POPULARITY
Solving Poker and Diplomacy, Debating RL+Reasoning with Ilya, what's *wrong* with the System 1/2 analogy, and where Test-Time Compute hits a wall Timestamps 00:00 Intro – Diplomacy, Cicero & World Championship 02:00 Reverse Centaur: How AI Improved Noam's Human Play 05:00 Turing Test Failures in Chat: Hallucinations & Steerability 07:30 Reasoning Models & Fast vs. Slow Thinking Paradigm 11:00 System 1 vs. System 2 in Visual Tasks (GeoGuessr, Tic-Tac-Toe) 14:00 The Deep Research Existence Proof for Unverifiable Domains 17:30 Harnesses, Tool Use, and Fragility in AI Agents 21:00 The Case Against Over-Reliance on Scaffolds and Routers 24:00 Reinforcement Fine-Tuning and Long-Term Model Adaptability 28:00 Ilya's Bet on Reasoning and the O-Series Breakthrough 34:00 Noam's Dev Stack: Codex, Windsurf & AGI Moments 38:00 Building Better AI Developers: Memory, Reuse, and PR Reviews 41:00 Multi-Agent Intelligence and the “AI Civilization” Hypothesis 44:30 Implicit World Models and Theory of Mind Through Scaling 48:00 Why Self-Play Breaks Down Beyond Go and Chess 54:00 Designing Better Benchmarks for Fuzzy Tasks 57:30 The Real Limits of Test-Time Compute: Cost vs. Time 1:00:30 Data Efficiency Gaps Between Humans and LLMs 1:03:00 Training Pipeline: Pretraining, Midtraining, Posttraining 1:05:00 Games as Research Proving Grounds: Poker, MTG, Stratego 1:10:00 Closing Thoughts – Five-Year View and Open Research Directions Chapters 00:00:00 Intro & Guest Welcome 00:00:33 Diplomacy AI & Cicero Insights 00:03:49 AI Safety, Language Models, and Steerability 00:05:23 O Series Models: Progress and Benchmarks 00:08:53 Reasoning Paradigm: Thinking Fast and Slow in AI 00:14:02 Design Questions: Harnesses, Tools, and Test Time Compute 00:20:32 Reinforcement Fine-tuning & Model Specialization 00:21:52 The Rise of Reasoning Models at OpenAI 00:29:33 Data Efficiency in Machine Learning 00:33:21 Coding & AI: Codex, Workflows, and Developer Experience 00:41:38 Multi-Agent AI: Collaboration, Competition, and Civilization 00:45:14 Poker, Diplomacy & Exploitative vs. Optimal AI Strategy 00:52:11 World Models, Multi-Agent Learning, and Self-Play 00:58:50 Generative Media: Image & Video Models 01:00:44 Robotics: Humanoids, Iteration Speed, and Embodiment 01:04:25 Rapid Fire: Research Practices, Benchmarks, and AI Progress 01:14:19 Games, Imperfect Information, and AI Research Directions
⭐ Join Rental Property Mastery, my community of rental investors on their way to financial freedom: https://coachcarson.com/rpm
Ever heard of model collapse? It occurs when AI models are trained on the outputs of previous AI models. And it's not pretty.I recently recorded an in-depth session for Rob Walling's SaaS Launchpad. This module is called AI in SaaS, and I share over a dozen pitfalls, unobvious risks, and hard-earned insights from my own AI SaaS building journey. The course was already great, but... you know... ;) Use the code ARVID150 to get $150 off until June 8th.The blog post: https://thebootstrappedfounder.com/building-ai-businesses-without-breaking-the-internet/ The podcast episode: https://tbf.fm/episodes/building-ai-businesses-without-breaking-the-internetCheck out Podscan, the Podcast database that transcribes every podcast episode out there minutes after it gets released: https://podscan.fmSend me a voicemail on Podline: https://podline.fm/arvidYou'll find my weekly article on my blog: https://thebootstrappedfounder.comPodcast: https://thebootstrappedfounder.com/podcastNewsletter: https://thebootstrappedfounder.com/newsletterMy book Zero to Sold: https://zerotosold.com/My book The Embedded Entrepreneur: https://embeddedentrepreneur.com/My course Find Your Following: https://findyourfollowing.comHere are a few tools I use. Using my affiliate links will support my work at no additional cost to you.- Notion (which I use to organize, write, coordinate, and archive my podcast + newsletter): https://affiliate.notion.so/465mv1536drx- Riverside.fm (that's what I recorded this episode with): https://riverside.fm/?via=arvid- TweetHunter (for speedy scheduling and writing Tweets): http://tweethunter.io/?via=arvid- HypeFury (for massive Twitter analytics and scheduling): https://hypefury.com/?via=arvid60- AudioPen (for taking voice notes and getting amazing summaries): https://audiopen.ai/?aff=PXErZ- Descript (for word-based video editing, subtitles, and clips): https://www.descript.com/?lmref=3cf39Q- ConvertKit (for email lists, newsletters, even finding sponsors): https://convertkit.com?lmref=bN9CZw
In an otherwise heavy week packed with Microsoft Build, Google I/O, and OpenAI io, the worst kept secret in biglab land was the launch of Claude 4, particularly the triumphant return of Opus, which many had been clamoring for. We will leave the specific Claude 4 recap to AINews, however we think that both Gemini's progress on Deep Think this week and Claude 4 represent the next frontier of progress on inference time compute/reasoning (at last until GPT5 ships this summer). Will Brown's talk at AIE NYC and open source work on verifiers have made him one of the most prominent voices able to publicly discuss (aka without the vaguepoasting LoRA they put on you when you join a biglab) the current state of the art in reasoning models and where current SOTA research directions lead. We discussed his latest paper on Reinforcing Multi-Turn Reasoning in LLM Agents via Turn-Level Credit Assignment and he has previewed his AIEWF talk on Agentic RL for those with the temerity to power thru bad meetup audio. Chapters 00:00 Introduction and Episode Overview 02:01 Discussion on Cloud 4 and its Features 04:31 Reasoning and Tool Use in AI Models 07:01 Extended Thinking in Claude and Model Differences 09:31 Speculation on Claude's Extended Thinking 11:01 Challenges and Controversies in AI Model Training 13:31 Technical Highlights and Code Trustworthiness 16:01 Token Costs and Incentives in AI Models 18:31 Thinking Budgets and AI Effort 21:01 Safety and Ethics in AI Model Development 23:31 Anthropic's Approach to AI Safety 26:01 LLM Arena and Evaluation Challenges 28:31 Developing Taste and Direction in AI Research 31:01 Recent Research and Multi-Turn RL 33:31 Tools and Incentives in AI Model Development 36:01 Challenges in Evaluating AI Model Outputs 38:31 Model-Based Rewards and Future Directions 41:01 Wrap-up and Future Plans
Tool Use and Model Context Protocol (MCP) Notes and resources at ocdevel.com/mlg/mla-24 Try a walking desk to stay healthy while you study or work! Tool Use in Vibe Coding Agents File Operations: Agents can read, edit, and search files using sophisticated regular expressions. Executable Commands: They can recommend and perform installations like pip or npm installs, with user approval. Browser Integration: Allows agents to perform actions and verify outcomes through browser interactions. Model Context Protocol (MCP) Standardization: MCP was created by Anthropic to standardize how AI tools and agents communicate with each other and with external tools. Implementation: MCP Client: Converts AI agent requests into structured commands. MCP Server: Executes commands and sends structured responses back to the client. Local and Cloud Frameworks: Local (S-T-D-I-O MCP): Examples include utilizing Playwright for local browser automation and connecting to local databases like Postgres. Cloud (SSE MCP): SaaS providers offer cloud-hosted MCPs to enhance external integrations. Expanding AI Capabilities with MCP Servers Directories: Various directories exist listing MCP servers for diverse functions beyond programming. modelcontextprotocol/servers Use Cases: Automation Beyond Coding: Implementing MCPs that extend automation into non-programming tasks like sales, marketing, or personal project management. Creative Solutions: Encourages innovation in automating routine tasks by integrating diverse MCP functionalities. AI Tools in Machine Learning Automating ML Process: Auto ML and Feature Engineering: AI tools assist in transforming raw data, optimizing hyperparameters, and inventing new ML solutions. Pipeline Construction and Deployment: Facilitates the use of infrastructure as code for deploying ML models efficiently. Active Experimentation: Jupyter Integration Challenges: While integrations are possible, they often lag and may not support the latest models. Practical Strategies: Suggests alternating between Jupyter and traditional Python files to maximize tool efficiency. Conclusion Action Plan for ML Engineers: Setup structured folders and documentation to leverage AI tools effectively. Encourage systematic exploration of MCPs to enhance both direct programming tasks and associated workflows.
Send us a textSevag Sarkissian, VP of Growth Marketing is back to explore how real estate professionals can thrive in 2025 by embracing change, adopting AI, and shifting their mindset. Sevag shares his insights on staying competitive in a rapidly evolving market, the importance of awareness over just intelligence, and how tools like ChatGPT can elevate productivity and creativity. Packed with energy, wisdom, and practical takeaways, this episode is your motivational boost to work smarter, not harder and to use the tools at your fingertips to grow your business.
The AI Breakdown: Daily Artificial Intelligence News and Discussions
A discussion originally from the Tool Use podcast. See full episode: https://www.youtube.com/watch?v=-fDu52FzmJc // https://podcasts.apple.com/ca/podcast/will-ai-agents-be-your-automation-breakthrough-ft-nlw/id1773693853?i=1000693646455Brought to you by:KPMG – Go to www.kpmg.us/ai to learn more about how KPMG can help you drive value with our AI solutions.Vanta - Simplify compliance - https://vanta.com/nlwThe Agent Readiness Audit from Superintelligent - Go to https://besuper.ai/ to request your company's agent readiness score.The AI Daily Brief helps you understand the most important news and discussions in AI. Subscribe to the podcast version of The AI Daily Brief wherever you listen: https://pod.link/1680633614Subscribe to the newsletter: https://aidailybrief.beehiiv.com/Join our Discord: https://bit.ly/aibreakdown
Humans aren't the only species that use tools, many animals do. Our closest relatives, the great apes, have long been known for their intelligence and ability to manipulate objects in their environment. Among the great apes, orangutans stand out for their complex tool use, which may hold important clues about how early human technologies evolved. A recent Outlook article authored by Dr. Robert Shumaker and Dr. Christopher Martin, both of the Indianapolis Zoo, delves into the existing research into ways that orangutans use stone tools, exploring the cognitive processes behind this behavior and what it can tell us about human evolution.
This episode covers the extensive research and investigations into Bigfoot, including behavior, diet, and evidence collection. Robert Pryor shares insights from his long career, focusing on Bigfoot's ecosystem adaptations and unique biological traits. Viewers learn about tracking techniques, signs like footprints and scat, and the use of tree markers for territory. Also discussed are vocalizations and sophisticated hunting methods. The episode addresses scientific skepticism and DNA research challenges, highlighting significant studies and controversies, including collaborations with Oxford University and the handling of DNA samples. It concludes with discussions on why finding Sasquatch bones is rare, and the complexities in verifying their existence.Listen To Backwoods Horror Stories Get Our FREE NewsletterGet Brian's Book Sasquatch Unleashed The Truth Behind The LegendLeave Us A VoicemailVisit Our WebsiteSupport Our SponsorsVisit Hangar 1 Publishing 00:00 Introduction to Robert Pryor 00:37 Welcome to the 2024 Ozark Mountain Victim Conference 00:54 The Journey of Fryder Exploration 03:09 Bigfoot Ecology and Behavior 03:37 Bigfoot's Diet and Foraging Habits 09:06 Tool Use and Hunting Techniques 12:04 Tracking and Footprints 28:15 Bigfoot Vocalizations 32:33 Feel the Rumble: Stunt Movers and Sound Effects 33:18 Foraging and Hunting Techniques 33:57 Sound Traps and Ambush Strategies 35:01 Occupation and Territorial Behavior 36:44 Dark Side of Behavior: Goat Kills 40:03 Scientific Viewpoint and DNA Evidence 47:13 Court Battles and Scientific Controversies 52:55 Q&A Session: DNA Studies and Missing BonesBecome a supporter of this podcast: https://www.spreaker.com/podcast/sasquatch-odyssey--4839697/support.
Guest: Yaniv Zaid Guest Bio: Known to the world as "Doctor Persuasion", economist and attorney Dr. Yaniv Zaid acts as a business consultant to government departments, private firms, and public organizations. He holds a PhD in law and utilizes his rich knowledge and experience to help others achieve success. Dr. Zaid is recognized worldwide as an expert in the fields of public speaking, marketing, sales, negotiation, and persuasion, and placed 3rd in the 2003 world ranking of public speakers. Following 20 years of international success, Dr. Zaid is the author of 11 best seller books - including "Public Speaking", "Creative Marketing" and "Sales Bible" - and is often invited to lecture and consult around the world. Guest Links: The 21st Century Sales Bible: Mastering the 10 Commandments of Marketing, Negotiation & Persuasion yaniv@yanivzaid.com Key Points: The Gap in Education · Many individuals lack training in public speaking, marketing, and sales despite these skills being essential in professional settings. · Law Schools often do not teach practical skills like marketing oneself, courtroom speaking, or legal writing. Similarly, economics programs often overlook sales and marketing. · Bridging the Gap: The speaker's mission is to educate others on essential skills such as marketing, sales, public speaking, and negotiation to fill the educational void. The Nature of Persuasion · Positive Influence: Persuasion is about positively changing people's minds and encouraging them to choose your offering over others. · Authenticity in Sales: Salespeople must genuinely believe in their products and show authenticity to build trust with clients. Target Audience · Focus on understanding and targeting the audience that genuinely needs your product to maximize effectiveness in persuasion. Sales Team Management · Hiring individuals who love sales and understand the product is crucial for an effective sales team. Salespeople should believe in the product they are selling for authenticity and effectiveness. · Encourage your sales team to improve their skills and knowledge. This can lead to better engagement with potential clients. Do's and Don'ts of Persuasion · Do Sell the Problem: Highlight the problem first; the bigger the problem, the more value clients place on the solution. · Don't Focus on Subjectivity: Avoid claiming that your product is great without evidence; instead, use social proof and objective evidence to support your claims. · Authenticity Over Perfection: Avoid branding yourself as a flawless expert. Instead, relate to your audience by sharing both successes and failures. This builds trust and makes you more relatable. Long-term Relationships · Sales should be based on mutual benefit and understanding, aiming for long-term client relationships rather than short-term gains. Storytelling as a Tool · Use stories to convey your message and include social proof within them. Stories engage the audience's emotions and can be more persuasive than just stating facts. About Salesology®: Conversations with Sales Leaders Download your free gift, The Salesology® Vault. The vault is packed full of free gifts from sales leaders, sales experts, marketing gurus and revenue generation experts. Download your free gift, 81 Tools to Grow Your Sales & Your Business Faster, More Easily & More Profitably. Save hours of work tracking down the right prospecting and sales resources and/or digital tools that every business owner and salesperson needs. If you are a business owner or sales manager with an underperforming sales team, let's talk. Click here to schedule a time. Please, subscribe to Salesology®: Conversations with Sales Leaders so that you don't miss a single episode, and while you're at it, won't you take a moment to write a short review and rate our show? It would be greatly appreciated! To learn more about our previous guests, listen to past episodes, and get to know your host, go to https://podcast.gosalesology.com/ and connect on LinkedIn and follow us on Facebook and Twitter and check out our website at http://www.gosalesology.
In this unique episode, Brian presents a panel discussion recorded after the Texas Bigfoot Conference in Fort Worth. Moderated by Mike Mays, notable speakers such as Josh Gates, Kathy Strain, Alton Higgins, Lyle Blackburn, Bill Munns, Daryl Collier, Jeff Meldrum, and Brian himself answer audience-submitted questions. Topics range from personal experiences with Bigfoot, reasons for the lack of physical evidence, technological challenges in research, to ethical considerations regarding capturing or killing a specimen. The panelists also share their most frightening encounters in the field, offering an in-depth look into the ongoing search for the elusive Sasquatch.Listen To Backwoods Horror Stories Get Our FREE NewsletterGet Brian's Book Sasquatch Unleashed The Truth Behind The LegendLeave Us A VoicemailVisit Our WebsiteSupport Our SponsorsVisit Hangar 1 Publishing 00:00 Introduction and Panel Overview 00:36 Audience Questions Begin 01:00 Destination Truth's Most Threatening Moments 01:44 Challenges in Proving Sasquatch's Existence 02:35 The Mystery of Physical Evidence 05:01 Technological and Environmental Hurdles 16:36 Hoaxes and Misidentifications 19:46 Tool Use and Behavioral Observations 21:30 Revisiting the Cabin Incident 25:51 The Patterson-Gimlin Footage Debate 31:11 The Necessity of a Type Specimen 32:57 Debate on DNA as Type Specimen 33:58 Handling a Sasquatch Body 36:07 Challenges of DNA Evidence 39:17 Conservation and Specimen Collection 41:50 Tranquilizers vs. Lethal Methods 47:05 Technology in Cryptid Research 56:14 Scariest Encounters in the FieldBecome a supporter of this podcast: https://www.spreaker.com/podcast/sasquatch-odyssey--4839697/support.
The discovery of Lucy in 1974 gave insight into early hominin body form but lacked hand bones. Subsequent findings revealed Australopithecus afarensis hand morphology. New discoveries and research since then, including associated hand skeletons and archaeological evidence, have enhanced our understanding of hominin hand evolution and tool use. These advancements inform us about Lucy's tool-related abilities and dexterity, reshaping our interpretation of early human behavior. Series: "CARTA - Center for Academic Research and Training in Anthropogeny" [Humanities] [Science] [Show ID: 39821]
The discovery of Lucy in 1974 gave insight into early hominin body form but lacked hand bones. Subsequent findings revealed Australopithecus afarensis hand morphology. New discoveries and research since then, including associated hand skeletons and archaeological evidence, have enhanced our understanding of hominin hand evolution and tool use. These advancements inform us about Lucy's tool-related abilities and dexterity, reshaping our interpretation of early human behavior. Series: "CARTA - Center for Academic Research and Training in Anthropogeny" [Humanities] [Science] [Show ID: 39821]
CARTA - Center for Academic Research and Training in Anthropogeny (Video)
The discovery of Lucy in 1974 gave insight into early hominin body form but lacked hand bones. Subsequent findings revealed Australopithecus afarensis hand morphology. New discoveries and research since then, including associated hand skeletons and archaeological evidence, have enhanced our understanding of hominin hand evolution and tool use. These advancements inform us about Lucy's tool-related abilities and dexterity, reshaping our interpretation of early human behavior. Series: "CARTA - Center for Academic Research and Training in Anthropogeny" [Humanities] [Science] [Show ID: 39821]
The discovery of Lucy in 1974 gave insight into early hominin body form but lacked hand bones. Subsequent findings revealed Australopithecus afarensis hand morphology. New discoveries and research since then, including associated hand skeletons and archaeological evidence, have enhanced our understanding of hominin hand evolution and tool use. These advancements inform us about Lucy's tool-related abilities and dexterity, reshaping our interpretation of early human behavior. Series: "CARTA - Center for Academic Research and Training in Anthropogeny" [Humanities] [Science] [Show ID: 39821]
The discovery of Lucy in 1974 gave insight into early hominin body form but lacked hand bones. Subsequent findings revealed Australopithecus afarensis hand morphology. New discoveries and research since then, including associated hand skeletons and archaeological evidence, have enhanced our understanding of hominin hand evolution and tool use. These advancements inform us about Lucy's tool-related abilities and dexterity, reshaping our interpretation of early human behavior. Series: "CARTA - Center for Academic Research and Training in Anthropogeny" [Humanities] [Science] [Show ID: 39821]
Nathan hosts Riley Goodside, the world's first staff prompt engineer at Scale AI, to discuss the evolution of prompt engineering. In this episode of The Cognitive Revolution, we explore how language models have progressed, making prompt engineering more like programming than poetry. Discover insights on enterprise AI applications, best practices for pushing LLMs to their limits, and the future of AI automation. Apply to join over 400 founders and execs in the Turpentine Network: https://hmplogxqz0y.typeform.com/to/JCkphVqj RECOMMENDED PODCAST: Complex Systems Patrick McKenzie (@patio11) talks to experts who understand the complicated but not unknowable systems we rely on. You might be surprised at how quickly Patrick and his guests can put you in the top 1% of understanding for stock trading, tech hiring, and more. Spotify: https://open.spotify.com/show/3Mos4VE3figVXleHDqfXOH Apple: https://podcasts.apple.com/us/podcast/complex-systems-with-patrick-mckenzie-patio11/id1753399812 SPONSORS: Building an enterprise-ready SaaS app? WorkOS has got you covered with easy-to-integrate APIs for SAML, SCIM, and more. Join top startups like Vercel, Perplexity, Jasper & Webflow in powering your app with WorkOS. Enjoy a free tier for up to 1M users! Start now at https://bit.ly/WorkOS-TCR Oracle Cloud Infrastructure (OCI) is a single platform for your infrastructure, database, application development, and AI needs. OCI has four to eight times the bandwidth of other clouds; offers one consistent price, and nobody does data better than Oracle. If you want to do more and spend less, take a free test drive of OCI at https://oracle.com/cognitive The Brave search API can be used to assemble a data set to train your AI models and help with retrieval augmentation at the time of inference. All while remaining affordable with developer first pricing, integrating the Brave search API into your workflow translates to more ethical data sourcing and more human representative data sets. Try the Brave search API for free for up to 2000 queries per month at https://bit.ly/BraveTCR Omneky is an omnichannel creative generation platform that lets you launch hundreds of thousands of ad iterations that actually work customized across all platforms, with a click of a button. Omneky combines generative AI and real-time advertising data. Mention "Cog Rev" for 10% off https://www.omneky.com/ Head to Squad to access global engineering without the headache and at a fraction of the cost: head to https://choosesquad.com/ and mention “Turpentine” to skip the waitlist. CHAPTERS: (00:00:00) About the Show (00:00:23) Sponsor: WorkOS (00:01:24) Introduction (00:06:23) LLMs using LLMs (00:09:38) Tool Use (00:11:06) How to manage the breadth of the task (00:14:51) Prompt engineering (00:16:24) Sponsors: Oracle | Brave (00:18:28) The importance of explicit reasoning (00:21:16) The importance of breaking down tasks (00:26:49) Multitasking fine-tuning (00:31:49) Sponsors: Omneky | Squad (00:33:36) Best models for fine-tuning (00:36:41) The Platonic Representation Hypothesis (00:42:02) How close are we to AGI? (00:45:44) How do you know if youre being too ambitious? (00:51:18) Best practices for generating good output (00:54:33) Backfills and synthetic transformations (00:56:59) Prompt engineering (01:05:54) AGI, modalities, and the limits of training (01:11:38) Compute thresholds (01:13:02) Jailbreaking models (01:16:09) Open-source models (01:20:08) Solving the ARC Challenge (01:23:20) How to Demonstrate Prompt Engineering Skills (01:25:27) Outro
If you see this in time, join our emergency LLM paper club on the Llama 3 paper!For everyone else, join our special AI in Action club on the Latent Space Discord for a special feature with the Cursor cofounders on Composer, their newest coding agent!Today, Meta is officially releasing the largest and most capable open model to date, Llama3-405B, a dense transformer trained on 15T tokens that beats GPT-4 on all major benchmarks:The 8B and 70B models from the April Llama 3 release have also received serious spec bumps, warranting the new label of Llama 3.1.If you are curious about the infra / hardware side, go check out our episode with Soumith Chintala, one of the AI infra leads at Meta. Today we have Thomas Scialom, who led Llama2 and now Llama3 post-training, so we spent most of our time on pre-training (synthetic data, data pipelines, scaling laws, etc) and post-training (RLHF vs instruction tuning, evals, tool calling).Synthetic data is all you needLlama3 was trained on 15T tokens, 7x more than Llama2 and with 4 times as much code and 30 different languages represented. But as Thomas beautifully put it:“My intuition is that the web is full of s**t in terms of text, and training on those tokens is a waste of compute.” “Llama 3 post-training doesn't have any human written answers there basically… It's just leveraging pure synthetic data from Llama 2.”While it is well speculated that the 8B and 70B were "offline distillations" of the 405B, there are a good deal more synthetic data elements to Llama 3.1 than the expected. The paper explicitly calls out:* SFT for Code: 3 approaches for synthetic data for the 405B bootstrapping itself with code execution feedback, programming language translation, and docs backtranslation.* SFT for Math: The Llama 3 paper credits the Let's Verify Step By Step authors, who we interviewed at ICLR:* SFT for Multilinguality: "To collect higher quality human annotations in non-English languages, we train a multilingual expert by branching off the pre-training run and continuing to pre-train on a data mix that consists of 90% multilingualtokens."* SFT for Long Context: "It is largely impractical to get humans to annotate such examples due to the tedious and time-consuming nature of reading lengthy contexts, so we predominantly rely on synthetic data to fill this gap. We use earlier versions of Llama 3 to generate synthetic data based on the key long-context use-cases: (possibly multi-turn) question-answering, summarization for long documents, and reasoning over code repositories, and describe them in greater detail below"* SFT for Tool Use: trained for Brave Search, Wolfram Alpha, and a Python Interpreter (a special new ipython role) for single, nested, parallel, and multiturn function calling.* RLHF: DPO preference data was used extensively on Llama 2 generations. This is something we partially covered in RLHF 201: humans are often better at judging between two options (i.e. which of two poems they prefer) than creating one (writing one from scratch). Similarly, models might not be great at creating text but they can be good at classifying their quality.Last but not least, Llama 3.1 received a license update explicitly allowing its use for synthetic data generation.Llama2 was also used as a classifier for all pre-training data that went into the model. It both labelled it by quality so that bad tokens were removed, but also used type (i.e. science, law, politics) to achieve a balanced data mix. Tokenizer size mattersThe tokens vocab of a model is the collection of all tokens that the model uses. Llama2 had a 34,000 tokens vocab, GPT-4 has 100,000, and 4o went up to 200,000. Llama3 went up 4x to 128,000 tokens. You can find the GPT-4 vocab list on Github.This is something that people gloss over, but there are many reason why a large vocab matters:* More tokens allow it to represent more concepts, and then be better at understanding the nuances.* The larger the tokenizer, the less tokens you need for the same amount of text, extending the perceived context size. In Llama3's case, that's ~30% more text due to the tokenizer upgrade. * With the same amount of compute you can train more knowledge into the model as you need fewer steps.The smaller the model, the larger the impact that the tokenizer size will have on it. You can listen at 55:24 for a deeper explanation.Dense models = 1 Expert MoEsMany people on X asked “why not MoE?”, and Thomas' answer was pretty clever: dense models are just MoEs with 1 expert :)[00:28:06]: I heard that question a lot, different aspects there. Why not MoE in the future? The other thing is, I think a dense model is just one specific variation of the model for an hyperparameter for an MOE with basically one expert. So it's just an hyperparameter we haven't optimized a lot yet, but we have some stuff ongoing and that's an hyperparameter we'll explore in the future.Basically… wait and see!Llama4Meta already started training Llama4 in June, and it sounds like one of the big focuses will be around agents. Thomas was one of the authors behind GAIA (listen to our interview with Thomas in our ICLR recap) and has been working on agent tooling for a while with things like Toolformer. Current models have “a gap of intelligence” when it comes to agentic workflows, as they are unable to plan without the user relying on prompting techniques and loops like ReAct, Chain of Thought, or frameworks like Autogen and Crew. That may be fixed soon?
In this episode the North American Wood Ape Conservancy (NAWAC) shares their long-term research endeavors at Area X in Eastern Oklahoma. Hosted by NAWAC Chairman Alton Higgins and field anthropologist Kathy Strain, the episode explores firsthand observations and experiences, including rock-throwing, vocalizations, tree manipulations, and potential sightings of wood apes. Listeners get insights into the behavioral patterns and social structures of these elusive creatures, drawing comparisons to other primates. The discussion highlights the scientific approach and meticulous documentation by the NAWAC team, emphasizing the need for conservation efforts to protect the apes' habitat.Get Our FREE NewsletterGet Brian's Book Sasquatch Unleashed The Truth Behind The LegendLeave Us A VoicemailVisit Our WebsiteSupport Our SponsorsVisit Hangar 1 Publishing00:00 Introduction to Area X 00:23 Operation Relentless: Current Field Study 00:43 Interview with Alton Higgins and Kathy Strain 02:18 Rock Throwing and Other Ape Activities 03:54 Tool Use and Behavioral Observations 11:36 Challenges and Unknowns in Research 15:14 The Importance of Data Collection 17:14 Future of Bigfoot Research and Conservation 22:26 Personal Experiences and Sightings 41:39 Campfire Conversations: Kathy and Monica's Nighttime Encounters 42:34 Bigfoot Activity: Rock Throws and Primate Displays 43:46 Mysterious Sounds and Eye Shine 46:57 Tree Push and Rock Throws: A Night of Intense Activity 52:32 Daytime Sightings and Rock Throws 53:55 Interview with Daryl Collier: Analyzing Field Notes 54:31 Rock Throws and Environmental Changes 01:02:59 Sightings and Behavioral Observations 01:11:55 Mysterious Rock Barrage 01:12:43 Frustrations in the Field 01:13:57 Rare Sightings and Mistakes 01:15:55 Collecting Physical Evidence 01:17:11 Targeted Rock Throws 01:19:39 Tree Incidents and Theories 01:25:32 Avoiding Cameras and Human Perception 01:35:17 The Urgency of Discovery 01:40:12 Concluding Thoughts and Future PlansBecome a supporter of this podcast: https://www.spreaker.com/podcast/sasquatch-odyssey--4839697/support.
Timestamps: 0:00 don't forget, now! 0:12 Claude, Perplexity, ChatGPT updates 1:58 TikTok working on U.S. algorithm 3:10 Google addresses Search leak 5:30 QUICK BITS INTRO 5:36 Spotify giving Car Thing refunds 6:14 Google Cloud explains account deletion 6:58 Internet Archive DDoS'd 7:26 Showrunner, the 'Netflix of AI' 8:03 next-level tank simulator News Sources: https://lmg.gg/mZ7Ty Learn more about your ad choices. Visit megaphone.fm/adchoices
With groundbreaking news from Anthropic, Apple, Microsoft, and OpenAI, it's a whirlwind of innovation and strategic maneuvers in the tech world.In this week's episode of Leveraging AI, host Isar Meitis dives into the latest AI developments that are set to revolutionize the industry. From Anthropic's new API capabilities to Apple's upcoming AI integrations, we've got all the critical updates you need to stay ahead.Recommendation: Whether you're a tech enthusiast or a business professional, understanding these shifts can provide a competitive edge. This episode will guide you through the most significant announcements and their implications for the future of AI.In this session, you'll discover:Anthropic's new API feature, Tool Use or Function Calling, and its vast potential.What to expect from Apple's Worldwide Developer Conference regarding AI advancements.The implications of Apple's collaboration with OpenAI and potential partnerships with Google.The impact of OpenAI's recent licensing deals with News Corp and Vox Media.Insights into the controversies surrounding OpenAI's leadership and strategic direction.Microsoft's latest AI integrations in Excel, Telegram, and GitHub Copilot Plus.The upcoming AI-powered video editing features in Google Photos.By the end of this episode, you'll have a comprehensive understanding of the latest AI trends and how they might impact your business and career.About Leveraging AI The 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/ Free AI Consultation: https://multiplai.ai/book-a-call/ If 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!
Xochitl explains how sea otters use tools to help open prey, with the help of UCSC researchers we can understand both tooth diseases and tool use!Sponsored by the Division of Student Affairs and Success and produced by our student podcasters, UCSC Slugcast supports free expression of ideas. Please be aware that the views and opinions expressed by speakers are that of the speaker alone and their appearance on the program does not imply an endorsement of them or any entity they represent. Similarly, views and opinions of University employees or students are their own and do not necessarily reflect the views of Slugcast, the Division of Student Affairs and Success, or the University.
On this week's show: What happens when the body's own immune system attacks the brain, and how otters' use of tools expands their diet First on the show this week, when rogue antibodies attack the brain, patients can show bizarre symptoms—from extreme thirst, to sleep deprivation, to outright psychosis. Contributing Correspondent Richard Stone joins host Sarah Crespi to discuss the hunt for biomarkers and treatments for this cluster of autoimmune disorders that were once mistaken for schizophrenia or even demonic possession. Next on this episode, producer Katherine Irving talks with Chris Law, an evolutionary biologist at the University of Washington and the University of Texas at Austin, about how sea otters gain energy benefits (and dental benefits) when they use tools to tackle tougher prey such as snails or large clams. This week's episode was produced with help from Podigy. About the Science Podcast Authors: Sarah Crespi; Richard Stone; Katherine Irving Episode page: https://www.science.org/doi/10.1126/science.z4pdg62 Learn more about your ad choices. Visit megaphone.fm/adchoices
On this week's show: What happens when the body's own immune system attacks the brain, and how otters' use of tools expands their diet First on the show this week, when rogue antibodies attack the brain, patients can show bizarre symptoms—from extreme thirst, to sleep deprivation, to outright psychosis. Contributing Correspondent Richard Stone joins host Sarah Crespi to discuss the hunt for biomarkers and treatments for this cluster of autoimmune disorders that were once mistaken for schizophrenia or even demonic possession. Next on this episode, producer Katherine Irving talks with Chris Law, an evolutionary biologist at the University of Washington and the University of Texas at Austin, about how sea otters gain energy benefits (and dental benefits) when they use tools to tackle tougher prey such as snails or large clams. This week's episode was produced with help from Podigy. About the Science Podcast Authors: Sarah Crespi; Richard Stone; Katherine Irving Episode page: https://www.science.org/doi/10.1126/science.z4pdg62 Learn more about your ad choices. Visit megaphone.fm/adchoices
Happy 300 episodes to our Curiosity Daily listeners! Today, you'll learn about a massive database of human genomes that's uncovering major new genetic discoveries, the scientists building stone age tools from scratch… and then putting them to work, and a treasure trove of deep sea discoveries in an underwater mountain range. Human Diversity “Ambitious survey of human diversity yields millions of undiscovered genetic variants.” by Max Kozlov. 2024. “NIH's All of Us Research Program Releases First Genomic Dataset of Nearly 100,000 Whole Genome Sequences.” NIH. 2022. “About.” NIH. 2021. Ancient Tool Use “Scientists try out stone age tools to understand how they were used.” EurekAlert! 2024. “Experiments with replicas of Early Upper Paleolithic edge-ground stone axes and adzes provide criteria for identifying tool functions.” by Akira Iwase, et al. 2024. Deep Sea Discoveries “See the strange new species discovered near Chile - with the help of a deep-diving sea robot.” by Laura Baisas. 2024. “Scientists Confirm Underwater Mountains Harbor Abundant Life Off Chile's Coast.” Schmidt Ocean Institute. 2024. “How much of the ocean has been explored?” NOAA. n.d. Follow Curiosity Daily on your favorite podcast app to get smarter with Calli and Nate — for free! Still curious? Get exclusive science shows, nature documentaries, and more real-life entertainment on discovery+! Go to https://discoveryplus.com/curiosity to start your 7-day free trial. discovery+ is currently only available for US subscribers. Hosted on Acast. See acast.com/privacy for more information.
In this episode of the All Creatures podcast, hosts Chris and Angie delve into the intriguing world of puffins, often referred to as the clowns of the sea. They discuss the physical characteristics, behavior, and migratory patterns of puffins, focusing on the Atlantic puffin's vulnerability and the significant threats to its population. The hosts share personal anecdotes and emphasize the importance of conservation efforts to protect these birds. The episode also highlights interesting facts about puffins, including their nesting habits, courtship rituals, and the various threats they face. Additionally, they touch on broader topics like the great auk's extinction, the impact of climate change on bird migration, and the potential intelligence of puffins as evidenced by observed tool-use behavior. A significant portion of the conversation is dedicated to puffin conservation, featuring the work of the organization Puffin Preservation and urging listeners to support puffin conservation efforts. Podcast Timeline 02:00 Welcome to the Puffin Podcast: Meet the Clowns of the Sea 02:42 Diving Deep into Puffin Conservation and Migration 05:44 From Personal Stories to Puffin Populations: A Conservation Chat 08:10 Exploring the Fascinating World of Puffins: Species, Behavior, and Threats 28:27 The Evolution and Lifespan of Puffins: A Deep Dive into Their World 36:54 The Clumsy Takeoff and Landing of Sea Birds 37:25 Evolutionary Trade-offs and Adaptations 38:49 The Unique Features of Puffin Beaks 40:23 The Mystery of the Glowing Beak 42:52 Diet and Coloration: The Carotenoid Connection 49:02 Incredible Migration Patterns of Puffins 54:41 Tool Use and Intelligence in Puffins 59:13 Breeding Habits and Nesting Rituals 01:06:18 Conservation Efforts and Puffin Preservation ------------------------------------------------------ Another thank you to all our Patreon supporters. You too can join for one cup of "good" coffee a month. With your pledge you can support your favorite podcast on Patreon and give back to conservation. With the funds we receive each month, we are have been sending money to conservation organizations monthly. We now send a check to every organization we cover, as we feel they all are deserving of our support. Thank you so much for your support and for supporting animal conservation. Please considering supporting us at Patreon HERE. We also want to thank you to all our listeners. We are giving back to every conservation organization we cover and you make that possible. We are committed to donating large portions of our revenue (at minimum 25%) to every organization we cover each week. Thank you for helping us to grow, and for helping to conserve our wildlife. Please contact us at advertising@airwavemedia.com if you would like to advertise on our podcast Learn more about your ad choices. Visit megaphone.fm/adchoices
Description: Co-host Ryan Piansky, a graduate student and patient advocate living with eosinophilic esophagitis (EoE) and eosinophilic asthma, and Holly Knotowicz, a speech-language pathologist and feeding specialist living with EoE who serves on APFED's Health Sciences Advisory Council, talk with Drs. Robin Shandas and Steven Ackerman, are scientists who work together at EnteroTrack. In this episode, Ryan and Holly interview Drs. Shandas and Ackerman about the development of the esophageal string test (EST), or simply the string test. They discuss the initial concept and the work that led to the development of the string test to collect samples from the upper gastrointestinal mucosa. The test involves swallowing a capsule, similar in size to a TicTacⓇ, with a string attached. It began as an overnight test, and today is a one-hour test. During the conversation, the researchers explained the use of the string test as a monitoring device for EoE, not as a diagnostic device. They describe how the test is administered with no endoscope, no anesthesia, no recovery time, and only minor discomfort for some patients. They discuss the age range for using the device (as young as 4) and the advantages it offers over traditional endoscopy and biopsy for frequent monitoring of eosinophilic esophagitis. Finally, they advise patients on how to access the test: Ask your clinician about it! Disclaimer: The information provided in this podcast is designed to support, not replace the relationship that exists between listeners and their healthcare providers. Opinions, information, and recommendations shared in this podcast are not a substitute for medical advice. Decisions related to medical care should be made with your healthcare provider. Opinions and views of guests and co-hosts are their own. Key Takeaways: [:50] Co-host Ryan Piansky welcomes co-host Holly Knotowicz. Holly introduces the topic of a new unsedated diagnostic tool for EoE, the esophageal string test, or for the purposes of today's podcast episode, the string test. [1:34] Holly introduces Drs. Robin Shandas and Steven Ackerman, are professional scientists who work together at EnteroTrack. Dr. Shandas is the Chief Executive Officer and Dr. Ackerman is the Chief Science Officer. [1:48] Drs. Shandas and Ackerman, with a team of entrepreneurs, engineers, medical doctors, and scientists, developed the EnteroTracker®, the esophageal string test, a new method for collecting and sampling upper gastrointestinal mucosa. [2:00] Holly thanks Drs. Shandas and Ackerman for joining us today. [2:19] Dr. Ackerman has worked on the biology and the roles of eosinophils in the pathogenesis of allergic diseases, including asthma, eosinophilic gastrointestinal disorders, and other hypereosinophilic syndromes for the past 40 years, starting when he was a post-doctoral fellow at the Mayo Clinic in Rochester, Minnesota. [2:44] About 18 years ago, Dr. Ackerman transitioned his lab to do more clinical translational research in eosinophilic GI diseases. His lab developed and validated the string test and the biomarkers and immunoassays that they use to evaluate the results. [3:10] Dr. Ackerman was also interested in studying how the eosinophil participates in tissue remodeling and fibrosis of the esophagus in EoE and other eosinophil-associated diseases. Notably, this started with a small research grant from APFED in 2007 to 2008, which was followed by a more substantial APFED grant in 2011. [3:51] Ryan was involved in APFED in 2007 and he remembers hearing about the early work that was going on for this new test. He thought it would be fantastic to have a test without anesthesia or a scope down his throat. It was exciting to hear about the development over time. [4:23] Dr. Ackerman and Dr. Glenn Furuta, a pediatric gastroenterologist, initially conducted eosinophilic disease research together at Harvard Medical School. Dr. Ackerman eventually went to Chicago but continued to interact with Dr. Furuta. They published papers together on eosinophils and their roles in GI diseases. [4:47] Dr. Ackerman was inspired to develop the string test when Dr. Furuta pointed out to him the tremendous need for a minimally-invasive or non-invasive test for EoE that would remove the need for children to have multiple repeat endoscopies and biopsies under sedation. [5:11] The endoscopies and biopsies were done to follow patient responses to treatment, especially food elimination diets. Dr. Ackerman's goal was to provide a way to evaluate disease activity in children with EoE in a minimally invasive or non-invasive fashion. [5:42] Dr. Shandas is a biomedical engineer, entrepreneur, and educator. He has been a scientist for over 30 years. He is a Distinguished Professor of Bioengineering at The University of Colorado. His passion is to take ideas that he thinks can help people and go through the work. It's really, really hard work to bring those ideas to the clinic! [6:07] Dr. Shandas has been working on the string test for some time. He started eight companies to support this type of activity. The idea of the string test, a simple, non-invasive way to collect samples from the esophagus to support frequent and granular monitoring of disease activity, was compelling when Drs. Furuta and Ackerman presented it to him. [6:38] After Drs. Furuta and Ackerman received funding from APFED they did a lot of basic science research to develop the foundation of the string test. They approached Dr. Shandas to see if he could take the idea and figure out how they could help patients with it. That was compelling to him. He learned a lot about GI and EoE in the process. [7:30] Dr. Ackerman and Dr. Furuta started working on the string test in 2006. It took more than 16 years from concept and pre-clinical lab studies to its marketing and adoption for routine clinical use. It was officially in the clinic at the beginning of 2023. [8:01] Holly Knotowicz was a clinical fellow at Children's Hospital Colorado in 2010 where they were talking about this and she recognized this would be life-changing for patients in all populations not to have to participate in sedation to get updated information. [8:32] The esophageal mucosa has biological information that may reflect the status of the disease in the esophagus. The conventional way of evaluating what is happening in the esophagus is to put a scope down the throat, pinch out a biopsy from the esophagus, and analyze that tissue for inflammatory cells and inflammatory markers. [9:19] Dr. Shandas learned that the mucosa lining the esophagus has a lot of biological information. Because it is a barrier between the outside and your body it has a lot of biochemical reactions; it may be exposed to environmental allergens or food issues. These things interact with your body through the mucosal layer. It's a complex layer. [10:08] Scientists are just starting to figure out the number of things that go on in the esophageal mucosal layer. The string test samples the esophageal mucosa from the throat down to the stomach, for the entire length of the esophagus. The mucosa is the repository of information on how your body interacts with EoE. [10:39] The molecules that the body releases as a function of the reaction that's taking place are expressed in the mucosal layer. The string test collects samples of that mucosal layer. It's a simple, non-painful, non-traditional method of sampling. [11:10] The researchers figured out there's a lot of biological information in the mucosa. Dr. Furuta and Dr. Ackerman have both done clinical studies comparing the string test against traditional biopsies to show that there are great correlations. So they set out to develop a clinically available test. [11:30] For the past eight years, EnteroTrack has done work to answer questions such as: What is the process of swallowing the device and getting the sample? What is the process of analyzing the sample? And how do we put all of that into the framework of a clinical product that can be available to patients and clinicians? [11:49] The string test is a phenomenally simple, easy way to sample mucosal content and analyze it for biomarkers, with no need for any sedation or anything to help you swallow the device. It's a very simple technique. [12:32] Dr. Ackerman was a skeptic when Dr. Furuta first approached him with the idea. He agreed on the need to sample at the site of inflammation. He was surprised at how good the test was in the 16-hour overnight format, compared to endoscopy and biopsy. It was hard to tell the figures apart between the two tests. [13:14] Dr. Ackerman learned they only needed to measure one eosinophil-specific biomarker, major basic protein-1, for a 16-hour test, but when they changed it to a one-hour test, they needed to measure two biomarkers to get enough of a sample. The second biomarker is eotaxin-3, which attracts the eosinophils into the esophagus. [14:18] Dr. Ackerman tells of clinical validation. After multiple applications to the FDA Rare Diseases Program, they were able to get substantial funding to do a Stage 2 clinical evaluation. It was a four-year grant but it took six years to recruit all the patients. They completed and published it in 2019 as the basis for the string test and EoE score. [15:03] Today, the string test is a one-hour test. The visit takes about 90 minutes, including the test, preparation, and removal of the string. The sample is sent to a reference lab. They analyze it and in about 10 days, send your doctor a report that describes the EoE score and the magnitude of the disease. [16:43] There is no recommended frequency for the string test. It can be done as often as your doctor wants to keep track of changes in your disease. Until the string test, there hasn't been a real solution to monitor the disease, with samples every week, two weeks, or four weeks. The researchers encourage patients to ask their clinicians for the string test. [18:19] After setting a baseline, the frequency of tests can be determined by the patient's progress and the clinician's findings. [18:45] The string test is not a replacement for endoscopy. Endoscopy plays an important role in definitively diagnosing the disease and getting a comprehensive sense of what's going on with the disease. The string test is a complement to endoscopy. [19:23] An endoscopic pinch biopsy samples a tiny fraction of the esophageal surface area. EoE is a patchy disease. It can manifest in different places in your esophagus. You may need to do multiple biopsies to make sure you're covering the esophagus nicely and identify a location where the disease manifests. [20:01] An advantage of the string test is that in one test, you get a sample from your throat to your stomach, the entire length of the esophagus. That allows you to get a very comprehensive evaluation of your esophagus. The ease, cost, time savings, out-patient setting, no endoscopy suite, no endoscopist, and no anesthesiologist, are advantages. [20:39] EnteroTrack is pursuing the possibility of doing point-of-care in-home use of the string test. This would be a significant advantage over the traditional way of monitoring the disease. [21:08] A string test cannot be used instead of an endoscopy to make the initial diagnosis of EoE. It's intended as a monitoring tool. Its clinical indication is not as a diagnostic test but for monitoring the disease, in combination with a lab analysis of the assay. Endoscopy is used for the diagnosis. [21:45] In areas where endoscopy and biopsy are not locally available, doctors might use the string test to screen patients to consider referring them to a center where they can get an endoscopy and a biopsy as an initial diagnostic tool. [22:49] Holly is excited to have access to the string test for patients on elimination diets, to test more frequently as foods are reintroduced one at a time to the diet, without putting the patient under anesthesia for a traditional endoscopy. [23:58] During the monitoring of EoE, if the doctor sees the data is trending in a certain way, it may be useful to get an endoscopy and biopsy to get additional information. [24:28] Ryan has had dozens of traditional endoscopies and biopsies so he's always excited to hear about new tests. He's been putting off an endoscopy and he needs one soon. Is Ryan, an adult, a candidate for this procedure at his local hospital? Yes! He has been diagnosed with EoE, so he can be monitored by the string test. [25:12] The youngest patient to swallow the device was a four-year-old and they did great with it; the test was very successful. EnteroTrack recommends that parents talk to their child's physician about the option to use this device. The key is if the patient, parent, and clinicians all feel that the patient can swallow the device. [25:35] The string test device is about the size of a TicTac®. It's important for the swallowing procedure to be successful. EnteroTrack works with clinical groups to give them tips on how best to ensure that the procedure is successful. [25:56] Holly is a feeding specialist. One of the things she helps people with eosinophilic diseases to learn is how to swallow pills. She helps them practice with different kinds of candies and capsules. [26:40] The clinical validation study included patients aged seven and older. Now that the product is released, clinicians, parents, and patients can together make the decision if they can swallow it. EnteroTrack looks at de-identified data to evaluate how the string test is doing. They are seeing younger age groups successfully swallow it. [27:25] There are factors that contribute to successful swallowing. One is the parental motivation of their children. Also, practice swallowing something the size of the device. [27:39] EnteroTrack is providing training to physicians and nurses at the centers that are adopting the test. They are aware of how to make the swallowing part of this test as successful as possible with the least amount of stress. [28:02] The EnteroTracker® is considered a minimal-risk device, and minimal-risk procedure. If you are anxious about swallowing, or you have challenges as you swallow it, you may gag a little bit. [28:37] In the Phase 2 data, there was a small percentage who had challenges in terms of gagging, when swallowing the device. All of those issues were resolved. Even if the patient gags, the clinician has tips to help the patient get through that. Repeated gulps of water can help bring the capsule down and alleviate those issues. [29:37] In the clinical validation studies, there were hundreds of string tests performed and there were no serious adverse reactions or problems other than the minor problem of some gagging. During the 16-hour tests, there was one patient who chewed through the string while sleeping and swallowed it. The string just passed through the GI tract. [32:47] Dr. Shandas has personally swallowed about 50 string tests to understand the patient experience. After the swallow, when the string is in the esophagus, there is a feeling of tickling in the side of the throat in the beginning, but that resolves in a couple of minutes and it is not noticed at all. It helps if the patient has something to occupy their attention. [34:05] The nurses and clinicians are trained to pull the device out, not fast and not slow. It should take three to five seconds. There is some discomfort as the string is coming out. It may feel like burning in the throat but it happens quickly and it resolves immediately. They have not seen big challenges with this, much less than in swallowing. [34:50] The string has two parts. The part that goes from your mouth down your throat is like dental floss. It's connected to the collection string that goes down the esophagus to the stomach. That collection string is a thicker nylon yarn designed to optimally collect mucosal samples. It would be uncomfortable in the throat, but not in the esophagus. [36:06] EnteroTrack has gotten a Medicare code for the assay and a Medicare reimbursement for the assay. The next step is getting private insurers to pay for the test. EnteroTrack is working on that. It will require educating insurance companies about the value of the test. Patients, clinicians, and groups like APFED can help. [37:51] The string test is clearly less expensive than doing an endoscopy and biopsy. That should be one of the motivating factors for insurance companies. It is significantly lower cost than the test that is used as the standard of care. [38:38] Some large centers have begun using the esophageal string test, including Denver, Phoenix, Tampa, Dallas, and Chicago. EnteroTrack has received interest from clinicians and patients around the country. [39:04] Patients and parents, let your clinicians know if you are interested in the esophageal string test (EST). Your doctor can contact EnteroTrack and ask how to get the test to their clinic. It's a simple process and the clinical training is very straightforward. [39:35] APFED plans to add information about the esophageal string test to its Specialist Finder at apfed.org/specialist to help patients find clinicians offering this test. [39:53] Drs. Ackerman and Shandas anticipate that the number of centers offering this test will grow exponentially as people see that it's working for early adopters. [40:10] Ryan and Holly thank Drs. Shandas and Ackerman for sharing their expertise to help others today in the podcast. Dr. Shandas thanks Ryan, Holly, and APFED for the opportunity to talk about the string test. [40:35] As entrepreneurs and researchers, they are constantly innovating, including looking for other applications for the string test. They have clinical trials going on to use the string test for esophageal issues, gastric issues, and small intestinal issues, analyzing the contents for different biomarkers that have correlations with a lot of upper GI diseases. [41:22] To learn more about eosinophilic esophagitis, visit apfed.org/eoe, to access the Specialist Finder, visit apfed.org/specialists, and to connect with patients and families living with eosinophilic disorders, visit apfed.org/eos-connections. Ryan and Holly thank Drs. Shandas and Ackerman again for an interesting conversation and thank APFED's education partners, linked below, for supporting this episode. Mentioned in This Episode: American Partnership for Eosinophilic Disorders (APFED) APFED on YouTube, Twitter, Facebook, Pinterest, Instagram Dr. Robin Shandas Dr. Steven Ackerman EnteroTrack, LLC Mayo Clinic Dr. Glenn Furuta The University of Colorado Real Talk: Eosinophilic Diseases Podcast Education Partners: This episode of APFED's podcast is brought to you thanks to the support of AstraZeneca, Bristol Myers Squibb, Sanofi, and Regeneron. Tweetables: “I was interested in studying how the eosinophil participates in tissue remodeling and fibrosis of the esophagus in EoE and other eosinophil-associated diseases. Notably, this started with a small research grant from APFED in 2007 to 2008.” — Dr. Steven Ackerman “The mucosa that lines your esophagus has a lot of biological information. … The mucosa is the barrier between the outside world and your body. It is this very interesting biological fluid and … it has a lot of biochemical reactions.” — Dr. Robin Shandas “Since we released the product earlier this year, we've had a four-year-old swallow the device; did great, very successful. And so, we recommend that patients, parents, talk to your physician.” — Dr. Robin Shandas Featured speakers: Dr. Steven Ackerman Dr. Robin Shandas
John Ferry is back to discuss new features available on the Use Rogue platform. Speaking from peresonel experience, I can tell you this is a game changer. It's not only saving time assembling responses to the government but is now able to ingest a federal RFP and produce a usable outline with poulated information. Listen to John explain the latest features and how companies are able to engage on more opportunities because of this amazing technology. If you are interested in Use Rogue go to the following link for incentives they are offering to our listeners: USE ROGUE ------------------------------------------------------------------ Reserve a copy of Richard's upcoming book: https://www.dodcontract.com/100meetings Schedule a 60 minute deep dive on your business and get your questions answered: https://www.dodcontract.com/DefenseContractAssessment Need someone to set up your government sales process? Richard's team is trained and closely managed to deliver results for our clients. Apply to see if your existing business is a good fit for our Done For You services: https://www.dodcontract.com/done-with-you-and-done-for-you-defense-contract-assessment-1
One of the great revolutions in the history of the development of humankind is the controlled use of fire. When did we as a species learn to keep and make fire? What changes did it make to humans socially and physically. Jeni, our resident Anthropologist, has the answers.
Owls zero in on their prey under snow by eavesdropping on the sounds they make; Elephant behaviour helps to maintain healthy, carbon-rich forests; Feisty songbirds swarm their predators – but only when the time's right; The ‘sensory moustache' that helps bats find sweet snacks; Cockatoos have a handy tool belt to fish for cashews; Seals may not tap their toes, but seals also appreciate a good musical rhythm; Listener Question: Why can't waste plastic be dumped into volcanoes?
Anthropogeny, has provided many new discoveries over the past decade, ranging from new fossil finds to ancient DNA data, including from extinct hominins. This CARTA symposium highlights where the future efforts should be focused and what type of novel collaborations are most promising for improving our understanding of the human phenomenon. Pascal Gagneux offers a zoological view of the human cultural animal. Robert Kluender talks about the evolution of language structure and the future of linguistics. Anne Stone discusses ancient DNA research and the surprising insights into human evolutionary history. Series: "CARTA - Center for Academic Research and Training in Anthropogeny" [Humanities] [Science] [Education] [Show ID: 38633]
Anthropogeny, has provided many new discoveries over the past decade, ranging from new fossil finds to ancient DNA data, including from extinct hominins. This CARTA symposium highlights where the future efforts should be focused and what type of novel collaborations are most promising for improving our understanding of the human phenomenon. Pascal Gagneux offers a zoological view of the human cultural animal. Robert Kluender talks about the evolution of language structure and the future of linguistics. Anne Stone discusses ancient DNA research and the surprising insights into human evolutionary history. Series: "CARTA - Center for Academic Research and Training in Anthropogeny" [Humanities] [Science] [Education] [Show ID: 38633]
Anthropogeny, has provided many new discoveries over the past decade, ranging from new fossil finds to ancient DNA data, including from extinct hominins. This CARTA symposium highlights where the future efforts should be focused and what type of novel collaborations are most promising for improving our understanding of the human phenomenon. Pascal Gagneux offers a zoological view of the human cultural animal. Robert Kluender talks about the evolution of language structure and the future of linguistics. Anne Stone discusses ancient DNA research and the surprising insights into human evolutionary history. Series: "CARTA - Center for Academic Research and Training in Anthropogeny" [Humanities] [Science] [Education] [Show ID: 38633]
Anthropogeny, has provided many new discoveries over the past decade, ranging from new fossil finds to ancient DNA data, including from extinct hominins. This CARTA symposium highlights where the future efforts should be focused and what type of novel collaborations are most promising for improving our understanding of the human phenomenon. Pascal Gagneux offers a zoological view of the human cultural animal. Robert Kluender talks about the evolution of language structure and the future of linguistics. Anne Stone discusses ancient DNA research and the surprising insights into human evolutionary history. Series: "CARTA - Center for Academic Research and Training in Anthropogeny" [Humanities] [Science] [Education] [Show ID: 38633]
Anthropogeny, has provided many new discoveries over the past decade, ranging from new fossil finds to ancient DNA data, including from extinct hominins. This CARTA symposium highlights where the future efforts should be focused and what type of novel collaborations are most promising for improving our understanding of the human phenomenon. Pascal Gagneux offers a zoological view of the human cultural animal. Robert Kluender talks about the evolution of language structure and the future of linguistics. Anne Stone discusses ancient DNA research and the surprising insights into human evolutionary history. Series: "CARTA - Center for Academic Research and Training in Anthropogeny" [Humanities] [Science] [Education] [Show ID: 38633]
CARTA - Center for Academic Research and Training in Anthropogeny (Video)
Anthropogeny, has provided many new discoveries over the past decade, ranging from new fossil finds to ancient DNA data, including from extinct hominins. This CARTA symposium highlights where the future efforts should be focused and what type of novel collaborations are most promising for improving our understanding of the human phenomenon. Pascal Gagneux offers a zoological view of the human cultural animal. Robert Kluender talks about the evolution of language structure and the future of linguistics. Anne Stone discusses ancient DNA research and the surprising insights into human evolutionary history. Series: "CARTA - Center for Academic Research and Training in Anthropogeny" [Humanities] [Science] [Education] [Show ID: 38633]
IAM Here And Isn't Going Nowhere. In Today's Episode, We Will Talk About Our Purpose. We Got Work To Do As Sonini's Children. The Time Is Now We Wake Up And Start Helping The Cause In Righoeness Not In Evilness. Our People Are Dying At A High Rate. Children Are Being Kidnapped Right Before Our Eyes And Things Are Only Going To Get Worst. Join Me As We Talk About IAM Is The Tool Used To Wake Up The Troops And The Youth. Subscribe to my page https://www.youtube.com/channel/UCpGoXQiHmhEhE31POEBVgFw?sub_confirmation=1 Location of the Real Jerusalem https://youtu.be/BOm05qVIJDo https://www.epidemicsound.com RUMBLE. https://rumble.com/account/ Facebook group page https://www.facebook.com/groups/repent12media Podcast Spotify link https://open.spotify.com/show/4tp2Jpcea4x45aQqDUaugM?si=4189de1a8a1f43cc Instagram Page https://www.instagram.com/repent_or_die_podcast1/ BitChute link. https://www.bitchute.com/profile/ Website. www.repentordiepodcast.net Donations Cash app. https://cash.app/$JunnsTheJew #repentordiepodcast #bible #TMH #SoNininaNini Music by Kids on the Porch(instrumental) Iso Indies and rapped by Junns The Original Copyright Disclaimer under section 107 of the Copyright Act 1976, allowance is made for “fair use” for purposes such as criticism, comment, news reporting, teaching, scholarship, education, and research. Fair use is a use permitted by copyright statute that might otherwise be infringing. --- Send in a voice message: https://podcasters.spotify.com/pod/show/repent-or-die-podcast/message
Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Results for a survey of tool use and workflows in alignment research, published by jacquesthibs on December 19, 2022 on LessWrong. In March 22nd, 2022, we released a survey with an accompanying post for the purpose of getting more insight into what tools we could build to augment alignment researchers and accelerate alignment research. Since then, we've also released a dataset, a manuscript (LW post), and the (relevant) Simulators post was released. This post is an overview of the survey results and leans towards being exhaustive. Feel free to skim. In our opinion, the most interesting questions are 6, 11, 12, and 13. We hope that this write-up of the survey results helps people who want to contribute to this type of work. Motivation for this work We are looking to build tools now rather than later because it allows us to learn what's useful before we have access to even more powerful models. Once GPT-(N-1) arrives, we want to be able to use it to generate extremely high-quality alignment work right out of the gate. This work involves both augmenting alignment researchers and using AI to generate alignment research. Both of these approaches fall under the “accelerating alignment” umbrella. Ideally, we want these kinds of tools to be used disproportionately for alignment work in the first six months of GPT-(N-1)'s release. We hope that the tools are useful before that time but, at the very least, we hope to have pre-existing code for interfaces, a data pipeline, and engineers already set to hit the ground running. Using AI to help improve alignment is not a new idea. From my understanding, this is a significant part of Paul Christiano's agenda and a significant part of his optimism about AI alignment. Of course, automating alignment is also OpenAI's main proposal and Jan Leike has been talking about it for a while. Ought has also pioneered doing work in this direction and I'm excited to see them devote more attention to building tools even more highly relevant to accelerating alignment research. Finally, as we said in the survey announcement post: In the long run, we're interested in creating seriously empowering tools that fall under categorizations like STEM AI, Microscope AI, superhuman personal assistant AI, or plainly Oracle AI. These early tools are oriented towards more proof-of-concept work, but still aim to be immediately helpful to alignment researchers. Our prior that this is a promising direction is informed in part by our own very fruitful and interesting experiences using language models as writing and brainstorming aids. One central danger of tools with the ability to increase research productivity is dual-use for capabilities research. Consequently, we're planning to ensure that these tools will be specifically tailored to the AI Safety community and not to other scientific fields. We do not intend to publish the specifics methods we use to create these tools. Caveat before we get started As mentioned Logan's post on Language Models Tools for Alignment Research (and many others we've talked to), could this work be repurposed for capabilities work? If made public with flashy demos, it's quite likely. That's why we'll be keeping most of this project private for alignment research only (the alignment text dataset is public). Survey Results We received 22 responses in total and some responses were optional so not all questions received responses from everyone. Of course we would have preferred even more responses, but this will have to do for now. We expect to iterate on tools with alignment researchers, so hopefully we get a lot of insights through user interviews of actual products/tools. If you are interested in answering some of the questions in the survey (all questions are optional!), here's the link. Leaving comments on this post would also be appre...
Our species is between 300,000 and 200,000 years old. For most of this one-quarter of a million years, up until just 12,000 years ago, it appears that our ancestors lived in small populations, in small-scale societies of which we can only guess the real nature. We remain in the dark about the age of some of the most diagnostic features of our species: our striding bipedalism, complex tool manufacture and use, fire use, language and societies defining their own identities, collaborating with and competing against neighboring societies. Data from studies of non-human animal behavior remind us of the underappreciated capacities of many other species. However, it has so far not provided evidence for any other species that shares the long list of distinctly human characteristics; chief among those, our species' capacity to not only simultaneous modify and threaten planetary ecosystems but also document and study such ecosystems across the globe. Series: "CARTA - Center for Academic Research and Training in Anthropogeny" [Humanities] [Science] [Show ID: 38298]
Our species is between 300,000 and 200,000 years old. For most of this one-quarter of a million years, up until just 12,000 years ago, it appears that our ancestors lived in small populations, in small-scale societies of which we can only guess the real nature. We remain in the dark about the age of some of the most diagnostic features of our species: our striding bipedalism, complex tool manufacture and use, fire use, language and societies defining their own identities, collaborating with and competing against neighboring societies. Data from studies of non-human animal behavior remind us of the underappreciated capacities of many other species. However, it has so far not provided evidence for any other species that shares the long list of distinctly human characteristics; chief among those, our species' capacity to not only simultaneous modify and threaten planetary ecosystems but also document and study such ecosystems across the globe. Series: "CARTA - Center for Academic Research and Training in Anthropogeny" [Humanities] [Science] [Show ID: 38298]
CEOs, founders, and executive-level SMEs are some of the most powerful voices your brand has at its disposal. So why don't more marketing teams take advantage of the top-tier minds in their organizations? The most common answer among executives is "I don't have time". The real answer is that they don't make it a priority. People often ask how Refine Labs gets so much traction on LinkedIn or on the State of Demand Podcast. The answer: Senior leadership sets the tone and shows what "good" looks like by creating content in dark social channels regularly. The rest of the team sees that and follows suit. In this episode, Chris talks about how he finds time to be active inside dark social channels while still managing the day-to-day responsibilities of a CEO in a rapid-growth company.
Dr. David Braun of George Washington University's Center for the Advanced Study of Human Paleobiology in the Anthropology Department chats about the cycles of tool use and niche construction. We talk about how one affects the other and vice versa in cycles, plus the interplay of greater environmental and climate change. Dr. Braun also discusses how we can look into the near and deep past to figure out environmental change. Find links to articles, books, and pics at AnthroBiology.com. Find the show on Instagram and Twitter @AnthroBiology. Email the host at gaby.lapera@anthrobiology.com.
Episode: 2200 Jays who are smarter than you'd ever have thought. Today, smart birds.