From a tiny island in Seychelles to the heartland of Ohio, we unpack a wild week in AI. First up: 1X’s “Neo” humanoid—$20k to buy or $500/month to rent—promising laundry, dishes, and errands soon…with a lot of teleoperation today. We debate whether tele-ops is a feature (not a bug), who it employs, and how quickly autonomy could follow. Then we zoom out to the money: Nvidia touches a $5T valuation, OpenAI reportedly eyes a $1T IPO, and the industry’s circular funding loops raise both eyebrows and opportunity. We also test-drive OpenAI’s Atlas browser (a Chromium fork with action-taking ambitions), and dig into Cursor’s agentic coding push, new in-house model, and blistering growth—plus the eternal “moat vs. momentum” question. Along the way: a live Neo preorder, enterprise ROI reality checks, and why agents may turn devs into project managers. If you’re curious where robotics, chips, and agentic software collide, this one’s for you. Ask a question on our Youtube Channel: https://www.youtube.com/@GenerativeAIMeetup Mark's Travel Vlog: https://www.youtube.com/@kumajourney11 Mark's Personal Youtube Channel: https://www.youtube.com/@markkuczmarski896 Attend a live event: https://genaimeetup.com/ Shashank Linked In: https://www.linkedin.com/in/shashu10/ 

In this episode of the Gen.AI Meetup Podcast, hosts Shashank and Mark dive into the latest AI developments that are reshaping how we create, code, and browse. They explore OpenAI's impressive Sora 2 video generation model and its built-in social network, compare it with Google's VO3, and discuss whether AI-generated content will become mainstream entertainment. The conversation shifts to the newest coding models, including Anthropic's Claude 4.5 Sonnet and Grok 4 Fast, examining their performance, pricing, and whether they're worth the cost for developers. Mark shares his experience vibe coding with Cursor and why faster, cheaper models might be better than the most powerful ones. The hosts also explore the maturing AI browser space, discussing Perplexity's Comet browser, Dia from the Browser Company, and Google's Gemini integration in Chrome. They debate whether these AI-native browsers can convince users to switch from Chrome and what features would actually make them indispensable. Finally, they tackle the big question: Is NVIDIA's $4.5 trillion valuation justified? They discuss the company's dominance in AI chips, the circular investment patterns in the industry, and whether specialized compute chips can compete with NVIDIA's end-to-end ecosystem. Timestamps: 0:00 - Intro & OpenAI's Sora 2 announcement 8:30 - Sora 2 vs Google VO3: The new video generation king 15:45 - Claude 4.5 Sonnet: Worth the premium price? 25:20 - Grok 4 Fast: Crazy cheap, crazy fast 35:15 - NVIDIA's dominance: Bubble or justified? 50:40 - AI browsers: Comet, Dia, and the future of browsing 1:02:15 - Ambient computing and what's next Mentioned Resources: OpenRouter - Multi-model API aggregator Cursor - AI-powered code editor Perplexity Comet - AI-native browser Upcoming event: Coding Agents Showcase - Jan 9th, Palo Altohttps://partiful.com/e/joRDIOYMqpogKjNtvlHY  Don't forget to RSVP for our Coding Agents event featuring Zed, Augment Code, Code Flash, Factory AI, and more! Spots are limited and filling fast. Have questions? Drop them in the YouTube comments and we'll answer them in future episodes!

Ask a question on our Youtube Channel: https://www.youtube.com/@GenerativeAIMeetup Mark's Travel Vlog: https://www.youtube.com/@kumajourney11 Mark's Personal Youtube Channel: https://www.youtube.com/@markkuczmarski896 Attend a live event: https://genaimeetup.com/ Shashank Linked In: https://www.linkedin.com/in/shashu10/    In this episode of the Generative AI Meetup Podcast, hosts Shashank and Mark dive into the latest breakthroughs in AI and technology. Fresh from his adventures across North America, Mark joins Shashank to discuss OpenAI's groundbreaking GPT-4B advancements in biology and medicine, including AI-driven cell reprogramming for potential longevity breakthroughs. They also cover Meta's newly announced Ray-Ban smart glasses with heads-up displays and gesture controls, Apple's iPhone 17 AI features like real-time translation and transcription (and how they stack up against Google's Pixel phones), massive funding for Figure AI's humanoid robots, Tesla's Optimus updates, and Waymo's expansion to San Francisco International Airport. Plus, insights on the US-China robotics race, upcoming coding agents meetups during Tech Week, and the future of consumer AI hardware. Tune in for an engaging mix of news, analysis, and excitement about AI's real-world impact! Mark Linked in: https://www.linkedin.com/in/markkuczmarski/ 

Ask a question on our Youtube Channel: https://www.youtube.com/@GenerativeAIMeetup Mark's Travel Vlog: https://www.youtube.com/@kumajourney11 Mark's Personal Youtube Channel: https://www.youtube.com/@markkuczmarski896 Attend a live event: https://genaimeetup.com/ Shashank Linked In: https://www.linkedin.com/in/shashu10/  Mark Linked in: https://www.linkedin.com/in/markkuczmarski/  In this episode of the Generative AI Meetup Podcast, hosts Shashank (a software engineer at Google Labs working on the AI vibe design tool Stitch) and Mark (a former Amazon engineer now building a stealth startup) dive into the world of "vibe coding"—a revolutionary approach to programming inspired by AI researcher Andrei Karpathy. Vibe coding lets you focus on the big picture and product vision while letting large language models (LLMs) handle the nitty-gritty details, melting away traditional coding hurdles. Shashank walks through his weekend project, Convo (convochat.io), an AI-powered app that analyzes exported chat backups from WhatsApp, Telegram, Facebook Messenger, or SMS. It delivers fun stats (like 11,000 messages sent to a friend over four years), conversation summaries, sentiment analysis, and even tips for better chats—all built with minimal manual coding. The duo shares practical tips for vibe coding success: brainstorming ideas with Claude, designing UIs in Stitch, building with tools like Cursor or Replit, using Git for checkpoints, picking popular frameworks (e.g., Tailwind CSS), writing tests, debugging with logs, optimizing SEO, and branding with AI-generated logos. They discuss pros (rapid prototyping, low costs—Shashank spent just $18), cons (scaling challenges, bug fixes), and best practices for beginners, including modularity, documentation, and refactoring. Whether you're a seasoned dev or a total newbie, this episode shows how AI tools can turn ideas into launched MVPs in days, not months. Tune in for inspiration, real-world examples, and motivation to vibe code your next project!

Ask a question on our Youtube Channel: https://www.youtube.com/@GenerativeAIMeetup Mark's Travel Vlog: https://www.youtube.com/@kumajourney11 Mark's Personal Youtube Channel: https://www.youtube.com/@markkuczmarski896 Attend a live event: https://genaimeetup.com/ Shashank Linked In: https://www.linkedin.com/in/shashu10/  Mark Linked in: https://www.linkedin.com/in/markkuczmarski/ 

Youtube Channel: https://www.youtube.com/@GenerativeAIMeetupMark's Travel Channel: https://www.youtube.com/@kumajourney11 Mark's Channel: https://www.youtube.com/@markkuczmarski896 Gen AI Meetup: https://genaimeetup.com/ Shashank Linked In: https://www.linkedin.com/in/shashu10/  Mark Linked in: https://www.linkedin.com/in/markkuczmarski/  Join hosts Shashank and Mark in this electrifying episode of the Gen.ai Meetup Podcast, where they unpack a whirlwind week of AI advancements reshaping the future of technology. From Anthropic's Claude 4.1—a subtle yet powerful upgrade boosting coding prowess and multi-file edits for enterprise dominance—to OpenAI's long-awaited open-source comeback with GPT-OSS models (a beefy 120B parameter beast and a tiny laptop-friendly version rivaling proprietary giants), the duo dives into benchmarks, real-world applications, and how tools like Ollama make deployment a breeze. They explore Gemini’s DeepThink, a reasoning powerhouse solving Olympiad-level math puzzles through extended inference, and Google’s groundbreaking “world model”—a seamless blend of video generation and game engine tech that lets you control characters in hyper-realistic, physics-aware simulations. Along the way, Shashank and Mark share candid insights on vibe coding pitfalls, side projects built with AI agents, OpenAI’s staggering valuations, and the open-source ecosystem’s role in driving innovation. Whether you’re a developer wrestling with agentic workflows, an enterprise leader eyeing LLM integrations, or an AI enthusiast dreaming of interactive worlds, this episode delivers expert analysis, practical tips, and forward-thinking speculation. Tune in for a fun, far-flung chat (Mark’s broadcasting from a Canadian road trip en route to the Arctic!) and discover why AI’s evolution is accelerating faster than ever. Drop your questions in the comments—we’ll tackle them next time! Timestamps: 00:00:00 - Introduction: Shashank welcomes listeners and introduces Mark, who's road-tripping in Canada to the Arctic Ocean.00:03:50 - Episode Overview: A quick rundown of the week’s major AI announcements.00:07:44 - Claude 4.1 from Anthropic: Discussing the incremental improvements of Claude Opus 4.1, its coding strengths, and enterprise adoption.00:16:32 - Claude’s Enterprise Impact: Why Claude leads in enterprise LLMs and its role in tools like Cursor for vibe coding.00:28:38 - Gemini’s DeepThink Feature: Deep dive into Gemini’s reasoning capabilities for complex math and problem-solving.00:29:28 - OpenAI’s GPT-OSS Release: OpenAI’s open-source models (120B and 20B parameters), their performance, and community implications.00:44:94 - OpenAI’s Valuation Debate: Exploring OpenAI’s $300B valuation and the strategic benefits of open-source releases.00:45:18 - Google’s World Model Announcement: Exploring the steerable 3D environments blending video generation and game engine tech.00:50:32 - World Model Applications: Potential uses in robotics, self-driving, and synthetic data generation.00:54:86 - Coding Agents and Side Projects: Shashank and Mark share experiences with vibe coding and AI-powered side projects.00:58:74 - Amazon’s Spec-Driven Development: Insights on Amazon’s Kero tool and the importance of detailed software specifications.00:58:94 - Ollama and Ollama Turbo: How Ollama simplifies model deployment and the new cloud-based Ollama Turbo service.01:07:26 - Prompt Engineering Tips: Practical advice on crafting effective prompts and iterating with LLMs for better outputs.01:11:50 - Closing and Call for Questions: Wrap-up and a call for listener questions in the YouTube comments. Subscribe and leave a comment with your questions for the next episode! #AI #GenAI #Claude4.1 #OpenAI #GPTOSS #GeminiDeepThink #WorldModels #Ollama #CodingAgents #TechPodcast

Youtube Channel: https://www.youtube.com/@GenerativeAIMeetupMark's Travel Channel: https://www.youtube.com/@kumajourney11 Mark's Channel: https://www.youtube.com/@markkuczmarski896 Gen AI Meetup: https://genaimeetup.com/ Shashank Linked In: https://www.linkedin.com/in/shashu10/  Mark Linked in: https://www.linkedin.com/in/markkuczmarski/ 

Listen to the podcast: https://podcast.genaimeetup.com/ Join the Meetup: https://genaimeetup.com/ Subscribe to Mark's youtube channel: https://www.youtube.com/@UCqLp0T6WBwxmuJdmo4mUW5A Youtube: https://www.youtube.com/@GenerativeAIMeetup 

Try Merlin: https://www.getmerlin.in/ X: https://x.com/MerlinAIByFoyer Youtube: https://www.youtube.com/@merlin_ai/videos  South Bay Gen AI: https://genaimeetup.com/  Mark's Youtube Channel: https://www.youtube.com/@markkuczmarski896 

https://rtrvr.ai/ Join hosts Shashank from Google AI Labs and Mark (currently on hiatus from Amazon) as they interview Arjun and Pavani, the founders of Retriever AI. This episode explores how Retriever's innovative browser extension brings agentic LLM capabilities directly to your desktop browser, offering unique advantages over cloud-based alternatives. Learn how Retriever can automate repetitive tasks, extract data across multiple websites, and interact with your personal accounts while maintaining security and privacy. The founders share their journey from big tech to startup life, demonstrate real-world use cases, and reveal their exciting vision for a federated network of browser agents that could revolutionize how we interact with the web.

https://www.genspark.ai/https://genaimeetup.com/Follow the podcast: https://podcast.genaimeetup.com/ Join hosts Shishank and Mark as they dive deep into the world of generative AI agents with Lenzoy Lin, Engineering Lead at Gens Park. Discover how this rapidly growing startup is revolutionizing productivity through their suite of AI agents - from their groundbreaking phone call agent to deep research tools and slide creation capabilities. Learn how Gens Park has grown from $10M to $22M in ARR in just one month, their approach to building reliable AI systems, and get a glimpse into the future of human-AI collaboration. Whether you're a tech enthusiast, entrepreneur, or AI professional, this episode offers valuable insights into one of 2025's most promising AI startups. In this episode: Lenzoy Lin's journey from Google to leading Gens Park's engineering teamHow Gens Park's mixture of agents approach solves complex tasksBehind the scenes of their phone call agent, deep research tools, and slide creation capabilitiesThe technical challenges of building reliable AI agents at scaleGens Park's position in the competitive AI landscape and future roadmap

https://www.realwear.com/ Timon Binder: https://www.linkedin.com/in/timon-binder/   Join us for a special episode as we sit down with Timon Binder, CTO of Realware, a leading AR hardware company transforming the enterprise landscape. Timon shares his journey from co-founding a startup in Switzerland to leading innovation in the US, revealing how Realware’s AR headsets are revolutionizing industries like manufacturing, logistics, and healthcare. We dive deep into: Realware’s approach to solving real-world problems with AR The challenges and opportunities in global hardware manufacturing How AI and voice assistants are reshaping user interaction The future of consumer AR and the transition from B2B to B2C Startup lessons, work culture differences, and advice for aspiring founders Whether you’re a tech enthusiast, entrepreneur, or curious about the future of augmented reality, this episode is packed with insights, practical advice, and candid stories from the frontlines of innovation. Call to Action:If you’re interested in AR, app development, or want to connect with Realware, check out the links in the episode description for open roles and collaboration opportunities!

Langflow: https://www.langflow.org/  https://www.producthunt.com/products/langflow  Langflow Desktop: https://www.langflow.org/desktop  In this insightful interview, Rodrigo from Langflow discusses the evolution and future of their low-code agent building platform. Starting with his background in machine learning and data science, Rodrigo explains how Langflow began as a vision to connect specialized AI models years before ChatGPT existed. The conversation covers Langflow's journey from open-source project to being acquired by DataStacks while maintaining its commitment to open-source principles. Rodrigo announces the exciting launch of Langflow Desktop, designed to democratize AI development by eliminating technical barriers through an intuitive drag-and-drop interface. Rodrigo details how Langflow serves both technical and non-technical users, supporting three main application types: LLM pipelines, RAG systems, and multi-agent applications. The interview highlights Langflow's integration with the new MCP protocol for more structured and efficient tool usage by AI agents. Looking to the future, Rodrigo envisions advanced agent orchestration systems where AI agents can assign tasks to each other, with humans serving as collaborators in the process. This episode offers valuable insights for anyone interested in the rapidly evolving landscape of AI agent development and deployment.

In this special episode, hosts Mark and Shashank take a break from their usual news coverage to explore the rapidly evolving world of AI agents. They define what agents are, examine their current capabilities, and make bold predictions about how these technologies will transform our lives over the next 1, 5, and 10 years. The hosts discuss how AI agents are already revolutionizing software development, research, and reasoning tasks, while exploring the imminent impact on knowledge work, self-driving vehicles, and scientific breakthroughs. Looking further ahead, they predict the widespread adoption of humanoid robots, the "YouTubeification" of product creation, and fundamental shifts in employment patterns. Whether you're new to AI or a seasoned professional, this episode offers fascinating insights into how exponential growth in AI capabilities will reshape our society, economy, and daily lives in ways we're only beginning to imagine.    

In this episode, hosts Mark and Shashank dive into recent developments in generative AI technology. They begin with NVIDIA's latest GTC announcements, including partnerships with GM for self-driving technology and advancements in robotics with Google DeepMind and Disney. The hosts debate the merits of camera-only versus LiDAR-based autonomous driving systems, referencing Mark Rober's viral comparison video. They also discuss NVIDIA's upcoming Ruben chip, which promises a 10-15x performance increase over the current Blackwell architecture. The conversation shifts to a correction about the DeepSeek model's computational requirements before culminating in a thought-provoking discussion about the challenges of creating generalized robotics systems and how simulation environments might accelerate development. Throughout the episode, the hosts share insights on what these technological advancements might mean for the future of AI and robotics.

This conversation delves into the latest developments in AI, particularly focusing on Google's Gemma models and their capabilities. The discussion covers the differences between various types of language models, the significance of multimodal inputs, and the training techniques employed in AI models. The hosts also explore the implications of open-source versus proprietary models, the hardware requirements for running these models, and the limitations of benchmarks in evaluating AI performance. Additionally, they touch on the future of robotics and the cultural differences in AI adoption, particularly between Japan and the United States.takeaways Open source models are pushing the boundaries of AI.Gemma models are capable of multimodal inputs.Different types of LLMs serve different purposes.Benchmarks can be misleading and should be approached with caution.Training techniques like RLHF are crucial for model performance.The hardware requirements for AI models vary significantly.Cultural differences affect the adoption of robotics and AI.Robots are increasingly filling labor gaps in societies with declining populations.AI benchmarks should be tailored to specific use cases.The future of robotics and AI feels imminent and exciting. Chapters 00:00 Introduction to the Week's AI Developments00:50 Exploring Google's Gemma Models03:21 Understanding Different Types of LLMs05:32 Gemma's Multimodal and Multilingual Capabilities08:45 Training Techniques Behind Gemma15:48 Open Source Models and Their Impact20:34 Benchmarking AI Models28:30 Gaming Benchmarks in AI34:10 The Ethics of Benchmarking in AI44:56 Language Learning and AI Models49:12 The Importance of Benchmarks52:35 Vibe Checks and User Preferences01:01:09 Top AI Models and Their Performance01:13:35 Robotics and the Future of AI01:27:20 Cultural Perspectives on Automation

In this conversation, Mark and Shashank discuss the latest advancements in AI, particularly focusing on China's lead in developing agentic models like Manus. They explore the capabilities of Manus, its performance on the Gaia benchmark, and the implications of AI in various sectors, including software development and customer service. The discussion also delves into the philosophical aspects of AI's potential to solve humanity's problems and the future of work as AI continues to evolve.takeaways China is leading in AI breakthroughs with models like Manus.Manus represents a shift towards more comprehensive AI solutions.The Gaia benchmark is crucial for evaluating AI performance.Manus scored impressively on the Gaia benchmark, nearing human performance.Future advancements in AI will focus on tool use and reasoning capabilities.The cost of training AI models is becoming a significant concern.AI can automate tedious tasks, impacting job markets.The marginal cost of software development is decreasing rapidly.AI has the potential to solve complex human problems.The evolution of AI will lead to more creative and powerful applications. Chapters00:00AI Breakthroughs: China's Lead in Agentic Computing02:53Understanding Manus: More Than Just a Model06:05Benchmarking AI: The Gaia Benchmark Explained08:54Human-Level Performance: Manus vs. Human Intelligence11:45The Future of AI: Tool Use and Reasoning Capabilities14:59Data and Integration: The Key to AI Success17:46Manus in Action: Real-World Applications20:57The Importance of Real-Time Data in AI24:00The Role of Human Oversight in AI Development27:10The Future of API Integration with AI29:59Building Modular AI Systems for Enhanced Functionality44:22The Evolution of AI in Software Engineering48:06The Rise of Smaller, Efficient AI Models51:30The Future of AI in Everyday Devices57:09The Cost of AI Agents vs Human Labor01:01:30AI in Customer Service and Beyond01:10:07The Impact of AI on Freelance Developers01:17:35The Future of AI: Solving Humanity's Problems Keywords: AI, Manus, China, agentic models, Gaia benchmark, human-level performance, tool use, reasoning capabilities, software development, customer service, AI agents, future of work, coding, creativity, philosophical implications

Mark's Channel: https://www.youtube.com/@UCqLp0T6WBwxmuJdmo4mUW5A Databricks Events: https://lu.ma/okxq0bt1 https://lu.ma/n7p3gdx0    In this episode of The Generative AI Meetup Podcast, we dive deep into the evolving landscape of AI coding assistants and their real-world applications. Our hosts explore the capabilities of Cursor, Claude 3.7, and other frontier models, discussing how these tools are transforming development workflows while acknowledging their current limitations. The conversation covers "vibe coding" - how non-developers can now create functional applications through prompt engineering, the challenges of scaling AI-generated code to enterprise levels, and the exciting potential of agentic behaviors in coding assistants. We also examine recent announcements from major tech companies including OpenAI's GPT-4.5, Amazon's Alexa Plus integration with Claude, and the competitive dynamics driving innovation in the AI space. Whether you're a developer looking to enhance your productivity or simply curious about how AI is reshaping software development, this episode offers valuable insights into the current state and future direction of generative AI tools. AI #GenerativeAI #CodingAssistants #Claude #Cursor #VibeCoding #AIAgents #TechTrends #SoftwareDevelopment #SmartAssistants

Youtube: https://www.youtube.com/@GenerativeAIMeetup   https://www.anthropic.com/news/the-anthropic-economic-indexhttps://x.com/xai/status/1891699715298730482 - Grok 3http://x.com/karpathy/status/1891720635363254772https://news.microsoft.com/source/features/innovation/microsofts-majorana-1-chip-carves-new-path-for-quantum-computing/https://en.wikipedia.org/wiki/Majorana_fermionhttps://www.youtube.com/watch?v=88vEsL5tgDIGrok 3-Elon Musk new model- Seems to be a very brute-force approach (which is working)- Definitely a state of the art model- 40 dollars a month now- https://techcrunch.com/2025/02/19/x-doubles-its-premium-plan-prices-after-xai-releases-grok-3/- 100,000 GPUsMajorana- Uses a particle called the Majorana- https://www.howtopronounce.com/ettore-majorana- https://en.wikipedia.org/wiki/Ettore_Majorana- Ettore Majorana disappeared soon after theorizing this particle in 1937 he disappeared in 1938 although thought to be alive- A majorana is a fermion that has its own antiparticle - this was originally theorized in 1937- The particle has its own charge... and own antiparticle- it has a new topological state... which is a new state of matter- Has has 8 quibits- Is designed to scale to a million quibits- To break bitcoin you might need ~1500 to 5000 logical quibits- A quibit can be used to represent potentially infinite statesFigure AI - Helix- https://www.figure.ai/news/helix- Vision-Language-Action (VLA)- Full-upper-body control- Multi Robot collaboration- Pick up anything- Single neural network to build this, no specific fine tuning- https://www.figure.ai/news/helix- Has a slow System 2 and a fast system 1 neural network- system 2 -- 7B- system 1 -- 80M   Get ready to explore the frontiers of technology in this exciting episode! We unpack the latest breakthroughs driving us toward Artificial General Intelligence (AGI) and beyond, with a mix of mind-blowing advancements and thought-provoking discussions: Grok 3 Unveiled: Elon Musk’s xAI has dropped Grok 3, a powerhouse AI model dominating benchmarks and redefining what large language models can achieve. We dive into its stellar performance, its bold “no-filter” approach to information, and how it stacks up against heavyweights like Claude 3.5, OpenAI’s offerings, and Google’s latest.   Quantum Leap Forward: Microsoft’s Majorana-based quantum chip is here, promising to scale quantum computing to a million qubits. We simplify the tech behind it, explore its potential to transform everything from cryptography to drug design, and ponder what it means for simulating reality itself.   Robots in Action: Figure AI’s Helix brings us two humanoid robots teaming up to tackle chores like grocery unpacking. We marvel at their teamwork, laugh at their quirky moves, and discuss the hurdles and possibilities of general-purpose robotics in our everyday lives.   But it’s not all tech demos and breakthroughs. We wrestle with the big stuff too: Are we inching closer to AGI? How do quantum computing, AI, and robotics fuel each other’s progress? And what happens when unrestricted AI or quantum tech shakes up ethics—like privacy or security? Perfect for AI buffs, tech pros, or anyone curious about tomorrow, this episode blends sharp insights, lively debates, and a dash of humor. Jump in to stay ahead in the fast-moving world of generative AI! Listen now and join the discussion! Chapters:   0:00 - Introduction: The AGI RevolutionOpening remarks about how close we are to AGI and the recent breakthroughs 2:15 - Grok 3: The New AI BenchmarkDiscussion of Grok 3's capabilities and performance compared to other models 7:30 - Elon's 122-Day Data CenterHow Elon Musk converted an old Electrolux factory into a massive AI data center 11:45 - The Computing Power Behind GrokDetails about the 100,000+ GPUs and diesel generators powering Grok's training 15:20 - Why Grok "Just Feels Better"Analysis of what makes Grok 3 outperform other models in real-world use cases 19:40 - AI Guardrails: Freedom

Podcast: https://podcast.genaimeetup.com/ Youtube: https://www.youtube.com/@GenerativeAIMeetup In this episode, we explore Stanford's groundbreaking S1 paper, which introduces a technique to transform any language model into a reasoning model for just $6 in computation costs. We dive deep into the implications of this research, discussing budget forcing techniques, the true costs of AI development, and the philosophical limits of artificial intelligence across different domains - from mathematical reasoning to language translation. The conversation extends to broader questions about superhuman AI capabilities and the fundamental limitations in various fields like translation, history, and agriculture. Join us for an insightful discussion on the future of AI reasoning and its practical applications. 0:00 - Intro and weekly AI news overview1:06 - Introduction to the S1 paper from Stanford1:40 - Explanation of reasoning models vs single-shot models2:22 - Details of Stanford's S1 technique and QEN32B model3:00 - Cost comparison with other models ($6 training cost)4:04 - Discussion of model distillation technique5:00 - Budget forcing explanation6:33 - Story about building an AI stock research agent8:22 - Philosophical discussion on reasoning limits