly in production, and "Search Intelligence" a persistent knowledge layer that teaches Claude how Fin designs and codes.We talk about what happens when the Figma file stops being the deliverable: how a complex filtering system got built as a working prototype instead of ten mockups, how handoff turns into a PR plus an auto-generated spec, and why engineers ended up being the biggest users of the design team's skills. If you want a concrete, unhyped look at what an AI-native design team actually does day to day, this one is for you.Key LearningsThere are two modes of AI prototyping and you need both. The design system is the guardrail that makes this safe. Because everyone builds on the same components and tokens, a PM or researcher can prototype without producing something off-brand or unusable.Skills are only as good as the information behind them. Fin's early skills gave mixed results — so they built "Search Intelligence", a persistent information layer that every skill reads from. Handoff becomes a PR + a generated spec. A "Search Handoff" skill reads the prototype code and writes the doc: architecture, what was built, which workarounds were needed, what broke. The engineer picks it up from there and focuses on plumbing, data and edge cases.The biggest users of the design team's skills are engineers. Everyone works in the same medium now. PM, designer and engineer used to live in Google Docs, Figma and VS Code. Now it's one shared artifact — which makes conversations concrete but also creates an identity crisis for designers who defined their value by the file.Curiosity + critical thinking are the two traits that survive.Play is back — and it counts. Side projects (a baby sleep tracker, a garden planner that checks on tomatoes) are now part of a portfolio.Domingo WidenLinkedIn: https://www.linkedin.com/in/domingowidenPortfolio: https://www.domingowiden.comX: https://x.com/DwidenR24Fin: https://fin.aiFin Operator (the tool Domingo prototypes on in production): https://fin.ai/operatorAI for Designers: 5-week Bootcamp

Shares of Moderna and Merck are surging on positive results from a Phase 3 trial of a groundbreaking cancer treatment. After a decade in development, individualized mRNA “vaccines” and established cancer treatment Keytruda result in reduced risk of recurrence or death. Moderna CEO Stephane Bancel and Merck Research Laboratories president Dr. Dean Li discuss this pivotal moment in medicine and how it fits into the long term visions for both Merck and Moderna. CEO of design and coding platform Figma Dylan Field discusses his company's wild year as a public company and building public trust in AI. Plus, it was primary night for Democrats in Florida, investors are scrutinizing revenues from Anthropic and OpenAI, and shares of Unitree are surging in Shanghai, presenting a potential challenge to U.S.-based humanoid robotics.    Stephane Bancel & Dr. Dean Li - 15:13 Dylan Field - 33:35   In this episode: Dylan Field, @zoink Becky Quick, @BeckyQuick Andrew Ross Sorkin, @andrewrsorkin Katie Kramer, @Kramer_Katie Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

The real skill with AI tools isn't prompting. It's knowing when to stop prompting, when to step in, do it by hand, and hand the result back to the machine.Darshan Gajara is a design leader who actually ships. His latest is Beanpresso, a coffee journaling app built in Lovable, launched, and refined in the wild. In this episode he shares his screen and walks through the whole build, the real project, the prompt history, the dead ends, and the moments a designer's eye had to take over, from first idea to live product, plus how he's approaching monetisation and marketing as a solo maker.We get into:Starting in Lovable and only bringing in Figma once the idea took shape, "give it a designer's touch"Why you don't need to design whole flows anymore: get one page right and Lovable infers your design systemThe handoff problem: screenshots plus precise description, component by component, down to corner radiiThe SVG trick: when the animation kept breaking, he built the UI by hand, named every layer, exported an SVG and told Lovable to animate exactly that"I'm still a designer": choosing to do the branding by hand, and why some things are better that wayMonetising a side project: premium taste insights, and a roaster marketplace his own users suggestedMarketing by making: free tools, a Berlin coffee guide the roasters reshared, and an AI writing setup trained on his own styleGET THE APPBeanpresso, coffee journaling for people who take their beans seriously: https://beanpresso.comMY REFLECTIONSI share my personal takeaways from each episode in the d.MBA newsletter, what I actually learned and what I'm stealing for my own builds: https://d.mba/newsletterLINKSDarshan on LinkedIn:   / darshangajara  Product Disrupt, Darshan's design learning resource: https://productdisrupt.comDarshan's site: https://darshan.designd.MBA, business education for designers: https://d.mba00:00 – "They just keep banging their heads, burning through tokens"00:36 – Who Darshan is, and what this episode is really about01:36 – Interview starts: scratching his own itch.05:51 – App demo: scanning a bag, community feed, taste stats10:51 – Validating the idea (and why the existing coffee apps weren't good enough)14:12 – Version one: built straight in Lovable, no design15:54 – The prompting workflow: voice notes → Notion AI → Claude → Lovable18:21 – Why he dropped the native app and went web19:33 – Mood boarding with Variant, and finding the visual language23:54 – Naming it, and why AI is terrible at logos27:15 – What AI is good at, and what he had to do by hand30:45 – Taking the design back into Lovable, component by component33:27 – Building new flows once Lovable knows your design system36:36 – The SVG animation trick38:27 – Don't expect one tool to do everything42:06 – Monetisation: premium stats and a roaster marketplace45:45 – Distribution: small free tools, Instagram, and SEO49:15 – Where to find Beanpresso and Darshan

Note: this episode contains screen sharing, so it's best watched on YouTube.What does a recruiter actually see in the ten seconds they spend on your LinkedIn profile, and what makes them scroll past?In this episode of Supra Insider, Marc Baselga and Ben Erez sit down with Chris Lee, founder of Product Scout, a boutique firm that helps early-stage founders make their first product hires. Chris explains the idea he calls candidate market fit, why he refuses to give profile feedback until he knows what role you're aiming at, and why a profile built to appeal to everyone ends up resonating with no one.Then he opens his actual tooling and walks through a live search for a head of product role at an AI-native pharma startup, showing the rubrics his agents score against and why almost everything at the top of the funnel comes down to keywords. He finishes by reviewing Marc's and Ben's own profiles against that role, out loud and unfiltered, and passes on both.If you're job searching and unsure whether your profile is working, weighing how specific to make your story, or a hiring manager curious how sourcing really happens now, this episode is for you.All episodes of the podcast are also available on Spotify, Apple and YouTube.New to the pod? Subscribe below to get the next episode in your inbox

Adam and Adir unpack San Francisco’s AI-fuelled tech boom, from Waymo and Zoox to Figma’s aggressive push and Lovable’s wild valuation. Later, Adir explains what DEF CON reveals about modern cybersecurity, before John Stensholt joins to discuss Tech Bros, Atlassian’s founder breakup and the strange billionaire world of Mike Cannon-Brookes and Scott Farquhar. GET YOUR TICKET TO THE CONTRARIANS LIVE! https://events.humanitix.com/the-contrarians-live John Stensholt's book 'Tech Bros' is available now: https://harpercollins.com.au/products/9781460769492 00:00 - San Francisco, Waymo and Zoox08:00 - Figma, AI and the API Boom17:15 - Lovable’s $13bn Valuation21:45 - Contrarians Live Show Announcement26:30 - LinkedIn’s Algorithm Problem31:30 - DEF CON, Hackers and Cybersecurity51:15 - John Stensholt on Tech Bros and Atlassian Join us on Substack for articles, news and more: https://www.thecontrarianspod.com/

Peter and Jesse are joined by Chengying "Z" Zheng, who shares her experience migrating a design team to be AI-first — off Figma and into code. She discusses how design leadership and operations divide the work, what happens to roles and titles when everyone becomes a builder, and how to make good design easier than bad design. Z on LinkedIn: https://www.linkedin.com/in/changyingz/ Z's Substack: https://changying.substack.com/ Jesse James Garrett: https://jessejamesgarrett.com/ Peter Merholz: https://petermerholz.com/

Blaine Billingsley is a Member of Technical Design Staff at OpenAI whose path into the field ran through music composition, spare-change website work at college, and a decade across Gmail, Airbnb, YouTube and Slack. At OpenAI he was hired to work on Presence, then redirected to solve a more immediate problem: making ChatGPT genuinely useful as a daily tool for designers. Working with a single engineer, he built the ChatGPT product design plugin — a bridge between the raw power of Codex and the day-to-day workflow of product teams. In this conversation with Randy Silver, he talks about what design actually means when your interface is a text box, why volume beats perfection in an age of infinite iteration, and why the fundamentals of good product thinking are more durable than the tools used to apply them.Key takeaways— The designer's job at an AI company has shifted from pixels to outputs. Deciding what "good" looks like, building the criteria to evaluate it, and heuristically assessing results is now a core part of the role — one that didn't exist in the same form at Gmail or Slack.— LLMs don't replace structured creativity techniques; they scale them. The crazy eights exercise squeezes eight ideas from a room of people in eight minutes. A well-directed ChatGPT session can return 80 in the same window, alone, while you're at lunch.— Volume beats perfection. The best way to make a pot isn't to try to make the best pot — it's to make a thousand pots. That logic now applies directly to prototyping: generate at scale, stay unattached, and find the nugget in the noise.— Evals are closer to synthetic user research than quality assurance. The hardest part isn't building the rubric — it's correctly anticipating what people will actually try to do. Show it to one more person and your assumptions will immediately break.— The second 80% problem hasn't gone away. Getting to a working prototype is dramatically faster; getting that prototype to production is still gruelling — and becomes harder still when platform direction shifts mid-sprint.— Small teams with AI assistance need to protect the rituals that keep them aligned. When two people can each produce a week's worth of work in an afternoon, parallel drift becomes the real collaboration risk.— Role boundaries are dissolving, but specialisations still matter. The question is less "what is your title" and more "what does the band need right now, and can you play that part?"— Experience brings judgment; freshness brings juice. The best work often comes from junior designers unconstrained by years of accumulated assumptions — and both things need to be in the room.Chapters (00:00) Introduction(01:14) Blaine's background: from music composition to product design (02:32) What design means at OpenAI (04:21) The ChatGPT product design plugin (06:12) Deciding what to build: the early exploration (10:05) Structured creativity and LLM-powered ideation (13:37) Volume over perfection: the thousand pots approach (16:13) Designing as a two-person team (20:29) What is the job now? (23:33) Evals as synthetic user research (27:08) Testing at scale when you can't know your users (30:15) How they actually did the research (33:01) The second 80%: from prototype to production (35:06) Staying aligned without roadmaps (38:19) Demo: the product design plugin (44:52) When a prototype isn't ready to ship (48:26) Design sprints, reimagined (49:27) Advice for joining an AI-first product team (52:17) The jazz analogy: experience, freshness and your role in the band Featured LinksChatGPT for Work: https://openai.com/chatgpt Codex: https://openai.com/codex Figma: https://figma.com FigJam: https://www.figma.com/figjam Linear: https://linear.appOur HostsLily Smith enjoys working as a consultant product manager with early-stage and growing startups and as a mentor to other product managers. She's currently Chief Product Officer at BBC Maestro, and has spent 13 years in the tech industry working with startups in the SaaS and mobile space. She's worked on a diverse range of products – leading the product teams through discovery, prototyping, testing and delivery. Lily also founded ProductTank Bristol and runs ProductCamp in Bristol and Bath.Randy Silver is a Leadership & Product Coach and Consultant. He gets teams unstuck, helping you to supercharge your results. Randy's held interim CPO and Leadership roles at scale-ups and SMEs, advised start-ups, and been Head of Product at HSBC and Sainsbury's. He participated in Silicon Valley Product Group's Coaching the Coaches forum, and speaks frequently at conferences and events. You can join one of communities he runs for CPOs (CPO Circles), Product Managers (Product In the {A}ether) and Product Coaches. He's the author of What Do We Do Now? A Product Manager's Guide to Strategy in the Time of COVID-19. A recovering music journalist and editor, Randy also launched Amazon's music stores in the US & UK.

Is your business creating more content with AI… but getting less attention from it?AI has made it ridiculously easy to produce images, ads, social posts, and product visuals. Unfortunately, easy doesn't mean effective. When everyone has access to the same tools, generic prompts tend to produce generic content—and generic content rarely moves the business needle.The solution isn't another 500-word “perfect prompt.” It's giving AI better context, references, brand knowledge, and creative direction—and knowing when human refinement still matters.In this episode of Leveraging AI, Isar Meitis sits down with Aastha Taneja to break down a practical workflow for creating AI-powered visual assets that look intentional, stay consistent with your brand, and are designed to generate engagement rather than simply fill your content calendar.Aastha demonstrates how she combines traditional creative thinking with tools including ChatGPT, Figma, Pinterest, image-generation platforms, and Photoshop. She also explains why she believes businesses should treat AI as an assistant—not outsource the entire creative process to it.In this session, you'll discover:Why so much AI-generated marketing content turns into forgettable “AI slop.”Why better AI creative starts before you write a prompt.How mood boards give AI a far clearer understanding of the visual direction you want.How to feed AI your website, brand assets, SOPs, typography, and existing creative to improve its output.The difference between borrowing a proven format and copying someone else's creative.How to maintain accurate product details when AI image generators get almost everything right—but miss one critical element.How AI can help businesses create high-quality Amazon listing images and other e-commerce assets.How reusable visual workflows can turn a manual creative process into a scalable content system.Aastha Taneja is a creative professional with years of experience developing digital assets for brands and through freelance work. She now combines that traditional design foundation with AI, helping businesses and professionals understand how to use AI tools more effectively for creative work and sharing those methods through corporate training.Her core philosophy is refreshingly practical: AI should amplify good creative thinking—not replace it.Connect with Aastha on LinkedIn:https://www.linkedin.com/in/taneja-aastha/ About Leveraging AIThe Ultimate AI Course for Business People: https://multiplai.ai/ai-course/YouTube Full Episodes: https://www.youtube.com/@Multiplai_AI/Connect with Isar Meitis: https://www.linkedin.com/in/isarmeitis/ Join our Live Sessions, AI Hangouts and newsletter: https://services.multiplai.ai/eventsIf you've enjoyed or benefited from some of the insights of this episode, leave us a five-star review on your favorite podcast platform, and let us know what you learned, found helpful, or liked most about this show!

UX portfolio advice is full of numbers that sound official but aren't backed by anything. This episode kicks off a five part series on UX portfolio myths, starting with the most common one: that you need three to five projects before you can apply.Sarah says the opposite is true. One strong, detailed project beats five surface level ones every time. She backs this up from an interview she did with Alexander Zeh, former Head of Product Design at ManyChat, who said he'd rather see one or two case studies that map directly to the job description than five that walk through every step of the design process without any real depth.The episode shares two specific examples. Laura got hired as a Principal Product Designer in cybersecurity with a single, unfinished feeling project still sitting in Figma slides, and was actually found and reached out to on LinkedIn before she even applied. Leon, laid off after 11 years at the same company and an author of a book on wire-framing, got hired after focusing on just one project instead of trying to rebuild an entire portfolio from scratch.If you've been stalling on applying because you don't have "enough" projects yet, this episode is a direct challenge to that belief, and a reason to ask where that number actually came from in the first place.Topics Discussed✅ Why the three to five project rule is a myth with no real backing, and where it likely came from✅ Why one detailed project beats five surface level ones, and what "confuse and lose" means for a UX portfolio✅ What a former Head of Product Design at ManyChat said he actually wants to see in a portfolio✅ How Laura got hired as a Principal Product Designer with one unfinished project, and was found on LinkedIn without ever applying✅ How Leon got hired after being laid off from an 11 year role, focusing on just one project instead of rebuilding his whole portfolio✅ Why working on your portfolio alone can make you more critical of it than recruiters and hiring managers actually are✅ The question to ask yourself if you believe you need three to five projects, and where to trace that belief back to✅ What Sarah says to look for in UX job descriptions instead of following a generic project countLinks & Resources

What happens when you pass every interview, get approved for an offer, and then hear nothing for five months?In this episode of Supra Insider, Marc Baselga interviews Ben Erez about team matching, the stage at large companies where an approved candidate waits for a hiring manager to claim them. Ben walks through how the process actually works from the inside, drawing on his own path into Facebook and what he has since heard from people going through it now.They explore why hiring managers exhaust internal candidates first, what a company-wide reshuffle does to everyone waiting outside, the reported changes at Meta that have left candidates in silence, and the moves that are still within a candidate's control, including how to see the roles that never get posted publicly and what to say when you reach out.If you're sitting in team matching wondering whether to keep waiting, preparing for a generalist loop at a big company, or a leader trying to understand why your approved candidate pool keeps growing, this episode is for you.All episodes of the podcast are also available on Spotify, Apple and YouTube.New to the pod? Subscribe below to get the next episode in your inbox

Adam and Adir discuss Zoox, Waymo, autonomous cars, Xero’s strange Vegas advertising, Sukhinder Singh Cassidy, Victoria’s work-from-home laws, Delta One, airline lounges, Jetstar’s carry-on bag crackdown, Ryanair, Canva’s slowing growth, AI design tools, Claude, OpenAI, Figma, Atlassian’s profit pivot and the future of Australia’s biggest tech companies. Join us on Substack for articles, news and more: https://www.thecontrarianspod.com/See omnystudio.com/listener for privacy information.

This Week In Startups is made possible by: DigitalOcean https://do.co/twist Sentry https://sentry.io/twist Lightfield https://lightfield.app Today's show: A hedge fund just blew up shorting SaaS stocks. In turns out, the software companies that went all in on AI are bouncing back. Figma's CEO — still riding high on strong revenue — forfeited roughly $46 million in stock awards to ease investor nerves. Twilio posted its strongest quarter in years. Airbnb's CEO credits AI as the single biggest factor in the company's incredible turnaround. Jason and Lon break down what's separating the winners this earnings season from the companies that are falling behind. PLUS we're chatting with Bluecore Energy CEO and founder Kofi Asante about his plan to power ports, and one day AI data centers, with small, floating nuclear reactors. AND we're kicking off the TWiST Weight Loss Challenge! Co-host Lon and Jason's brother, Jamie, are both starting GLP-1s. We'll track their progress and follow along as they attempt to collect $15,000 EACH in prize money. Guest Kofi Asante on LinkedIn: https://www.linkedin.com/in/kofiasante1/ Bluecore Energy: https://www.bluecore.energy/ Jamie Calacanis on X: https://x.com/CalacanisJamie Relevant Links Bloomberg: Figma CEO forfeits $46M: https://www.bloomberg.com/news/articles/2026-08-05/figma-ceo-forfeits-46-million-in-stock-after-share-decline Twilio announces second quarter 2026 results: https://investors.twilio.com/news-releases/news-release-details/twilio-announces-second-quarter-2026-results TechCrunch: Twilio co-founder Jeff Lawson buys The Onion: https://techcrunch.com/2024/04/26/area-man-twilio-co-founder-jeff-lawson-buys-the-onion Chesky: Airbnb will spend a lot more on AI: https://www.cnbc.com/2026/08/07/chesky-airbnb-ai-earnings.html "Unreasonable Hospitality" by Will Guidara on Amazon: https://www.amazon.com/Unreasonable-Hospitality-Remarkable-Giving-People/dp/0593418573 The Times: Meet the unreasonable hospitality master who inspired "The Bear": https://www.thetimes.com/magazines/the-sunday-times-magazine/article/will-guidara-hospitality-the-bear-tkgzqz33w Port of Long Beach partnership on nuclear-powered shipping: https://gcaptain.com/port-of-long-beach-marad-launch-first-u-s-partnership-on-nuclear-powered-shipping/ Reuters: ByteDance targets mega AI model: https://www.reuters.com/technology/bytedance-targets-mega-ai-model-nearing-anthropics-mythos-ft-reports-2026-08-07/ Axios: OpenAI says its AI agents breached its own systems before Hugging Face: https://www.axios.com/2026/08/06/openai-hugging-face-black-hat Frank Lloyd Wright's Ennis House: https://franklloydwright.org/site/ennis-house/ Dwell: Why "The Studio" built a faux Frank Lloyd Wright HQ: https://www.dwell.com/article/the-studio-apple-tv-filming-locations-production-design-frank-lloyd-wright-john-lautner-97c68081 "Predators" documentary trailer: https://www.youtube.com/watch?v=ibVZkvpAJjI A24 "Primetime" trailer: https://www.youtube.com/watch?v=5fHXyqQOKL8 Alex Honnold free solos Taipei 101: https://www.cnn.com/2026/01/25/sport/alex-honnold-taipei-101-free-solo-netflix-intl-hnk Harmonic.ai: https://harmonic.ai/ Ro.co: http://ro.co/twist Timestamps: 0:00 Figma, Twilio & Airbnb earnings: who's winning at AI? 9:50 DigitalOcean - Head to https://do.co/twist to start building on DigitalOcean's AI-Native Cloud today — and cut your AI workload costs by up to 50%. 11:32 The real cost of stock-based compensation 17:38 Kofi Asante of Blue Core Energy joins the show 19:47 Sentry - Your team should be focused on shipping features — not chasing down bugs. New users can get $240 in free credits when they go to https://sentry.io/twist and use the code TWIST 20:51 What is a small modular reactor (SMR)? 26:00 Floating nuclear reactors for ports and data centers 29:47 Lightfield - Name one person who's ever enjoyed updating a CRM. Exactly. Lightfield's AI agent does it for you — it even prospects and books your meetings. Used by thousands of startups. Free at https://lightfield.app 31:56 Nuclear waste, submarine risks, and public perception 42:41 The generational shift in how we think about nuclear 48:19 ByteDance is training a 10 trillion parameter model 53:47 When AI models plan heists: the Hugging Face caper 58:02 Jason's Roko weight loss competition 1:11:38 Off duty: To Catch a Predator, Alex Honnold, and comedy Subscribe to the TWiST500 newsletter: https://ticker.thisweekinstartups.com Check out the TWIST500: https://www.twist500.com Subscribe to This Week in Startups on Apple: https://rb.gy/v19fcp   Follow Lon: X: https://x.com/lons   Follow Jason: X: https://twitter.com/Jason LinkedIn: https://www.linkedin.com/in/jasoncalacanis   Check out all our partner offers: https://partners.launch.co/   Great TWIST interviews: Will Guidara, Eoghan McCabe, Steve Huffman, Brian Chesky, Bob Moesta, Aaron Levie, Sophia Amoruso, Reid Hoffman, Frank Slootman, Billy McFarland   Check out Jason's suite of newsletters: https://substack.com/@calacanis   Follow TWiST: Twitter: https://twitter.com/TWiStartups YouTube: https://www.youtube.com/thisweekin Instagram: https://www.instagram.com/thisweekinstartups TikTok: https://www.tiktok.com/@thisweekinstartups Substack: https://twistartups.substack.com

Anastasios Angelopoulos is the co-founder and CEO of Arena, the real-world evaluation platform that has become a leading referee of the global AI model race. Arena has raised $250 million, with the latest round valuing the company at $1.7BN. Arena recently surpassed $100M ARR just eight months after launching its enterprise offering, powered by more than 30 million monthly users. AGENDA: 00:00 – Intro: "Kimi beat every American model": What Nobody Wants to Admit… 05:20 – Is this the true commoditization of models? Are they just a utility layer now? 07:30 – Do Chinese open source models cannibalize the closed frontier labs? 10:30 – Why has America's open source community lagged so badly behind China? 17:20 – Will Chinese models be banned in the US — and does hosting locally really kill the backdoor risk? 23:20 – Are enterprises really terrified of working with the frontier labs? 27:20 – Why hasn't inference got cheaper — and what happens when Anthropic's "disgustingly high" margins go public? 30:20 – Who should decide if a model is safe to release: the government, a neutral body, or nobody? 33:10 – Are we about to see cyberattacks like we've never seen before? (The fake candidate who passed every interview) 37:00 – 75 Neo labs: what separates the winners from the two-thirds worth nothing? 40:45 – Is data actually a commodity — and can data providers be $100BN companies? 48:00 – Can you be the referee when the players are paying you? (And Arena's real revenue) 50:45 – Will the model providers eat the application layer? Are Harvey, Lagora and Figma in trouble? 54:10 – Quickfire: Why hasn't NVIDIA bounced on the rise of open source, who hits $10 trillion first, and does the compute debt cycle end in insolvency?    

Google didn't ship its big model, but they shipped a TON of new useful AI you can use today. And Google wasn't the only company updating their features behind the scenes. Replit is bringin vibe designing, ChatGPT got a lot more useful on the web, and Meta is changing from chatbot to agent. We'll get you caught up quickly. Chrome adds Some Gemini Spark, Replit Design makes impact, Buzz brings AI Agent Teamwork and 7 more AI Features you Should use Today -- an Everyday AI Chat with Jordan WilsonNewsletter: Sign up for our free daily newsletterMore on this Episode: Episode PageToday's Episode on LinkedIn: Thoughts on this? Join the convo on LinkedIn and connect with other AI leaders.Upcoming Episodes: Check out the upcoming Everyday AI Livestream lineupWebsite: YourEverydayAI.comEmail The Show: info@youreverydayai.comConnect with Jordan on LinkedInTopics Covered in This Episode:Replit Design Suite Launches With Free MobbinChatGPT Chrome Extension Adds YouTube SummarizationChatGPT Side Chat Integrates Tabs and Highlighted TextMeta AI Rolls Out Recurring Agent TasksGoogle Gemini Generates Images in Google DocsGemini AI Summarizes Comments, Edits in DocsGoogle Gemini Spark Agent Arrives in ChromeChrome Agent Uses Saved Accounts and PasswordsGoogle Lyria 3.5 Music Model ReleasedBuzz by Block Unites Team and Agent CollaborationTimestamps:00:00 Recent AI updates and developments05:01 Creating with Replit and AI models09:42 Real-time research tracking benefits10:34 Meta AI new recurring features13:35 New features of Meta AI17:53 Google Spark integrates with Chrome22:09 Google DeepMind's new music model25:25 Buzz from Block messaging tool29:42 Building a collaborative platform31:23 AI feature updates recapKeywords: Gemini Spark, Google Chrome AI integration, Google Docs AI features, AI image generation, Gemini in Docs, ChatGPT Chrome extension, YouTube video summarization, OpenAI ChatGPT update, Codex, Vibe design, Replit design suite, Mobbin integration, AI reference library, Design export automation, Project management AI, Figma competitor, Replit creative tools, Meta AI, Muse Spark 1.1, Agentic model, Recurring AI tasks, AI scheduling, Daily briefings, AI productivity tools, Google Lyria 3.5, AI music model, Flow Music, Suno, Yudio, AI generated lyrics, Vocal delivery in AI music, Licensing in AI music, Buzz collaboration platform, Block, Square, AI agent teamwork, Slack-like AI platform, Open source collaboration, Agent governance, Cryptographic identity, Agentic browser, Automated web errands, Chrome passwords integration, Google Drive data access, Multi-agent collaboration, Research automation, Enterprise AI workflow, AI productivity boost.Send Everyday AI and Jordan a text message. (We can't reply back unless you leave contact info) Ready for ROI on GenAI? Go to youreverydayai.com/partner 

Anthropic ha pasado en unos pocos años de ser el gran reformador de la inteligencia artificial a ser el gran sospechoso. Fundada en 2021 por Dario Amodei y otros ex empleados de OpenAI, Anthropic se presentaba como la conciencia moral de la inteligencia artificial, el laboratorio que anteponía la prudencia a la ambición. Hoy buena parte de Silicon Valley la mira con recelo y constatan que es como cualquier otra, una empresa que compite con las mismas armas que las demás, solo que envueltas en el celofán de las buenas intenciones. El primer aldabonazo llegó en abril con Claude Design, un competidor directo de Figma cuyo consejero delegado, Dylan Field, reprochó en público la falta de sinceridad de un socio con el que habían estado trabajando mano a mano. En junio se sumó Fable 5, la versión restringida del modelo Mythos, que degrada en silencio ciertas respuestas, no admite retención cero de datos y guarda las conversaciones con los usuarios durante 30 días. Al mismo tiempo Amodei se ha puesto en contra de los modelos de pesos abiertos, más baratos y difíciles de controlar. Lo que muchos sospechan es que está creando un pánico para después ofrecerse a resolverlo. El calendario tampoco ayuda, la compañía planea salir a bolsa en otoño y esos modelos abiertos amenazan sus elevados márgenes. Nvidia, Microsoft, Meta, OpenAI, Google y más de 70 empresas han hecho pública una carta abierta en la que defienden los modelos abiertos. Eso ha dejado a Anthropic prácticamente sola. Amodei no ha tardado en responder arguyendo que su oposición se debe a que quiere evitar que gobiernos autoritarios amenacen la seguridad nacional. El otro campo de batalla es China. Hace dos meses, en una carta dirigida a dos senadores, Anthropic acusaba a Alibaba del mayor ataque de destilación del que se tiene noticia. Al parecer Alibaba creo unas 25.000 cuentas falsas que hicieron 29 millones de preguntas a Claude para extraerle información sobre sus mejores capacidades. Lo curioso aquí es que la propia Anthropic ha destilado todo internet sin pagar derechos de autor y ahora exige a los chinos un respeto por la propiedad intelectual que ellos no tuvieron con nadie. Detrás de esta guerra de los pesos abiertos, a las grandes de la inteligencia artificial se les mueve el suelo bajo los pies, pero por otra cosa bien distinta. Están perdiendo mercado. Como veíamos en el programa del lunes, sus clientes han descubierto que no necesitan gastar una fortuna en tokens de un mismo proveedor y están empezando a repartir su consumo entre modelos más caros y capaces y otros más baratos para tareas rutinarias. Así se entiende mejor la doble cruzada que ha emprendido Anthropic contra la destilación china y contra los modelos abiertos. Ambas apuntan al mismo blanco, la competencia barata que machaca sus márgenes, envuelta, eso sí, en el noble lenguaje de la seguridad nacional. En La ContraRéplica: 0:00 Introducción 3:21 Todos contra Anthropic 31:28 El responsable del incendio de Ávila 37:55 La despoblación rural 43:27 Medios e IA · Canal de Telegram: https://t.me/lacontracronica · “Contra el pesimismo”… https://amzn.to/4m1RX2R · “Hispanos. Breve historia de los pueblos de habla hispana”… https://amzn.to/428js1G · “La ContraHistoria del comunismo”… https://amzn.to/39QP2KE · “La ContraHistoria de España. Auge, caída y vuelta a empezar de un país en 28 episodios”… https://amzn.to/3kXcZ6i · “Contra la Revolución Francesa”… https://amzn.to/4aF0LpZ · “Lutero, Calvino y Trento, la Reforma que no fue”… https://amzn.to/3shKOlK Apoya La Contra en: · Patreon... https://www.patreon.com/diazvillanueva · iVoox... https://www.ivoox.com/podcast-contracronica_sq_f1267769_1.html · Paypal... https://www.paypal.me/diazvillanueva Sígueme en: · Web... https://diazvillanueva.com · Twitter... https://twitter.com/diazvillanueva · Facebook... https://www.facebook.com/fernandodiazvillanueva1/ · Instagram... https://www.instagram.com/diazvillanueva · Linkedin… https://www.linkedin.com/in/fernando-d%C3%ADaz-villanueva-7303865/ · Flickr... https://www.flickr.com/photos/147276463@N05/?/ · Pinterest... https://www.pinterest.com/fernandodiazvillanueva Encuentra mis libros en: · Amazon... https://www.amazon.es/Fernando-Diaz-Villanueva/e/B00J2ASBXM #FernandoDiazVillanueva #anthropic #ia Escucha el episodio completo en la app de iVoox, o descubre todo el catálogo de iVoox Originals

Claude Design went viral at launch, but has been kinda crickets since.

Noam Segal is a longtime research leader across Airbnb, Meta, Twitter, Zapier, Intercom, and Figma, a certified coach, AI builder, and my community research lead. Together, we run the annual Tech Worker Sentiment Survey, now in its second year and one of the largest of its kind: a quantitative study of how people in tech actually feel about their jobs, AI, burnout, and the future of their careers. This year's survey captured responses from thousands of workers across product, engineering, design, research, marketing, data, and sales, and the results are striking.In our in-depth conversation, we discuss:1. Why AI has split the tech workforce almost exactly in half—one half that's thriving, another that's shaken2. The four emotional archetypes defining tech workers right now (the Energized, the Conflicted, the Disoriented, and the Resentful)3. Why burnout has jumped an alarming 11 points in a single year4. Why nobody in tech would recommend their job to someone entering the industry today5. The #1 fear in tech right now (it's not job loss to AI)6. Why managers are the single biggest lever for employee well-being7. Concrete advice for what employees and leaders can do right now—Brought to you by:WorkOS—Make your app enterprise-ready, with SSO, SCIM, RBAC, and more: https://workos.com/lennyMercury—Radically different banking, now with Command: https://mercury.com/command?utm_source=lennys&utm_medium=sponsored_newsletter&utm_campaign=26q3_brand_campaign—Episode transcript: https://www.lennysnewsletter.com/p/how-tech-workers-actually-feel-about—Archive of all Lenny's Podcast transcripts: https://www.dropbox.com/scl/fo/yxi4s2w998p1gvtpu4193/AMdNPR8AOw0lMklwtnC0TrQ?rlkey=j06x0nipoti519e0xgm23zsn9&st=ahz0fj11&dl=0—Where to find Noam Segal:• X: https://x.com/noamseg• LinkedIn: https://www.linkedin.com/in/noamsegal—Where to find Lenny:• Newsletter: https://www.lennysnewsletter.com• X: https://twitter.com/lennysan• LinkedIn: https://www.linkedin.com/in/lennyrachitsky/—In this episode, we cover:(00:00) Introduction to Noam Segal(02:34) About the survey: methodology and scope(06:04) The core finding: AI has split the tech workforce in half(13:03) The AI identity stance(14:40) The four archetypes: Energized, Conflicted, Disoriented, Resentful(19:35) Burnout is surging (and why shipping faster is making it worse)(22:53) A glimmer of hope(24:55) Layoff worries(29:15) The career recommendation NPS score(36:45) The ladder metaphor: rungs disappearing beneath our feet(45:14) AI is making us faster, not better(52:53) The #1 fear: being squeezed to do more for the same pay(55:55) The emotional landscape and “smiling exhaustion”(01:01:02) Designers and researchers: the most negative group two years running(01:06:27) Who's happiest(01:12:18) Managers: the single biggest lever on well-being(01:18:47) The industry is “chaotic”(01:24:53) What employees and leaders can do right now(01:31:32) AI guilt and closing thoughts—Referenced:• How tech workers are feeling in 2026: a workforce splitting in two: https://www.lennysnewsletter.com/p/how-tech-workers-are-feeling-in-2026• How tech's most resilient workers handle burnout: https://www.lennysnewsletter.com/p/how-techs-most-resilient-workers• Please stop the AI Confidence Theater: https://www.elenaverna.com/p/please-stop-the-ai-confidence-theater• Velocity over everything: How Ramp became the fastest-growing SaaS startup of all time | Geoff Charles (VP of Product): https://www.lennysnewsletter.com/p/velocity-over-everything-how-ramp• NPS Is The Worst: https://www.npsistheworst.com• The Terminator: https://www.imdb.com/title/tt0088247• Skynet: https://terminator.fandom.com/wiki/Skynet• Inside Devin: The world's first autonomous AI engineer that's set to write 50% of its company's code by end of year | Scott Wu (CEO and co-founder of Cognition): https://www.lennysnewsletter.com/p/inside-devin-scott-wu• Devin: https://devin.ai• An AI state of the union: We've passed the inflection point, dark factories are coming, and automation timelines | Simon Willison: https://www.lennysnewsletter.com/p/an-ai-state-of-the-union• Redeploying Fable 5: https://www.anthropic.com/news/redeploying-fable-5• Why half of product managers are in trouble | Nikhyl Singhal (Meta, Google): https://www.lennysnewsletter.com/p/why-half-of-product-managers-are-in-trouble• Inside Linear: Building with taste, craft, and focus | Karri Saarinen (co-founder, designer, CEO): https://www.lennysnewsletter.com/p/inside-linear-building-with-taste• Building beautiful products with Stripe's Head of Design | Katie Dill (Stripe, Airbnb, Lyft): https://www.lennysnewsletter.com/p/building-beautiful-products-with• The design process is dead. Here's what's replacing it. | Jenny Wen (head of design at Claude): https://www.lennysnewsletter.com/p/the-design-process-is-dead• OpenAI Codex lead on the new shape of product work | Andrew Ambrosino: https://www.lennysnewsletter.com/p/openai-codex-lead-on-the-new-shape• Elon Musk: ‘Chances are we're all living in a simulation': https://www.theguardian.com/technology/2016/jun/02/elon-musk-tesla-space-x-paypal-hyperloop-simulation—Production and marketing by https://penname.co/. For inquiries about sponsoring the podcast, email podcast@lennyrachitsky.com.—Lenny may be an investor in the companies discussed. To hear more, visit www.lennysnewsletter.com

This Week In Startups is made possible by: NetSuite https://NetSuite.ai/TWIST Squarespace https://squarespace.com/twist YSecurity https://YSecurity.io/TWIST Today's show: *America posted 400,000 farm jobs last year, and fewer than 1% got a single domestic applicant. Danny Bernstein of Reservoir believes it's time to start automating this backbreaking labor, like picking stone fruit in 110°F temperatures. So he built the world's first on-farm robotics incubator on 40 acres of California farmland. Here's why he argues that automating farms isn't just a new opportunity, it's a national security issue. PLUS Jason responded to Gal Shir's viral "AI beat me at design" tweet, got into a kerfuffle with Figma CEO Dylan Field, and wound up making a brand new J-Trade. AND we're talking about the Dept. of Education's new "do no harm" policy, the Brown University AI cheating chart that's blowing up social media, and Lon has fresh streaming recommendations in an all-new Off Duty. Guest: Danny Bernstein on X: https://x.com/bernsteind Reservoir Farms: https://reservoir.co/ Reservoir VC: https://reservoir.vc/ Relevant Links: Bonsai Robotics: https://bonsairobotics.ai/ Root AI acquired by AppHarvest: https://www.therobotreport.com/root-ai-acquired-by-appharvest-for-60m/ John Deere: https://www.deere.com/en-us/ Western Growers Association: https://www.wga.com/ Tanimura & Antle: https://www.taproduce.com/ Naturipe Berry Growers: https://www.naturipefarms.com/ Driscoll's: https://www.driscolls.com/ Taylor Farms: https://www.taylorfarms.com/ The Wonderful Company: https://www.wonderful.com/ Capital Factory: https://www.capitalfactory.com/ Gal Shir "quitting design" post: https://x.com/galshirart/status/2074854464729629060 Dylan Field response to Gal Shir: https://x.com/zoink/status/2075290218660298807 JCal response to Field and "J-Trade": https://x.com/Jason/status/2075481565115654305 Figma: https://www.figma.com/ Dept. of Education: "Do No Harm" policy announcement: https://www.ed.gov/about/news/press-release/us-department-of-education-issues-final-rule-hold-all-colleges-and-universities-accountable-low-earning-programs NPR coverage on Dept. of Education earnings test: https://www.npr.org/2026/06/30/nx-s1-5835631/turner-camhi-do-no-harm-college-loans Paul Graham "cheating chart" post: https://x.com/paulg/status/2075031014628311236 Off Duty Recommendations: "Lioness" on Paramount+: https://www.youtube.com/watch?v=jNRQ0PR4a8U "Mayor of Kingstown" on Paramount+: https://www.youtube.com/watch?v=VkQzvwxOp0s "Human Vapor" on Netflix: https://www.youtube.com/watch?v=7xe6dRKVAb8 "Sugar" on Apple TV+: https://www.youtube.com/watch?v=twvPGxuEOEA "Wind River" (now on Netflix): https://www.youtube.com/watch?v=CZgN0dpFoaE "Not Fade Away" by Peter Barton & Laurence Shames: https://www.amazon.com/Not-Fade-Away-Short-Lived/dp/1579546889 Timestamps: 0:00 Jason's ongoing World Tour 1:43 Danny Bernstein joins live from Reservoir Farms 3:38 What is "specialty crop agriculture" 5:15 Why strawberries are the "white whale" of AgTech 6:55 The labor crisis in farming 9:20 Why AgTech never scaled 9:51 NetSuite - For the first time, you can try NetSuite Next for free. If your revenues are at least in the seven figures, go to https://NetSuite.ai/TWIST 10:51 Inside Reservoir's business model 17:44 Agriculture as national security 20:09 Squarespace - Turn your idea into a beautiful website! Go to https://www.squarespace.com/twist for a free trial. When you're ready to launch, use offer code TWIST to save 10% off your first purchase of a website or domain. 22:15 Did AI beat a designer at design? 30:56 YSecurity - The on-demand security team for startups. Need enterprise-grade security without hiring a $400k CISO? YSecurity gives you 40+ expert engineers, matched to exactly what you need, by the hour, with your first six hours completely free. Go to https://YSecurity.io/TWIST 37:23 The "do no harm" college earnings test 43:27 Brown University students cheated on their midterms 52:22 Lon's streaming recommendations 59:46 Jason's new snake grabber   Subscribe to the TWiST500 newsletter: https://ticker.thisweekinstartups.com Check out the TWIST500: https://www.twist500.com Subscribe to This Week in Startups on Apple: https://rb.gy/v19fcp   Follow Lon: X: https://x.com/lons   Follow Alex: X: https://x.com/alex LinkedIn: ⁠https://www.linkedin.com/in/alexwilhelm   Follow Jason: X: https://twitter.com/Jason LinkedIn: https://www.linkedin.com/in/jasoncalacanis   Check out all our partner offers: https://partners.launch.co/   Great TWIST interviews: Will Guidara, Eoghan McCabe, Steve Huffman, Brian Chesky, Bob Moesta, Aaron Levie, Sophia Amoruso, Reid Hoffman, Frank Slootman, Billy McFarland   Check out Jason's suite of newsletters: https://substack.com/@calacanis   Follow TWiST: Twitter: https://twitter.com/TWiStartups YouTube: https://www.youtube.com/thisweekin Instagram: https://www.instagram.com/thisweekinstartups TikTok: https://www.tiktok.com/@thisweekinstartups Substack: https://twistartups.substack.com

This Week In Startups is made possible by: Digital Ocean - do.co/twist Agree.com - agree.com Every.io - every.io.   Today's show: How many startups matter in tech? Fewer than you think. That's why venture capitalists are tripping over themselves to get onto their cap tables, no matter the cost. Why? Footwork's Nikhil Basu Trivedi argues that the Valley has never been more "power-law-pilled" than it is today. Basu Trivedi joined Cendana Capital's Michael Kim and TWiST's Alex Wilhelm to go deep on secondary markets, the state of startup M&A, why the SaaSpocalypse may be temporary, and what could trigger a retrenchment of the AI trade. It's Wednesday, so it's time for our venture capital roundtable to go deep on how VCs are investing today, and where on the horizon they have their eyes fixed! Guest links: Nikhil Basu Trivedi https://x.com/nbt Footwork https://www.footwork.vc/ Michael Kim https://x.com/MKRocks Cendana Capital https://www.cendanacapital.com/ Show links: The USVC-Anduril blowup https://x.com/ankurnagpal/status/2072701195714531398 Kline Hill Cendana Partners https://www.secondariesinvestor.com/kline-hill-and-cendana-raise-400m-for-second-vc-secondaries-fund/ GPTZero's exit https://gptzero.me/news/preserving-whats-human/ Salesforce buys Fin https://www.salesforce.com/news/press-releases/2026/06/15/salesforce-signs-definitive-agreement-to-acquire-fin/ Vercel buys Better Auth https://vercel.com/blog/vercel-acquires-better-auth Figma buys Bud https://techcrunch.com/2026/07/07/figma-acquires-team-behind-a-vibe-coding-app/ Protoge https://withprotege.ai/ Windborne https://windbornesystems.com/ Etched https://www.etched.com/ Lovable's reported raise https://sifted.eu/articles/lovable-300m-13-2bn-valuation Josh Browder https://x.com/Joshuabrowder   Timestamps: 0:00 Introduction: Nikhil Basu Trivedi (Footwork) & Michael Kim (Cendana Capital) 1:59 The Anduril vs. USVC secondary market blowup 4:08 Why Silicon Valley is 'power-law-pilled' 8:23 Plaud: If your work depends on conversations — interviews, meetings, calls — you need a Plaud NotePin. You can check it out at https://Plaud.ai/twist and use code TWIST for 10% off! 9:37 Information asymmetry in the secondary markets 9:45 Every.io — For all of your incorporation, banking, payroll, benefits, accounting, taxes or other back-office administration needs, visit https://every.io 15:02 Is SPV fraud smoke or fire? 16:32 Superhuman acquires GPTZero 19:54 Agree.com - Stop chasing invoices and automate your entire contract-to-cash stack. Go to https://agree.com and tell them Jason sent you to get 50% off for life! 21:10 The M&A wave 27:19 The SaaSpocalypse debate 29:59 DigitalOcean - Head to https://do.co/twist to start building on DigitalOcean's AI-Native Cloud today — and cut your AI workload costs by up to 50%. 30:44 Data's moment in the energy → compute → data loop 35:01 Where will AI value accrue? 40:04 What could cause an AI correction? 42:17 Why some companies are "too big to miss" 46:23 China's possible open-weight model ban 53:28 Young founders: Etched, Thiel Fellows, Z Fellows, Neo 55:33 Portfolio spotlight: WindBorne's weather balloons and data moat 58:47 Michael's favorite fund manager: Josh Browder Subscribe to the TWiST500 newsletter: https://ticker.thisweekinstartups.com Check out the TWIST500: https://www.twist500.com Subscribe to This Week in Startups on Apple: https://rb.gy/v19fcp   Follow Lon: X: https://x.com/lons   Follow Alex: X: https://x.com/alex LinkedIn: ⁠https://www.linkedin.com/in/alexwilhelm   Follow Jason: X: https://twitter.com/Jason LinkedIn: https://www.linkedin.com/in/jasoncalacanis   Check out all our partner offers: https://partners.launch.co/   Great TWIST interviews: Will Guidara, Eoghan McCabe, Steve Huffman, Brian Chesky, Bob Moesta, Aaron Levie, Sophia Amoruso, Reid Hoffman, Frank Slootman, Billy McFarland   Check out Jason's suite of newsletters: https://substack.com/@calacanis   Follow TWiST: Twitter: https://twitter.com/TWiStartups YouTube: https://www.youtube.com/thisweekin Instagram: https://www.instagram.com/thisweekinstartups TikTok: https://www.tiktok.com/@thisweekinstartups Substack: https://twistartups.substack.com

This Week In Startups is made possible by: Northwest Registered Agent https://northwestregisteredagent.com/twist Vanta https://www.vanta.com/twist Sentry https://sentry.io/twist Today's show: *There are $100 trillion in global assets sitting on top of what Hanover Park co-founder/CEO Chris Hladczuk calls "human duct tape": armies of accountants in offices patching together work from various legacy tools (QuickBooks, Excel) that are holding funds' own data hostage. Can all of this be replaced with AI? Find out how their startup went from overseeing $1B to $20B in assets in just 15 months. PLUS, we flash back to March 2020, when Jason and Figma co-founder/CEO Dylan Field broke down the design tool's initial go-to-market strategy, made some WILDLY inaccurate COVID predictions, and considered anxiety about "SaaS burnout" years before the category went full apocalyptic. Guests: Chris Hladczuk on X: https://x.com/chrishlad Hanover Park: https://www.hanoverpark.com/ Dylan Field: https://x.com/zoink Figma: https://www.figma.com/ Relevant Links: Turner Novak on X: https://x.com/TurnerNovak Banana Capital: https://www.bananacapital.vc/ Emergence Capital: https://www.emcap.com/ Lux Capital: https://www.luxcapital.com/ Susa Ventures: https://susaventures.com/ Bill.com: https://www.bill.com/ METR: https://metr.org/ Granola AI note taker: https://www.granola.ai/ Vanta: https://www.vanta.com/ Foo Camp on YouTube: https://www.youtube.com/c/foocamp TechCrunch Mahalo coverage: https://techcrunch.com/2014/01/27/inside-mobile-news-launch/ Timestamps: 0:00 Hanover Park & the fund admin problem 3:32 Why funds outsource instead of building 5:44 Why fund accounting is so complex 10:38 The "one-click migration" goal 10:48 Northwest Registered Agent - Get more when you start your business with Northwest. In 10 clicks and 10 minutes, you can form your company and walk away with a real business identity — Learn more at https://northwestregisteredagent.com/twist 13:51 Context vs. intelligence gaps 16:11 No PMs, No Designers 20:46 Vanta - Get $1000 off your SOC 2 at https://www.vanta.com/twist 23:18 This is a $100T opportunity 25:36 Flashback w/ Dylan Field of Figma 29:15 Sentry - Your team should be focused on shipping features — not chasing down bugs. New users can get $240 in free credits when they go to https://sentry.io/twist and use the code TWIST 31:22 Pre-AI enterprise security worries 36:30 SaaS overload and SaaS burnout 37:30 The evolution of Figma pricing 42:38 The rise and fall of Mahalo dot com 51:38 The work-from-home revolution begins Subscribe to the TWiST500 newsletter: https://ticker.thisweekinstartups.com Check out the TWIST500: https://www.twist500.com Subscribe to This Week in Startups on Apple: https://rb.gy/v19fcp Follow Lon: X: https://x.com/lons Follow Alex: X: https://x.com/alex LinkedIn: ⁠https://www.linkedin.com/in/alexwilhelm Follow Jason: X: https://twitter.com/Jason LinkedIn: https://www.linkedin.com/in/jasoncalacanis Thank you to our partners: (0:00) PARTNER - AD BLURB (0:00) PARTNER - AD BLURB (0:00) PARTNER - AD BLURB Check out all our partner offers: https://partners.launch.co/ Great TWIST interviews: Will Guidara, Eoghan McCabe, Steve Huffman, Brian Chesky, Bob Moesta, Aaron Levie, Sophia Amoruso, Reid Hoffman, Frank Slootman, Billy McFarland Check out Jason's suite of newsletters: https://substack.com/@calacanis Follow TWiST: Twitter: https://twitter.com/TWiStartups YouTube: https://www.youtube.com/thisweekin Instagram: https://www.instagram.com/thisweekinstartups TikTok: https://www.tiktok.com/@thisweekinstartups Substack: https://twistartups.substack.com