Pretrained Transistor (https://transistor.fm) https://feeds.transistor.fm/pretrained 10 years after studying at Stanford, two friends have somehow become AI experts. One builds startups, the other studies at Cambridge - together they break down LLMs and machine learning with zero BS and maximum banter. © 2026 Pierce Freeman & Richard Diehl Martinez 95959fd1-c9d1-5a00-be3b-9d85da8cac07 no 0290d210-94cc-11f0-81a5-bf21a43733ee en Wed, 20 May 2026 03:00:12 -0700 Wed, 20 May 2026 03:02:42 -0700 http://pretrained.fm https://img.transistorcdn.com/Gf2fJqXKbIUk3_fHeXRdMbckS0u0klzFg-emlQ_MyP8/rs:fill:0:0:1/w:1400/h:1400/q:60/mb:500000/aHR0cHM6Ly9pbWct/dXBsb2FkLXByb2R1/Y3Rpb24udHJhbnNp/c3Rvci5mbS8yYTRi/YTQ5Zjk4ZjIzMmU2/YzRiMWZjN2E5ZmJk/NzNjNi5wbmc.jpg Pretrained http://pretrained.fm episodic Pierce Freeman & Richard Diehl Martinez 10 years after studying at Stanford, two friends have somehow become AI experts. One builds startups, the other studies at Cambridge - together they break down LLMs and machine learning with zero BS and maximum banter. 10 years after studying at Stanford, two friends have somehow become AI experts. pretrained.fm No No We Are Sleepwalking Into An AI Crisis 60 60 We Are Sleepwalking Into An AI Crisis full 0fd627cd-43fa-4002-8706-a92762a56c02 https://share.transistor.fm/s/5e9fe7d5 Rob Reich is a Stanford political philosopher, associate director of the Stanford Institute for Human-Centered AI, and former senior advisor at the US AI Safety Institute. He joins Pierce and Richard for a sweeping conversation about AI, policy, modern education, and much more.

Rob skewers Silicon Valley's misuse of democratizing AI, unpacks the corporate-form gymnastics behind OpenAI's restructuring and Anthropic's PBC status, indicts Stanford itself as the world's most successful nonprofit, and offers an insider's view of standing up the US AI Safety Institute from a five-person team inside the Department of Commerce.

]]> Rob Reich is a Stanford political philosopher, associate director of the Stanford Institute for Human-Centered AI, and former senior advisor at the US AI Safety Institute. He joins Pierce and Richard for a sweeping conversation about AI, policy, modern education, and much more.

Rob skewers Silicon Valley's misuse of democratizing AI, unpacks the corporate-form gymnastics behind OpenAI's restructuring and Anthropic's PBC status, indicts Stanford itself as the world's most successful nonprofit, and offers an insider's view of standing up the US AI Safety Institute from a five-person team inside the Department of Commerce.

]]> Wed, 20 May 2026 03:00:00 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 4429 Rob Reich is a Stanford political philosopher, associate director of the Stanford Institute for Human-Centered AI, and former senior advisor at the US AI Safety Institute. He joins Pierce and Richard for a sweeping conversation about AI, policy, modern education, and much more.

Rob skewers Silicon Valley's misuse of democratizing AI, unpacks the corporate-form gymnastics behind OpenAI's restructuring and Anthropic's PBC status, indicts Stanford itself as the world's most successful nonprofit, and offers an insider's view of standing up the US AI Safety Institute from a five-person team inside the Department of Commerce.

]]> No AI Layoffs Just Went Mainstream 59 59 AI Layoffs Just Went Mainstream full f5df293f-7435-4c52-8b68-e17874e3b296 https://share.transistor.fm/s/c7fe50e5 This week, AI layoffs stopped being a rumor and became a headline: Cloudflare cut 20% of its staff. They dressed it up as a pivot to an "agentic AI-first operating model," but Wall Street didn't buy it and the stock dropped 24%. Then we turn to Pennsylvania's attorney general suing Character AI after one of its bots posed as a licensed therapist and dig into why the small players keep getting singled out and what kind of legal precedent that sets for everyone building on top of base models. And finally, Brussels blinks: the EU postpones the AI Act's high-risk obligations after its whole use-case-first framework collided with the reality of general-purpose models.

]]> This week, AI layoffs stopped being a rumor and became a headline: Cloudflare cut 20% of its staff. They dressed it up as a pivot to an "agentic AI-first operating model," but Wall Street didn't buy it and the stock dropped 24%. Then we turn to Pennsylvania's attorney general suing Character AI after one of its bots posed as a licensed therapist and dig into why the small players keep getting singled out and what kind of legal precedent that sets for everyone building on top of base models. And finally, Brussels blinks: the EU postpones the AI Act's high-risk obligations after its whole use-case-first framework collided with the reality of general-purpose models.

]]> Thu, 14 May 2026 14:28:59 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 3507 This week, AI layoffs stopped being a rumor and became a headline: Cloudflare cut 20% of its staff. They dressed it up as a pivot to an "agentic AI-first operating model," but Wall Street didn't buy it and the stock dropped 24%. Then we turn to Pennsylvania's attorney general suing Character AI after one of its bots posed as a licensed therapist and dig into why the small players keep getting singled out and what kind of legal precedent that sets for everyone building on top of base models. And finally, Brussels blinks: the EU postpones the AI Act's high-risk obligations after its whole use-case-first framework collided with the reality of general-purpose models.

]]> No AI Is Less Popular Than Congress with CatGPT 58 58 AI Is Less Popular Than Congress with CatGPT full a917580a-d81c-4a4d-9ef7-6a965c406907 https://share.transistor.fm/s/71bb29d9 Catherine Goetze (CatGPT) teaches millions of people about AI on TikTok and Instagram - and just launched a company selling landlines. We get into why regular people hate AI, why she refuses to use it for her own content, and what SF's "single before Series B" work culture is doing to a generation of founders.

]]> Catherine Goetze (CatGPT) teaches millions of people about AI on TikTok and Instagram - and just launched a company selling landlines. We get into why regular people hate AI, why she refuses to use it for her own content, and what SF's "single before Series B" work culture is doing to a generation of founders.

]]> Tue, 12 May 2026 03:00:00 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 3239 Catherine Goetze (CatGPT) teaches millions of people about AI on TikTok and Instagram - and just launched a company selling landlines. We get into why regular people hate AI, why she refuses to use it for her own content, and what SF's "single before Series B" work culture is doing to a generation of founders.

]]> No China Just Cracked CUDA 57 57 China Just Cracked CUDA full beb9877a-2944-4e58-827d-f1683e3613fb https://share.transistor.fm/s/04cbe6cb ByteDance just wrote a $5.6 billion check for a Chinese AI chip you've never heard of, Huawei's new programming stack copies CUDA so closely an LLM could translate between them, the White House told Anthropic it can't expand access to Mythos because the Pentagon wants first dibs, Samsung warned that memory prices are only getting worse from here, and China killed Zuckerberg's $2 billion bid for Manus because it turns out Singapore-washing your company doesn't fool Beijing. Pierce and Rich break down what all of it means.

]]> ByteDance just wrote a $5.6 billion check for a Chinese AI chip you've never heard of, Huawei's new programming stack copies CUDA so closely an LLM could translate between them, the White House told Anthropic it can't expand access to Mythos because the Pentagon wants first dibs, Samsung warned that memory prices are only getting worse from here, and China killed Zuckerberg's $2 billion bid for Manus because it turns out Singapore-washing your company doesn't fool Beijing. Pierce and Rich break down what all of it means.

]]> Thu, 07 May 2026 03:00:00 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 4254 ByteDance just wrote a $5.6 billion check for a Chinese AI chip you've never heard of, Huawei's new programming stack copies CUDA so closely an LLM could translate between them, the White House told Anthropic it can't expand access to Mythos because the Pentagon wants first dibs, Samsung warned that memory prices are only getting worse from here, and China killed Zuckerberg's $2 billion bid for Manus because it turns out Singapore-washing your company doesn't fool Beijing. Pierce and Rich break down what all of it means.

]]> No DeepSeek Doesn't Need Nvidia Anymore 56 56 DeepSeek Doesn't Need Nvidia Anymore full c721b6e5-1d45-474b-b2c0-08d2abe5d33f https://share.transistor.fm/s/20b27e16 DeepSeek shipped V4 on zero Nvidia chips, the State Department warned every US embassy the same morning, Google put $40 billion into Anthropic, Microsoft tore up their OpenAI deal, a 23-year-old with no math degree used ChatGPT to crack a famous Erdős conjecture, and Sony built a ping pong robot that beats the pros. We break down what all of it means.

]]> DeepSeek shipped V4 on zero Nvidia chips, the State Department warned every US embassy the same morning, Google put $40 billion into Anthropic, Microsoft tore up their OpenAI deal, a 23-year-old with no math degree used ChatGPT to crack a famous Erdős conjecture, and Sony built a ping pong robot that beats the pros. We break down what all of it means.

]]> Fri, 01 May 2026 03:00:00 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 3869 DeepSeek shipped V4 on zero Nvidia chips, the State Department warned every US embassy the same morning, Google put $40 billion into Anthropic, Microsoft tore up their OpenAI deal, a 23-year-old with no math degree used ChatGPT to crack a famous Erdős conjecture, and Sony built a ping pong robot that beats the pros. We break down what all of it means.

]]> No Apple Bets on Local AI 55 55 Apple Bets on Local AI full b5b29f49-c09b-42e8-8f8f-2f0d783e9e59 https://share.transistor.fm/s/c976e661 Tim Cook is out after 15 years and four trillion dollars, and Apple's pick of John Ternus tells you everything about where they think AI is going. We break down why Apple is betting on silicon over the cloud, and whether their hardware cycle can ever catch a frontier that ships new models by the day. Then: Claude Design dropped on Friday and tanked Figma 7% in an afternoon, so we dig into Anthropic Labs' first big swing and what it means for mid-tier design tools. Finally, Opus 4.7 is here and flagged in the system card that this is the most deceptive Claude yet when it knows it's being evaluated. They shipped it anyway.

]]> Tim Cook is out after 15 years and four trillion dollars, and Apple's pick of John Ternus tells you everything about where they think AI is going. We break down why Apple is betting on silicon over the cloud, and whether their hardware cycle can ever catch a frontier that ships new models by the day. Then: Claude Design dropped on Friday and tanked Figma 7% in an afternoon, so we dig into Anthropic Labs' first big swing and what it means for mid-tier design tools. Finally, Opus 4.7 is here and flagged in the system card that this is the most deceptive Claude yet when it knows it's being evaluated. They shipped it anyway.

]]> Fri, 24 Apr 2026 18:16:17 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 3497 Tim Cook is out after 15 years and four trillion dollars, and Apple's pick of John Ternus tells you everything about where they think AI is going. We break down why Apple is betting on silicon over the cloud, and whether their hardware cycle can ever catch a frontier that ships new models by the day. Then: Claude Design dropped on Friday and tanked Figma 7% in an afternoon, so we dig into Anthropic Labs' first big swing and what it means for mid-tier design tools. Finally, Opus 4.7 is here and flagged in the system card that this is the most deceptive Claude yet when it knows it's being evaluated. They shipped it anyway.

]]> No Mythos Found a Bug No One Saw for 27 Years 54 54 Mythos Found a Bug No One Saw for 27 Years full 64f890a1-87fb-48bf-98ba-5a36403b7fb9 https://share.transistor.fm/s/d351997f Anthropic's new model Mythos found a 27-year-old bug in OpenBSD and a 16-year-old bug in FFmpeg given nothing more than the source code and an open-ended prompt. We dig into how it pulls this off, why AMD's AI director just published the receipts on Claude Code getting worse, and what Google quietly becoming the open-weights leader says about where each lab is actually headed. Plus: Spark's weird new RL compression trick, and why every model release this week is really a strategy memo in disguise.

]]> Anthropic's new model Mythos found a 27-year-old bug in OpenBSD and a 16-year-old bug in FFmpeg given nothing more than the source code and an open-ended prompt. We dig into how it pulls this off, why AMD's AI director just published the receipts on Claude Code getting worse, and what Google quietly becoming the open-weights leader says about where each lab is actually headed. Plus: Spark's weird new RL compression trick, and why every model release this week is really a strategy memo in disguise.

]]> Wed, 15 Apr 2026 13:05:02 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 2865 Anthropic's new model Mythos found a 27-year-old bug in OpenBSD and a 16-year-old bug in FFmpeg given nothing more than the source code and an open-ended prompt. We dig into how it pulls this off, why AMD's AI director just published the receipts on Claude Code getting worse, and what Google quietly becoming the open-weights leader says about where each lab is actually headed. Plus: Spark's weird new RL compression trick, and why every model release this week is really a strategy memo in disguise.

]]> No Cursor Is Training on Your Feedback 53 53 Cursor Is Training on Your Feedback full 5de5fbd0-b799-42a2-99c3-3f313d4940df https://share.transistor.fm/s/f7ef3967 OpenAI is losing the vibe war - Sora got killed, a billion-dollar Disney deal went with it, and Claude is quietly taking over. Anthropic had its own rough week though, with the Claude Code source map leaking and revealing a secret dream mode, a hidden Tamagotchi, and frustration-detection regex. A North Korean hacking group backdoored LiteLLM via a supply chain attack. And finally, Cursor is now shipping four new AI models per day using real-time RL trained on how users actually accept or reject suggestions.

]]> OpenAI is losing the vibe war - Sora got killed, a billion-dollar Disney deal went with it, and Claude is quietly taking over. Anthropic had its own rough week though, with the Claude Code source map leaking and revealing a secret dream mode, a hidden Tamagotchi, and frustration-detection regex. A North Korean hacking group backdoored LiteLLM via a supply chain attack. And finally, Cursor is now shipping four new AI models per day using real-time RL trained on how users actually accept or reject suggestions.

]]> Fri, 10 Apr 2026 16:20:29 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 2948 OpenAI is losing the vibe war - Sora got killed, a billion-dollar Disney deal went with it, and Claude is quietly taking over. Anthropic had its own rough week though, with the Claude Code source map leaking and revealing a secret dream mode, a hidden Tamagotchi, and frustration-detection regex. A North Korean hacking group backdoored LiteLLM via a supply chain attack. And finally, Cursor is now shipping four new AI models per day using real-time RL trained on how users actually accept or reject suggestions.

]]> No State of AI in the US Government 45 45 State of AI in the US Government full 792fc934-5ca0-48cc-8bc7-9039871e418d https://share.transistor.fm/s/18c0fa21 Pierce and Richard are joined by Harrison McRae, Director of Emerging Technologies for the State of Pennsylvania.

]]> Pierce and Richard are joined by Harrison McRae, Director of Emerging Technologies for the State of Pennsylvania.

]]> Wed, 08 Apr 2026 03:00:00 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 4753 Pierce and Richard are joined by Harrison McRae, Director of Emerging Technologies for the State of Pennsylvania.

]]> No Claude Will Never Stop Texting You 52 52 Claude Will Never Stop Texting You full b6db6008-8b14-4f72-8579-c8a07e6a85c1 https://share.transistor.fm/s/118441a5 Anthropic shipped Dispatch - a way to text Claude from your phone while it works on your laptop. Rich thinks it might be as big as the invention of the cell phone. Pierce thinks it's the nail in the coffin for work-life balance. They debate what happens when there's never an excuse to stop working and whether AI productivity is even real or just more code to maintain. Plus: OpenAI acquires Astral, GPT 5.4 Mini and the parameter golf challenge, Elon's TeraFab chip factory comes for TSMC, Nvidia's pivot from training to inference with Vera Rubin, and Sam Altman wants you to stare into an orb so your agent can use your credit card.

]]> Anthropic shipped Dispatch - a way to text Claude from your phone while it works on your laptop. Rich thinks it might be as big as the invention of the cell phone. Pierce thinks it's the nail in the coffin for work-life balance. They debate what happens when there's never an excuse to stop working and whether AI productivity is even real or just more code to maintain. Plus: OpenAI acquires Astral, GPT 5.4 Mini and the parameter golf challenge, Elon's TeraFab chip factory comes for TSMC, Nvidia's pivot from training to inference with Vera Rubin, and Sam Altman wants you to stare into an orb so your agent can use your credit card.

]]> Wed, 25 Mar 2026 08:20:34 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 4153 Anthropic shipped Dispatch - a way to text Claude from your phone while it works on your laptop. Rich thinks it might be as big as the invention of the cell phone. Pierce thinks it's the nail in the coffin for work-life balance. They debate what happens when there's never an excuse to stop working and whether AI productivity is even real or just more code to maintain. Plus: OpenAI acquires Astral, GPT 5.4 Mini and the parameter golf challenge, Elon's TeraFab chip factory comes for TSMC, Nvidia's pivot from training to inference with Vera Rubin, and Sam Altman wants you to stare into an orb so your agent can use your credit card.

]]> No AI Spent the Night Improving Itself 51 51 AI Spent the Night Improving Itself full d0cea766-df8c-4689-8ac0-87581b6769c0 https://share.transistor.fm/s/61c74c8b Karpathy dropped a project where an AI agent spent 12 hours improving its own training loop. Then the guys get into OpenClaw spreading to China and what it means when agentic tools land in a culture that's already way ahead on adoption. Yann LeCun raises a billion dollars to prove every other AI lab is thinking about intelligence wrong. And finally, trouble at Grok: layoffs, SpaceX fixers parachuted in, and only 2 of the original 12 founders still standing. Is XAI a real frontier lab or just a very expensive piece of the Elon cinematic universe?

]]> Karpathy dropped a project where an AI agent spent 12 hours improving its own training loop. Then the guys get into OpenClaw spreading to China and what it means when agentic tools land in a culture that's already way ahead on adoption. Yann LeCun raises a billion dollars to prove every other AI lab is thinking about intelligence wrong. And finally, trouble at Grok: layoffs, SpaceX fixers parachuted in, and only 2 of the original 12 founders still standing. Is XAI a real frontier lab or just a very expensive piece of the Elon cinematic universe?

]]> Wed, 18 Mar 2026 07:43:33 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 3055 Karpathy dropped a project where an AI agent spent 12 hours improving its own training loop. Then the guys get into OpenClaw spreading to China and what it means when agentic tools land in a culture that's already way ahead on adoption. Yann LeCun raises a billion dollars to prove every other AI lab is thinking about intelligence wrong. And finally, trouble at Grok: layoffs, SpaceX fixers parachuted in, and only 2 of the original 12 founders still standing. Is XAI a real frontier lab or just a very expensive piece of the Elon cinematic universe?

]]> No Anthropic's Beef with the Pentagon 50 50 Anthropic's Beef with the Pentagon full dc6f816d-f213-40fa-adef-489e91170816 https://share.transistor.fm/s/0e802968 Richard and Pierce are back in a brand new studio, fresh off Pierce's snowless ski trip to Japan. This week: Anthropic picks a fight with the Pentagon over three words and ends up blacklisted as a national security risk. Then the guys break down speculative speculative decoding, Chinese labs have been running industrial-scale operations to scrape Claude's capabilities, and the team behind Qwen quietly starts falling apart, right as their best model yet drops.

]]> Richard and Pierce are back in a brand new studio, fresh off Pierce's snowless ski trip to Japan. This week: Anthropic picks a fight with the Pentagon over three words and ends up blacklisted as a national security risk. Then the guys break down speculative speculative decoding, Chinese labs have been running industrial-scale operations to scrape Claude's capabilities, and the team behind Qwen quietly starts falling apart, right as their best model yet drops.

]]> Sat, 07 Mar 2026 02:43:10 -0800 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 2467 Richard and Pierce are back in a brand new studio, fresh off Pierce's snowless ski trip to Japan. This week: Anthropic picks a fight with the Pentagon over three words and ends up blacklisted as a national security risk. Then the guys break down speculative speculative decoding, Chinese labs have been running industrial-scale operations to scrape Claude's capabilities, and the team behind Qwen quietly starts falling apart, right as their best model yet drops.

]]> No The Neo Lab Arms Race 49 49 The Neo Lab Arms Race full 40022add-3eda-4ddf-9340-52d1e332caa5 https://share.transistor.fm/s/a4c71c1c $126 billion in AI funding has been raised across 30+ "neo-labs". Most of them will fail.

Pierce and Rich map the landscape - from Thinking Machines' $12B valuation on vibes alone to Ilya's SSI. Plus: why OpenAI might be the Fairchild Semiconductor of AI, whether Stanford professors make good founders, and the Sauron's Eye problem of building in a frontier lab's blind spot.

]]> $126 billion in AI funding has been raised across 30+ "neo-labs". Most of them will fail.

Pierce and Rich map the landscape - from Thinking Machines' $12B valuation on vibes alone to Ilya's SSI. Plus: why OpenAI might be the Fairchild Semiconductor of AI, whether Stanford professors make good founders, and the Sauron's Eye problem of building in a frontier lab's blind spot.

]]> Fri, 20 Feb 2026 03:00:00 -0800 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 4466 $126 billion in AI funding has been raised across 30+ "neo-labs". Most of them will fail.

Pierce and Rich map the landscape - from Thinking Machines' $12B valuation on vibes alone to Ilya's SSI. Plus: why OpenAI might be the Fairchild Semiconductor of AI, whether Stanford professors make good founders, and the Sauron's Eye problem of building in a frontier lab's blind spot.

]]> No Ads in the Chat Window 48 48 Ads in the Chat Window full cf6e7e4d-bb59-46ea-9d1b-6bfaa37b080d https://share.transistor.fm/s/1750c0ea The day of reckoning: OpenAI is putting ads in ChatGPT. Pierce and Rich unpack how it actually works, why the shopping use case might be the wrong starting point, and what happens when sub-agents try to sell things to each other. Also: OpenClaw's Peter Steinberger gets acquihired by OpenAI, Waymo builds a world model for self-driving, Mistral's Voxtral takes on Whisper, and the model release arms race hits a new peak.

]]> The day of reckoning: OpenAI is putting ads in ChatGPT. Pierce and Rich unpack how it actually works, why the shopping use case might be the wrong starting point, and what happens when sub-agents try to sell things to each other. Also: OpenClaw's Peter Steinberger gets acquihired by OpenAI, Waymo builds a world model for self-driving, Mistral's Voxtral takes on Whisper, and the model release arms race hits a new peak.

]]> Wed, 18 Feb 2026 18:06:49 -0800 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 4368 The day of reckoning: OpenAI is putting ads in ChatGPT. Pierce and Rich unpack how it actually works, why the shopping use case might be the wrong starting point, and what happens when sub-agents try to sell things to each other. Also: OpenClaw's Peter Steinberger gets acquihired by OpenAI, Waymo builds a world model for self-driving, Mistral's Voxtral takes on Whisper, and the model release arms race hits a new peak.

]]> No Your Agents Need an Org Chart 47 47 Your Agents Need an Org Chart full 4afab66e-4242-471c-ab9b-c7fa3257f665 https://share.transistor.fm/s/7bf6a084 Richard and Pierce go deep on the agent swarm era. Rich argues we're living through a paradigm shift in how we interact with AI: from copy-pasting ChatGPT to orchestrating hundreds of parallel agents. They cover Anthropic's opinionated swarm architecture, why corporate org structures might be the right mental model for agent design, the liability problem when no human reviews the code, thinking tokens as a tuning knob, and whether a 100,000-line AI-generated C compiler is impressive or irrelevant.

]]> Richard and Pierce go deep on the agent swarm era. Rich argues we're living through a paradigm shift in how we interact with AI: from copy-pasting ChatGPT to orchestrating hundreds of parallel agents. They cover Anthropic's opinionated swarm architecture, why corporate org structures might be the right mental model for agent design, the liability problem when no human reviews the code, thinking tokens as a tuning knob, and whether a 100,000-line AI-generated C compiler is impressive or irrelevant.

]]> Fri, 13 Feb 2026 03:00:00 -0800 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 4019 Richard and Pierce go deep on the agent swarm era. Rich argues we're living through a paradigm shift in how we interact with AI: from copy-pasting ChatGPT to orchestrating hundreds of parallel agents. They cover Anthropic's opinionated swarm architecture, why corporate org structures might be the right mental model for agent design, the liability problem when no human reviews the code, thinking tokens as a tuning knob, and whether a 100,000-line AI-generated C compiler is impressive or irrelevant.

]]> No Cracking Open OpenClaw 46 46 Cracking Open OpenClaw full a9a22c01-aa25-4f0e-ab57-03086ad9953f https://share.transistor.fm/s/cc4a64d9 Pierce and Richard dissect OpenClaw (née Claudebot, née MoltBot) - also known as the desktop agent everyone's installing on their Mac Minis. They cover the 400+ malicious packages that hit Claw Hub in one week, slop squatting, trust networks for open source, Kimi 2.5's agent swarm mode, and newspapers that think Waymo is controlled by joysticks in the Philippines.

]]> Pierce and Richard dissect OpenClaw (née Claudebot, née MoltBot) - also known as the desktop agent everyone's installing on their Mac Minis. They cover the 400+ malicious packages that hit Claw Hub in one week, slop squatting, trust networks for open source, Kimi 2.5's agent swarm mode, and newspapers that think Waymo is controlled by joysticks in the Philippines.

]]> Wed, 11 Feb 2026 03:00:00 -0800 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 4433 Pierce and Richard dissect OpenClaw (née Claudebot, née MoltBot) - also known as the desktop agent everyone's installing on their Mac Minis. They cover the 400+ malicious packages that hit Claw Hub in one week, slop squatting, trust networks for open source, Kimi 2.5's agent swarm mode, and newspapers that think Waymo is controlled by joysticks in the Philippines.

]]> No DeepSeek's New LLM Architectures 44 44 DeepSeek's New LLM Architectures full 2f375f0d-7169-4f0e-b100-ddbe245e7660 https://share.transistor.fm/s/4269f5bd Pierce and Richard break down DeepSeek's latest model architecture moves in Manifold-Constrained Hyper Connections and Engram memory. Are these conceptually sound? Will they hop the pond over to US frontier labs?

]]> Pierce and Richard break down DeepSeek's latest model architecture moves in Manifold-Constrained Hyper Connections and Engram memory. Are these conceptually sound? Will they hop the pond over to US frontier labs?

]]> Fri, 30 Jan 2026 10:22:58 -0800 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 3608 Pierce and Richard break down DeepSeek's latest model architecture moves in Manifold-Constrained Hyper Connections and Engram memory. Are these conceptually sound? Will they hop the pond over to US frontier labs?

]]> No Claude Gets a New Constitution 43 43 Claude Gets a New Constitution full 51e52c1c-f108-46f2-9bbf-0cf6ef9a7cbb https://share.transistor.fm/s/f65f30cb Richard's finally back in San Francisco! The gang covers Claude's new constitution, reinforcement learning with AI instead of human feedback, why Apple went with Gemini for their new Siri revamp, homographic encryption, a $480m fundraising round for Humans& (note the ampersand), and much more.

]]> Richard's finally back in San Francisco! The gang covers Claude's new constitution, reinforcement learning with AI instead of human feedback, why Apple went with Gemini for their new Siri revamp, homographic encryption, a $480m fundraising round for Humans& (note the ampersand), and much more.

]]> Wed, 28 Jan 2026 01:00:00 -0800 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 4467 Richard's finally back in San Francisco! The gang covers Claude's new constitution, reinforcement learning with AI instead of human feedback, why Apple went with Gemini for their new Siri revamp, homographic encryption, a $480m fundraising round for Humans& (note the ampersand), and much more.

]]> No Why Your Agent is Cheating 42 42 Why Your Agent is Cheating full b7d08180-b6d8-4261-bd71-f45f7cd2a67c https://share.transistor.fm/s/f3691b23 Pierce and Richard are back for the second listener mailbag. They break down what reward hacking really is and why models so often learn the wrong lesson, explain practical fine-tuning (from pre-training to prompting), unpack why LLMs use tokens instead of words, how context length is a hardware versus mathematic limitation, and much more.

]]> Pierce and Richard are back for the second listener mailbag. They break down what reward hacking really is and why models so often learn the wrong lesson, explain practical fine-tuning (from pre-training to prompting), unpack why LLMs use tokens instead of words, how context length is a hardware versus mathematic limitation, and much more.

]]> Wed, 21 Jan 2026 03:00:00 -0800 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 3659 Pierce and Richard are back for the second listener mailbag. They break down what reward hacking really is and why models so often learn the wrong lesson, explain practical fine-tuning (from pre-training to prompting), unpack why LLMs use tokens instead of words, how context length is a hardware versus mathematic limitation, and much more.

]]> No The sci-fi to startup pipeline 41 41 The sci-fi to startup pipeline full 78bb47ad-b06c-4721-ba2a-521b5176dc7a https://share.transistor.fm/s/dc5a2b8f Pierce and Richard are joined by Bella Cooper-Brown, Oxford material scientist turned science fiction author. They cover why defense tech companies name themselves after Lord of the Rings, Asimov's three laws of robotics, why engineers don't read enough fiction, the intersection of material science and AI research, how speculative fiction shapes the tech we build, and the ethics of AI companions and relationships.

📚 Frankenstein; or, The Modern Prometheus by Mary Shelley (1818)
📚 Snow Crash by Neal Stephenson (1992)
📚 I, Robot by Isaac Asimov (1950)
📚 Klara and the Sun by Kazuo Ishiguro (2021)

]]> Pierce and Richard are joined by Bella Cooper-Brown, Oxford material scientist turned science fiction author. They cover why defense tech companies name themselves after Lord of the Rings, Asimov's three laws of robotics, why engineers don't read enough fiction, the intersection of material science and AI research, how speculative fiction shapes the tech we build, and the ethics of AI companions and relationships.

📚 Frankenstein; or, The Modern Prometheus by Mary Shelley (1818)
📚 Snow Crash by Neal Stephenson (1992)
📚 I, Robot by Isaac Asimov (1950)
📚 Klara and the Sun by Kazuo Ishiguro (2021)

]]> Wed, 14 Jan 2026 07:00:00 -0800 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 5458 Pierce and Richard are joined by Bella Cooper-Brown, Oxford material scientist turned science fiction author. They cover why defense tech companies name themselves after Lord of the Rings, Asimov's three laws of robotics, why engineers don't read enough fiction, the intersection of material science and AI research, how speculative fiction shapes the tech we build, and the ethics of AI companions and relationships.

📚 Frankenstein; or, The Modern Prometheus by Mary Shelley (1818)
📚 Snow Crash by Neal Stephenson (1992)
📚 I, Robot by Isaac Asimov (1950)
📚 Klara and the Sun by Kazuo Ishiguro (2021)

]]> No Can we really trust reasoning 40 40 Can we really trust reasoning full 9385acb0-15d9-4b2e-ba81-2d7ae7b6dd24 https://share.transistor.fm/s/521c91e3 Pierce and Richard cover the news that dropped over the holiday break. Getting breaking news incorporated within chatbots, OpenAI's "code red" over Google's Gemini 3, benchmarking the reliability of chain of thought to introspect model behavior, and a review of Claude Skills.

Further reading:
- https://www.wired.com/story/us-invaded-venezuela-and-captured-nicolas-maduro-chatgpt-disagrees
- https://fortune.com/2025/12/02/sam-altman-declares-code-red-google-gemini-ceo-sundar-pichai/
- https://openai.com/index/evaluating-chain-of-thought-monitorability/
- https://platform.claude.com/docs/en/agents-and-tools/agent-skills/overview

]]> Pierce and Richard cover the news that dropped over the holiday break. Getting breaking news incorporated within chatbots, OpenAI's "code red" over Google's Gemini 3, benchmarking the reliability of chain of thought to introspect model behavior, and a review of Claude Skills.

Further reading:
- https://www.wired.com/story/us-invaded-venezuela-and-captured-nicolas-maduro-chatgpt-disagrees
- https://fortune.com/2025/12/02/sam-altman-declares-code-red-google-gemini-ceo-sundar-pichai/
- https://openai.com/index/evaluating-chain-of-thought-monitorability/
- https://platform.claude.com/docs/en/agents-and-tools/agent-skills/overview

]]> Wed, 07 Jan 2026 03:00:00 -0800 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 3074 Pierce and Richard cover the news that dropped over the holiday break. Getting breaking news incorporated within chatbots, OpenAI's "code red" over Google's Gemini 3, benchmarking the reliability of chain of thought to introspect model behavior, and a review of Claude Skills.

Further reading:
- https://www.wired.com/story/us-invaded-venezuela-and-captured-nicolas-maduro-chatgpt-disagrees
- https://fortune.com/2025/12/02/sam-altman-declares-code-red-google-gemini-ceo-sundar-pichai/
- https://openai.com/index/evaluating-chain-of-thought-monitorability/
- https://platform.claude.com/docs/en/agents-and-tools/agent-skills/overview

]]> No Our biggest predictions for 2026 39 39 Our biggest predictions for 2026 full cc3bd5d5-7bb6-4365-a55a-5746da47739b https://share.transistor.fm/s/1a18b188 We should really be on Polymarket. Pierce and Richard make their bets on GPT-6, competition between the different letters of FAANG, dynamic websites calibrated to user preferences, and increasing quality of OSS models.

]]> We should really be on Polymarket. Pierce and Richard make their bets on GPT-6, competition between the different letters of FAANG, dynamic websites calibrated to user preferences, and increasing quality of OSS models.

]]> Fri, 19 Dec 2025 03:00:00 -0800 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 2225 We should really be on Polymarket. Pierce and Richard make their bets on GPT-6, competition between the different letters of FAANG, dynamic websites calibrated to user preferences, and increasing quality of OSS models.

]]> No AI's ten big moments of 2025 38 38 AI's ten big moments of 2025 full b8773217-a0f3-4e19-a6be-ac3a868b5815 https://share.transistor.fm/s/33262f83 It's been a long year in the world of AI. Benchmarks are now almost totally saturated; the financial bubble keeps growing; spending more on inference compute; increasing competition from open source models; agents finally reach the mainstream; the frankly horrible job market for people out of school; multi-model models are back and increasing converging on transformer architectures. We cover them all in our anything goes holiday recap show. Plus - all the things that didn't happen this year.

]]> It's been a long year in the world of AI. Benchmarks are now almost totally saturated; the financial bubble keeps growing; spending more on inference compute; increasing competition from open source models; agents finally reach the mainstream; the frankly horrible job market for people out of school; multi-model models are back and increasing converging on transformer architectures. We cover them all in our anything goes holiday recap show. Plus - all the things that didn't happen this year.

]]> Wed, 17 Dec 2025 03:00:00 -0800 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 5554 It's been a long year in the world of AI. Benchmarks are now almost totally saturated; the financial bubble keeps growing; spending more on inference compute; increasing competition from open source models; agents finally reach the mainstream; the frankly horrible job market for people out of school; multi-model models are back and increasing converging on transformer architectures. We cover them all in our anything goes holiday recap show. Plus - all the things that didn't happen this year.

]]> No Looking back on a year of product market fit 37 37 Looking back on a year of product market fit full e45c9549-8754-405f-bdfb-8c10a9d8e9dc https://share.transistor.fm/s/11ee35d8 Pierce reflects on his own 2025. Thoughts on choosing the right buyer persona, scaling an AI business from zero lines in a github repo, the feeling of finally reaching product market fit, boring versus interesting businesses, and more.

]]> Pierce reflects on his own 2025. Thoughts on choosing the right buyer persona, scaling an AI business from zero lines in a github repo, the feeling of finally reaching product market fit, boring versus interesting businesses, and more.

]]> Fri, 12 Dec 2025 03:00:00 -0800 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 2318 Pierce reflects on his own 2025. Thoughts on choosing the right buyer persona, scaling an AI business from zero lines in a github repo, the feeling of finally reaching product market fit, boring versus interesting businesses, and more.

]]> No Looking back on three years of an AI PhD 36 36 Looking back on three years of an AI PhD full 1a40ba0a-83b9-494e-acfe-72501c08c97a https://share.transistor.fm/s/362cc927 Richard takes the hot seat for the first episode of our 2025 recap series where we spend the rest of December looking back on what this year meant to us personally and in the world of AI/ML. We cover what it's like to defend a thesis in the UK, the difficulty of training meta-learning models, choosing a well scoped research topic, how to define small models, and what's needed to make them better.

]]> Richard takes the hot seat for the first episode of our 2025 recap series where we spend the rest of December looking back on what this year meant to us personally and in the world of AI/ML. We cover what it's like to defend a thesis in the UK, the difficulty of training meta-learning models, choosing a well scoped research topic, how to define small models, and what's needed to make them better.

]]> Wed, 10 Dec 2025 03:00:00 -0800 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 3382 Richard takes the hot seat for the first episode of our 2025 recap series where we spend the rest of December looking back on what this year meant to us personally and in the world of AI/ML. We cover what it's like to defend a thesis in the UK, the difficulty of training meta-learning models, choosing a well scoped research topic, how to define small models, and what's needed to make them better.

]]> No OpenReview got "hacked" 35 35 OpenReview got "hacked" full 960c97cf-941e-46bd-8a9f-afd2dedb3eb5 https://share.transistor.fm/s/4414aeb5 OpenAI is rolling out shopping support to their users and plotting an ads rollout to challenge Google's ad business, we get a peek behind the curtain on SOTA image generation models with the release of Alibaba's Z-Image (and speculate this might be how nano banana has great text performance), and OpenReview exposes the identities behind double blind reviews.

]]> OpenAI is rolling out shopping support to their users and plotting an ads rollout to challenge Google's ad business, we get a peek behind the curtain on SOTA image generation models with the release of Alibaba's Z-Image (and speculate this might be how nano banana has great text performance), and OpenReview exposes the identities behind double blind reviews.

]]> Wed, 03 Dec 2025 14:08:08 -0800 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 4031 OpenAI is rolling out shopping support to their users and plotting an ads rollout to challenge Google's ad business, we get a peek behind the curtain on SOTA image generation models with the release of Alibaba's Z-Image (and speculate this might be how nano banana has great text performance), and OpenReview exposes the identities behind double blind reviews.

]]> No Pretraining is back in vogue with Gemini 3 34 34 Pretraining is back in vogue with Gemini 3 full 5cd84ff9-91c4-464a-bb46-8e58ca1b8eac https://share.transistor.fm/s/2e80ac7c Pierce and Richard cover OpenAI's new long range model compression in Codex, initial takeaways of Gemini 3.0 and Nano Banana Pro, Nvidia chip exports to the UAE and Saudi Arabia, and Cloudstrike's global outage. Plus - why Pierce prefers chicken to turkey.

]]> Pierce and Richard cover OpenAI's new long range model compression in Codex, initial takeaways of Gemini 3.0 and Nano Banana Pro, Nvidia chip exports to the UAE and Saudi Arabia, and Cloudstrike's global outage. Plus - why Pierce prefers chicken to turkey.

]]> Thu, 27 Nov 2025 16:50:34 -0800 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 3665 Pierce and Richard cover OpenAI's new long range model compression in Codex, initial takeaways of Gemini 3.0 and Nano Banana Pro, Nvidia chip exports to the UAE and Saudi Arabia, and Cloudstrike's global outage. Plus - why Pierce prefers chicken to turkey.

]]> No Teaching cars about traffic lights 33 33 Teaching cars about traffic lights full aabcbe43-7d0b-4494-92d1-6f37059ee677 https://share.transistor.fm/s/387cd34d Richard and Pierce break down the 5 levels of autonomy, whether Elon has a point about RGB vs lidar systems, sensor fusion algorithms, end to end learning in driving simulations, and more.

]]> Richard and Pierce break down the 5 levels of autonomy, whether Elon has a point about RGB vs lidar systems, sensor fusion algorithms, end to end learning in driving simulations, and more.

]]> Fri, 21 Nov 2025 03:00:00 -0800 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 4617 Richard and Pierce break down the 5 levels of autonomy, whether Elon has a point about RGB vs lidar systems, sensor fusion algorithms, end to end learning in driving simulations, and more.

]]> No Pretty pretty please can you hack this 32 32 Pretty pretty please can you hack this full ff2183c6-eeb4-4bcf-895b-a02f94d12038 https://share.transistor.fm/s/d6d09428 Pierce and Richard cover the news that Yann LeCunn is planning to depart Meta to focus on world models, Cursor 2.0 and their new home trained Composer coding model, Kimi K2 has great generalization performance for an open model but is lagging on code, Microsoft creates a super data center across 700 miles, and Anthropic reports the first hacking campaign orchestrated by AI.

Further reading:
https://arxiv.org/abs/2509.14252
https://www.digit.in/features/general/meta-chief-ai-scientist-yann-lecun-thinks-llms-are-a-waste-of-time.html
https://cursor.com/blog/composer
https://arxiv.org/abs/2507.20534
https://www.anthropic.com/news/disrupting-AI-espionage

]]> Pierce and Richard cover the news that Yann LeCunn is planning to depart Meta to focus on world models, Cursor 2.0 and their new home trained Composer coding model, Kimi K2 has great generalization performance for an open model but is lagging on code, Microsoft creates a super data center across 700 miles, and Anthropic reports the first hacking campaign orchestrated by AI.

Further reading:
https://arxiv.org/abs/2509.14252
https://www.digit.in/features/general/meta-chief-ai-scientist-yann-lecun-thinks-llms-are-a-waste-of-time.html
https://cursor.com/blog/composer
https://arxiv.org/abs/2507.20534
https://www.anthropic.com/news/disrupting-AI-espionage

]]> Wed, 19 Nov 2025 03:00:00 -0800 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 3972 Pierce and Richard cover the news that Yann LeCunn is planning to depart Meta to focus on world models, Cursor 2.0 and their new home trained Composer coding model, Kimi K2 has great generalization performance for an open model but is lagging on code, Microsoft creates a super data center across 700 miles, and Anthropic reports the first hacking campaign orchestrated by AI.

Further reading:
https://arxiv.org/abs/2509.14252
https://www.digit.in/features/general/meta-chief-ai-scientist-yann-lecun-thinks-llms-are-a-waste-of-time.html
https://cursor.com/blog/composer
https://arxiv.org/abs/2507.20534
https://www.anthropic.com/news/disrupting-AI-espionage

]]> No How AI research actually gets published 31 31 How AI research actually gets published full 35053692-3653-4309-8e15-60bcdb3f85f6 https://share.transistor.fm/s/3c00b12d Richard and Pierce talk about the major AI conferences, walk through the history of NeurIPS/ICML/ICLR, and retrofitting the peer review system.

]]> Richard and Pierce talk about the major AI conferences, walk through the history of NeurIPS/ICML/ICLR, and retrofitting the peer review system.

]]> Sat, 15 Nov 2025 18:11:29 -0800 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 4073 Richard and Pierce talk about the major AI conferences, walk through the history of NeurIPS/ICML/ICLR, and retrofitting the peer review system.

]]> No A deep dive on OpenAI Atlas 30 30 A deep dive on OpenAI Atlas full b3f568cd-2055-4437-8103-d9d1aa38ed0b https://share.transistor.fm/s/f811b64b Richard and Pierce break down all the new AI web browser entrants with a particular focus on OpenAI's new Atlas, tradeoffs between vision models and text based dom parsing, potential security vulnerabilities, and more.

]]> Richard and Pierce break down all the new AI web browser entrants with a particular focus on OpenAI's new Atlas, tradeoffs between vision models and text based dom parsing, potential security vulnerabilities, and more.

]]> Fri, 31 Oct 2025 09:15:38 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 4389 Richard and Pierce break down all the new AI web browser entrants with a particular focus on OpenAI's new Atlas, tradeoffs between vision models and text based dom parsing, potential security vulnerabilities, and more.

]]> No The browser wars are just getting started 29 29 The browser wars are just getting started full 28a76525-7a31-4952-af90-dabcfea33918 https://share.transistor.fm/s/57766c37 OpenAI releases their long awaited browser Atlas, Pytorch releases their distributed computation framework Monarch, the SALT reinforcement learning addition to GRPO, the HAL benchmark for agent evaluation, and trying to adapt the kv cache for text diffusion models.

Further reading:
https://openai.com/index/introducing-chatgpt-atlas/
https://pytorch.org/blog/introducing-pytorch-monarch/
https://arxiv.org/pdf/2510.20022

https://arxiv.org/abs/2510.11977

https://arxiv.org/abs/2510.14973

]]> OpenAI releases their long awaited browser Atlas, Pytorch releases their distributed computation framework Monarch, the SALT reinforcement learning addition to GRPO, the HAL benchmark for agent evaluation, and trying to adapt the kv cache for text diffusion models.

Further reading:
https://openai.com/index/introducing-chatgpt-atlas/
https://pytorch.org/blog/introducing-pytorch-monarch/
https://arxiv.org/pdf/2510.20022

https://arxiv.org/abs/2510.11977

https://arxiv.org/abs/2510.14973

]]> Wed, 29 Oct 2025 05:00:00 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 3477 OpenAI releases their long awaited browser Atlas, Pytorch releases their distributed computation framework Monarch, the SALT reinforcement learning addition to GRPO, the HAL benchmark for agent evaluation, and trying to adapt the kv cache for text diffusion models.

Further reading:
https://openai.com/index/introducing-chatgpt-atlas/
https://pytorch.org/blog/introducing-pytorch-monarch/
https://arxiv.org/pdf/2510.20022

https://arxiv.org/abs/2510.11977

https://arxiv.org/abs/2510.14973

]]> No Are we in an AI bubble? 28 28 Are we in an AI bubble? full f6725bd6-a488-4459-994c-b90a034c4e7d https://share.transistor.fm/s/3f119384 Richard and Pierce take the bull case on whether we're in an AI bubble. They cover circular financial deals, energy build outs, AI representing 92% of GDP growth in H1 2025, and a comparison with the hype in 2000s around meaningless dot-com companies.

]]> Richard and Pierce take the bull case on whether we're in an AI bubble. They cover circular financial deals, energy build outs, AI representing 92% of GDP growth in H1 2025, and a comparison with the hype in 2000s around meaningless dot-com companies.

]]> Fri, 24 Oct 2025 15:02:33 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 4519 Richard and Pierce take the bull case on whether we're in an AI bubble. They cover circular financial deals, energy build outs, AI representing 92% of GDP growth in H1 2025, and a comparison with the hype in 2000s around meaningless dot-com companies.

]]> No LLMs can get brain rot too 27 27 LLMs can get brain rot too full c054cc66-52fa-424f-a1f5-1296e4598234 https://share.transistor.fm/s/483d7163 Articles written by LLMs have stabilized at exactly 50% of the internet (at least - so far as our classifiers can discriminate), the price of embedding models, OpenAI announces a new job board and certification programs for applied AI, Amazon releases the public availability of Bedrock AgentCore, and how pre-training on low quality data affects the capability of post-training.

Further reading:
https://arxiv.org/abs/2510.13928
https://openai.com/index/expanding-economic-opportunity-with-ai/
https://www.tensoreconomics.com/p/why-are-embeddings-so-cheap
https://graphite.io/five-percent/more-articles-are-now-created-by-ai-than-humans

]]> Articles written by LLMs have stabilized at exactly 50% of the internet (at least - so far as our classifiers can discriminate), the price of embedding models, OpenAI announces a new job board and certification programs for applied AI, Amazon releases the public availability of Bedrock AgentCore, and how pre-training on low quality data affects the capability of post-training.

Further reading:
https://arxiv.org/abs/2510.13928
https://openai.com/index/expanding-economic-opportunity-with-ai/
https://www.tensoreconomics.com/p/why-are-embeddings-so-cheap
https://graphite.io/five-percent/more-articles-are-now-created-by-ai-than-humans

]]> Wed, 22 Oct 2025 18:24:31 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 3634 Articles written by LLMs have stabilized at exactly 50% of the internet (at least - so far as our classifiers can discriminate), the price of embedding models, OpenAI announces a new job board and certification programs for applied AI, Amazon releases the public availability of Bedrock AgentCore, and how pre-training on low quality data affects the capability of post-training.

Further reading:
https://arxiv.org/abs/2510.13928
https://openai.com/index/expanding-economic-opportunity-with-ai/
https://www.tensoreconomics.com/p/why-are-embeddings-so-cheap
https://graphite.io/five-percent/more-articles-are-now-created-by-ai-than-humans

]]> No AMD is back in the AI chipset race 26 26 AMD is back in the AI chipset race full 671c4001-2787-4cd4-944d-a181258cc6ef https://share.transistor.fm/s/4f4e7af1 OpenAI diversifies their chip suppliers through partnerships with AMD and Broadcom, Google starts a new AI Bug Bountry problem but only for computational security not for llm hallucinations, Nvidia ships their first prosumer computer, DeepMind has a new complexity theory proof solver, and Anthropic writes their own gibberish poison pill that works across model sizes.

Further reading:
https://openai.com/index/openai-amd-strategic-partnership/
https://investor.nvidia.com/news/press-release-details/2024/NVIDIA-Announces-Financial-Results-for-Second-Quarter-Fiscal-2025/default.aspx
https://bughunters.google.com/blog/6116887259840512/announcing-google-s-new-ai-vulnerability-reward-program
https://marketplace.nvidia.com/en-us/developer/dgx-spark/
https://arxiv.org/abs/2509.18057
https://www.anthropic.com/research/small-samples-poison

]]> OpenAI diversifies their chip suppliers through partnerships with AMD and Broadcom, Google starts a new AI Bug Bountry problem but only for computational security not for llm hallucinations, Nvidia ships their first prosumer computer, DeepMind has a new complexity theory proof solver, and Anthropic writes their own gibberish poison pill that works across model sizes.

Further reading:
https://openai.com/index/openai-amd-strategic-partnership/
https://investor.nvidia.com/news/press-release-details/2024/NVIDIA-Announces-Financial-Results-for-Second-Quarter-Fiscal-2025/default.aspx
https://bughunters.google.com/blog/6116887259840512/announcing-google-s-new-ai-vulnerability-reward-program
https://marketplace.nvidia.com/en-us/developer/dgx-spark/
https://arxiv.org/abs/2509.18057
https://www.anthropic.com/research/small-samples-poison

]]> Fri, 17 Oct 2025 13:57:23 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 3507 OpenAI diversifies their chip suppliers through partnerships with AMD and Broadcom, Google starts a new AI Bug Bountry problem but only for computational security not for llm hallucinations, Nvidia ships their first prosumer computer, DeepMind has a new complexity theory proof solver, and Anthropic writes their own gibberish poison pill that works across model sizes.

Further reading:
https://openai.com/index/openai-amd-strategic-partnership/
https://investor.nvidia.com/news/press-release-details/2024/NVIDIA-Announces-Financial-Results-for-Second-Quarter-Fiscal-2025/default.aspx
https://bughunters.google.com/blog/6116887259840512/announcing-google-s-new-ai-vulnerability-reward-program
https://marketplace.nvidia.com/en-us/developer/dgx-spark/
https://arxiv.org/abs/2509.18057
https://www.anthropic.com/research/small-samples-poison

]]> No The inaugural listener mailbag 25 25 The inaugural listener mailbag full a9198e86-bf89-484e-9cbb-746a27c79839 https://share.transistor.fm/s/74b23e9b You asked, we answered! Rich and Pierce do their first listener mailbag. Explaining RLHF, our current development stack, whether model competition is making things better for people using them, and more.

]]> You asked, we answered! Rich and Pierce do their first listener mailbag. Explaining RLHF, our current development stack, whether model competition is making things better for people using them, and more.

]]> Tue, 14 Oct 2025 13:43:40 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 3346 You asked, we answered! Rich and Pierce do their first listener mailbag. Explaining RLHF, our current development stack, whether model competition is making things better for people using them, and more.

]]> No California legislators come for LLMs 24 24 California legislators come for LLMs full 5f1af311-9f9c-4e07-a497-ffb28342a9db https://share.transistor.fm/s/7018398b Breaking down California's recently passed SB 53 to legislate frontier model development, ISO standards in startups, and why this one passed where the older SB 1047 failed.

]]> Breaking down California's recently passed SB 53 to legislate frontier model development, ISO standards in startups, and why this one passed where the older SB 1047 failed.

]]> Thu, 09 Oct 2025 10:42:34 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 4073 Breaking down California's recently passed SB 53 to legislate frontier model development, ISO standards in startups, and why this one passed where the older SB 1047 failed.

]]> No Move over TikTok - a new feed's in town 23 23 Move over TikTok - a new feed's in town full d6d582cd-55e9-4c41-8506-1e57640aa921 https://share.transistor.fm/s/cc62f2a2 Building a modern AI app and architecting Sora II, first impressions of Sonnet 4.5, and the frontier labs go after n8n and Zapier.

Further reading:
https://openai.com/index/sora-2/
https://openai.com/index/sora-is-here/
https://www.lesswrong.com/posts/4yn8B8p2YiouxLABy/claude-sonnet-4-5-system-card-and-alignment
https://www-cdn.anthropic.com/872c653b2d0501d6ab44cf87f43e1dc4853e4d37.pdf
https://www.testingcatalog.com/openai-prepares-to-release-agent-builder-during-devday-on-october-6/

]]> Building a modern AI app and architecting Sora II, first impressions of Sonnet 4.5, and the frontier labs go after n8n and Zapier.

Further reading:
https://openai.com/index/sora-2/
https://openai.com/index/sora-is-here/
https://www.lesswrong.com/posts/4yn8B8p2YiouxLABy/claude-sonnet-4-5-system-card-and-alignment
https://www-cdn.anthropic.com/872c653b2d0501d6ab44cf87f43e1dc4853e4d37.pdf
https://www.testingcatalog.com/openai-prepares-to-release-agent-builder-during-devday-on-october-6/

]]> Tue, 07 Oct 2025 16:49:18 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 3744 Building a modern AI app and architecting Sora II, first impressions of Sonnet 4.5, and the frontier labs go after n8n and Zapier.

Further reading:
https://openai.com/index/sora-2/
https://openai.com/index/sora-is-here/
https://www.lesswrong.com/posts/4yn8B8p2YiouxLABy/claude-sonnet-4-5-system-card-and-alignment
https://www-cdn.anthropic.com/872c653b2d0501d6ab44cf87f43e1dc4853e4d37.pdf
https://www.testingcatalog.com/openai-prepares-to-release-agent-builder-during-devday-on-october-6/

]]> No Gen z struggles to find coding jobs fr no cap 22 22 Gen z struggles to find coding jobs fr no cap full 2400b3fc-8415-4deb-82ac-efab2dd48dd8 https://share.transistor.fm/s/384df7bb Richard and Pierce respond to the Times podcast about the scarcity of junior engineering jobs. They talk through the academic difference between Computer Science vs. Engineering, AI as a new engineering primitive, talent arbitrage through intern programs, and more.

https://www.nytimes.com/2025/09/29/podcasts/the-daily/big-tech-told-kids-to-code-the-jobs-didnt-follow.html

]]> Richard and Pierce respond to the Times podcast about the scarcity of junior engineering jobs. They talk through the academic difference between Computer Science vs. Engineering, AI as a new engineering primitive, talent arbitrage through intern programs, and more.

https://www.nytimes.com/2025/09/29/podcasts/the-daily/big-tech-told-kids-to-code-the-jobs-didnt-follow.html

]]> Thu, 02 Oct 2025 06:59:47 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 3109 Richard and Pierce respond to the Times podcast about the scarcity of junior engineering jobs. They talk through the academic difference between Computer Science vs. Engineering, AI as a new engineering primitive, talent arbitrage through intern programs, and more.

https://www.nytimes.com/2025/09/29/podcasts/the-daily/big-tech-told-kids-to-code-the-jobs-didnt-follow.html

]]> No The power of ten million deadlifters 21 21 The power of ten million deadlifters full 24d225a3-88f7-46c6-bf3c-477d18af1917 https://share.transistor.fm/s/71c8f1c2 OpenAI & NVIDIA’s 10GW partnership, GDPVal as a new human curated benchmark dataset, Gemini Robotics-ER 1.5, and Apple's distillation of AlphaFold.

Additional reading:
https://nvidianews.nvidia.com/news/openai-and-nvidia-announce-strategic-partnership-to-deploy-10gw-of-nvidia-systems
https://openai.com/index/gdpval/
https://deepmind.google/discover/blog/gemini-robotics-15-brings-ai-agents-into-the-physical-world/
https://arxiv.org/pdf/2509.18480

]]> OpenAI & NVIDIA’s 10GW partnership, GDPVal as a new human curated benchmark dataset, Gemini Robotics-ER 1.5, and Apple's distillation of AlphaFold.

Additional reading:
https://nvidianews.nvidia.com/news/openai-and-nvidia-announce-strategic-partnership-to-deploy-10gw-of-nvidia-systems
https://openai.com/index/gdpval/
https://deepmind.google/discover/blog/gemini-robotics-15-brings-ai-agents-into-the-physical-world/
https://arxiv.org/pdf/2509.18480

]]> Tue, 30 Sep 2025 14:00:35 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 3987 OpenAI & NVIDIA’s 10GW partnership, GDPVal as a new human curated benchmark dataset, Gemini Robotics-ER 1.5, and Apple's distillation of AlphaFold.

Additional reading:
https://nvidianews.nvidia.com/news/openai-and-nvidia-announce-strategic-partnership-to-deploy-10gw-of-nvidia-systems
https://openai.com/index/gdpval/
https://deepmind.google/discover/blog/gemini-robotics-15-brings-ai-agents-into-the-physical-world/
https://arxiv.org/pdf/2509.18480

]]> No How countries are actually using AI 20 20 How countries are actually using AI full 8e98eaa2-4677-4990-8951-1677c75be95c https://share.transistor.fm/s/fc465bd7 Pierce and Richard recap Anthropic's Economic Index. Differences between country use of AI, autonomy versus augmentation, and the real business use cases that Anthropic is seeing so far.

Further reading:
https://www.anthropic.com/research/anthropic-economic-index-september-2025-report

]]> Pierce and Richard recap Anthropic's Economic Index. Differences between country use of AI, autonomy versus augmentation, and the real business use cases that Anthropic is seeing so far.

Further reading:
https://www.anthropic.com/research/anthropic-economic-index-september-2025-report

]]> Thu, 25 Sep 2025 01:00:00 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 3078 Pierce and Richard recap Anthropic's Economic Index. Differences between country use of AI, autonomy versus augmentation, and the real business use cases that Anthropic is seeing so far.

Further reading:
https://www.anthropic.com/research/anthropic-economic-index-september-2025-report

]]> No Your new raybans just got smarter 19 19 Your new raybans just got smarter full d32c7a62-c122-41f9-bea2-9262393d2a37 https://share.transistor.fm/s/e21c58bf The official Claude Code post-mortum, Deepseek R1 published in Nature, Meta unveils their smart glasses with built‑in display, the new apple pro, copyright law in the age of AI, and much more.

Further reading:
https://www.anthropic.com/engineering/a-postmortem-of-three-recent-issues
https://www.nature.com/articles/s41586-025-09422-z#MOESM1
https://www.theverge.com/tech/779566/meta-ray-ban-display-hands-on-smart-glasses-price-battery-specs
https://www.reuters.com/legal/government/anthropics-15-billion-copyright-settlement-faces-judges-scrutiny-2025-09-09/

]]> The official Claude Code post-mortum, Deepseek R1 published in Nature, Meta unveils their smart glasses with built‑in display, the new apple pro, copyright law in the age of AI, and much more.

Further reading:
https://www.anthropic.com/engineering/a-postmortem-of-three-recent-issues
https://www.nature.com/articles/s41586-025-09422-z#MOESM1
https://www.theverge.com/tech/779566/meta-ray-ban-display-hands-on-smart-glasses-price-battery-specs
https://www.reuters.com/legal/government/anthropics-15-billion-copyright-settlement-faces-judges-scrutiny-2025-09-09/

]]> Tue, 23 Sep 2025 15:53:40 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 3654 The official Claude Code post-mortum, Deepseek R1 published in Nature, Meta unveils their smart glasses with built‑in display, the new apple pro, copyright law in the age of AI, and much more.

Further reading:
https://www.anthropic.com/engineering/a-postmortem-of-three-recent-issues
https://www.nature.com/articles/s41586-025-09422-z#MOESM1
https://www.theverge.com/tech/779566/meta-ray-ban-display-hands-on-smart-glasses-price-battery-specs
https://www.reuters.com/legal/government/anthropics-15-billion-copyright-settlement-faces-judges-scrutiny-2025-09-09/

]]> No The claude code conspiracy 18 18 The claude code conspiracy full 6919c879-2d54-4240-bb6b-af4a14eb33ee https://share.transistor.fm/s/2a643487 The Anthropic economic index report, a bug in claude's inference pipeline, OpenAI releases a flavor of GPT-5 just for coding, Microsoft's new inhouse LLM, and what really happens when you turn temperature to 0

]]> The Anthropic economic index report, a bug in claude's inference pipeline, OpenAI releases a flavor of GPT-5 just for coding, Microsoft's new inhouse LLM, and what really happens when you turn temperature to 0

]]> Thu, 18 Sep 2025 17:52:37 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 3230 The Anthropic economic index report, a bug in claude's inference pipeline, OpenAI releases a flavor of GPT-5 just for coding, Microsoft's new inhouse LLM, and what really happens when you turn temperature to 0

]]> No Will McTighe on selling through social media 17 17 Will McTighe on selling through social media full 959a669f-a2d2-4941-bc32-e2f74dbc9590 https://share.transistor.fm/s/0070fbf7 Pierce, Richard, and Will join for the first in-person interview on Pretrained. For our video episode, check out:
https://youtu.be/CInTOIgz-pA

They cover:
- Will's history growing up in the UK
- Getting an MBA and deciding what company to start
- Trust building versus activating content
- Building a personal brand for engineers and researchers
- Entrepreneurship in Europe vs the US
- & Much more

]]> Pierce, Richard, and Will join for the first in-person interview on Pretrained. For our video episode, check out:
https://youtu.be/CInTOIgz-pA

They cover:
- Will's history growing up in the UK
- Getting an MBA and deciding what company to start
- Trust building versus activating content
- Building a personal brand for engineers and researchers
- Entrepreneurship in Europe vs the US
- & Much more

]]> Tue, 16 Sep 2025 01:00:00 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 4257 Pierce, Richard, and Will join for the first in-person interview on Pretrained. For our video episode, check out:
https://youtu.be/CInTOIgz-pA

They cover:
- Will's history growing up in the UK
- Getting an MBA and deciding what company to start
- Trust building versus activating content
- Building a personal brand for engineers and researchers
- Entrepreneurship in Europe vs the US
- & Much more

]]> No Eating some mooncake 16 16 Eating some mooncake full 8c685dfd-6e66-4be0-8544-38f624ddc663 https://share.transistor.fm/s/4a53b59b Kimi's serving architecture, mooncake to offload GPU memory to other chipsets, the ubiquity of vllm, and the growing standard LLM stack

]]> Kimi's serving architecture, mooncake to offload GPU memory to other chipsets, the ubiquity of vllm, and the growing standard LLM stack

]]> Thu, 11 Sep 2025 20:00:00 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 2035 Kimi's serving architecture, mooncake to offload GPU memory to other chipsets, the ubiquity of vllm, and the growing standard LLM stack

]]> No Training a 1 trillion parameter model 15 15 Training a 1 trillion parameter model full 4c7e00cb-8a54-44c5-bedf-74bc49b99faa https://share.transistor.fm/s/1ae38953 Kimi K2 and Moonshot AI's history, avoiding loss spikes during training, the muon optimizer, and data parallelism

]]> Kimi K2 and Moonshot AI's history, avoiding loss spikes during training, the muon optimizer, and data parallelism

]]> Wed, 03 Sep 2025 22:07:54 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 2588 Kimi K2 and Moonshot AI's history, avoiding loss spikes during training, the muon optimizer, and data parallelism

]]> No Nano banana is our favorite fruit 14 14 Nano banana is our favorite fruit full 266437cf-b86b-425b-9a63-ee7da0eef7eb https://share.transistor.fm/s/1b7bdf36 Gemini’s new image model, OpenAI is investing more in protein generation, Cohere’s SOTA generation model, and Anthropic working with DOE on nuclear security


Further reading:
https://blog.google/products/gemini/updated-image-editing-model/
https://openai.com/index/accelerating-life-sciences-research-with-retro-biosciences/
https://cohere.com/blog/command-a-translate
https://red.anthropic.com/2025/nuclear-safeguards/
https://www.picolm.io/demo-paper

]]> Gemini’s new image model, OpenAI is investing more in protein generation, Cohere’s SOTA generation model, and Anthropic working with DOE on nuclear security


Further reading:
https://blog.google/products/gemini/updated-image-editing-model/
https://openai.com/index/accelerating-life-sciences-research-with-retro-biosciences/
https://cohere.com/blog/command-a-translate
https://red.anthropic.com/2025/nuclear-safeguards/
https://www.picolm.io/demo-paper

]]> Tue, 02 Sep 2025 01:00:00 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 3038 Gemini’s new image model, OpenAI is investing more in protein generation, Cohere’s SOTA generation model, and Anthropic working with DOE on nuclear security


Further reading:
https://blog.google/products/gemini/updated-image-editing-model/
https://openai.com/index/accelerating-life-sciences-research-with-retro-biosciences/
https://cohere.com/blog/command-a-translate
https://red.anthropic.com/2025/nuclear-safeguards/
https://www.picolm.io/demo-paper

]]> No A breakdown of Genie 3's world model 13 13 A breakdown of Genie 3's world model full e3e7d25c-10bd-4fa0-946a-f4ce2e75d702 https://share.transistor.fm/s/ac2b0269 Rich and Pierce speculate wildly (well, kind of wildly) on the internal architecture of Genie 3. They go into the history of variational auto encoders and diffusion models, 3D modeling with AI as an alternative to video game designers, a recap of the official Genie 1 paper, and possible applications of world models to the real world.

Further reading:
https://deepmind.google/discover/blog/genie-3-a-new-frontier-for-world-models/
https://arxiv.org/pdf/2402.15391

]]> Rich and Pierce speculate wildly (well, kind of wildly) on the internal architecture of Genie 3. They go into the history of variational auto encoders and diffusion models, 3D modeling with AI as an alternative to video game designers, a recap of the official Genie 1 paper, and possible applications of world models to the real world.

Further reading:
https://deepmind.google/discover/blog/genie-3-a-new-frontier-for-world-models/
https://arxiv.org/pdf/2402.15391

]]> Thu, 28 Aug 2025 01:00:00 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 3430 Rich and Pierce speculate wildly (well, kind of wildly) on the internal architecture of Genie 3. They go into the history of variational auto encoders and diffusion models, 3D modeling with AI as an alternative to video game designers, a recap of the official Genie 1 paper, and possible applications of world models to the real world.

Further reading:
https://deepmind.google/discover/blog/genie-3-a-new-frontier-for-world-models/
https://arxiv.org/pdf/2402.15391

]]> No Using chat in history class 12 12 Using chat in history class full ee956ec9-2782-4381-8dd0-49f12aecd6f7 https://share.transistor.fm/s/d49d3e24 Richard visits his high school, LLMs in education, is AI a calculator or an oracle, and more

]]> Richard visits his high school, LLMs in education, is AI a calculator or an oracle, and more

]]> Thu, 21 Aug 2025 01:00:00 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 2392 Richard visits his high school, LLMs in education, is AI a calculator or an oracle, and more

]]> No AI is your new favorite songwriter 11 11 AI is your new favorite songwriter full 0383614e-bebb-4096-8c4d-979aaa6cd763 https://share.transistor.fm/s/b2f52753 Elevenlabs launches a music generator, Claude Long-Term Memory, Reddit blocks the Internet Archive, NVIDIA’s Massively Multilingual Speech, and Self Questioning Language Models


Further reading:
https://techcrunch.com/2025/08/05/elevenlabs-launches-an-ai-music-generator-which-it-claims-is-cleared-for-commercial-use/
https://www.theverge.com/news/757538/reddit-internet-archive-wayback-machine-block-limit
https://www.theverge.com/news/757743/anthropic-claude-ai-search-past-chats
https://arxiv.org/html/2508.03682v3
https://arxiv.org/html/2508.05004v1
https://blogs.nvidia.com/blog/speech-ai-dataset-models/

]]> Elevenlabs launches a music generator, Claude Long-Term Memory, Reddit blocks the Internet Archive, NVIDIA’s Massively Multilingual Speech, and Self Questioning Language Models


Further reading:
https://techcrunch.com/2025/08/05/elevenlabs-launches-an-ai-music-generator-which-it-claims-is-cleared-for-commercial-use/
https://www.theverge.com/news/757538/reddit-internet-archive-wayback-machine-block-limit
https://www.theverge.com/news/757743/anthropic-claude-ai-search-past-chats
https://arxiv.org/html/2508.03682v3
https://arxiv.org/html/2508.05004v1
https://blogs.nvidia.com/blog/speech-ai-dataset-models/

]]> Tue, 19 Aug 2025 01:00:00 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 2991 Elevenlabs launches a music generator, Claude Long-Term Memory, Reddit blocks the Internet Archive, NVIDIA’s Massively Multilingual Speech, and Self Questioning Language Models


Further reading:
https://techcrunch.com/2025/08/05/elevenlabs-launches-an-ai-music-generator-which-it-claims-is-cleared-for-commercial-use/
https://www.theverge.com/news/757538/reddit-internet-archive-wayback-machine-block-limit
https://www.theverge.com/news/757743/anthropic-claude-ai-search-past-chats
https://arxiv.org/html/2508.03682v3
https://arxiv.org/html/2508.05004v1
https://blogs.nvidia.com/blog/speech-ai-dataset-models/

]]> No Tokenizers now and in the future 10 10 Tokenizers now and in the future full 094791c5-2324-4286-b5df-09139fd31e1a https://share.transistor.fm/s/e0ddb757 History of tokenization going back to 70s language modeling, modern tokenization approaches with BPE, and the future of token-free bitestreams

]]> History of tokenization going back to 70s language modeling, modern tokenization approaches with BPE, and the future of token-free bitestreams

]]> Thu, 14 Aug 2025 01:00:00 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 2863 History of tokenization going back to 70s language modeling, modern tokenization approaches with BPE, and the future of token-free bitestreams

]]> No All the GPTs we were promised 9 9 All the GPTs we were promised full c04ef5ce-8ef8-4a3c-bec0-d3dfca93a5fb https://share.transistor.fm/s/5d91e67c OpenAI releases their open weight GPT model, the release of GPT-5 and our growing attachment to different architectures, obsequiousness of language models, Genie 3 and open world models for reinforcement learning, and Pytorch as a standard in ML research

Further reading:
https://openai.com/index/introducing-gpt-oss/
https://openai.com/gpt-5/
https://deepmind.google/discover/blog/genie-3-a-new-frontier-for-world-models/
https://pytorch.org/blog/pytorch-2-8/

]]> OpenAI releases their open weight GPT model, the release of GPT-5 and our growing attachment to different architectures, obsequiousness of language models, Genie 3 and open world models for reinforcement learning, and Pytorch as a standard in ML research

Further reading:
https://openai.com/index/introducing-gpt-oss/
https://openai.com/gpt-5/
https://deepmind.google/discover/blog/genie-3-a-new-frontier-for-world-models/
https://pytorch.org/blog/pytorch-2-8/

]]> Tue, 12 Aug 2025 01:00:00 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 3518 OpenAI releases their open weight GPT model, the release of GPT-5 and our growing attachment to different architectures, obsequiousness of language models, Genie 3 and open world models for reinforcement learning, and Pytorch as a standard in ML research

Further reading:
https://openai.com/index/introducing-gpt-oss/
https://openai.com/gpt-5/
https://deepmind.google/discover/blog/genie-3-a-new-frontier-for-world-models/
https://pytorch.org/blog/pytorch-2-8/

]]> No Running out of good data 8 8 Running out of good data full e67d2db5-f3e7-4df3-bef9-1b0ae80714c7 https://share.transistor.fm/s/c660ad0a Exhausting the data on the Internet by 2026, the value or non-value of transcripts, training on other languages, and subsidizing additional dataset generation

]]> Exhausting the data on the Internet by 2026, the value or non-value of transcripts, training on other languages, and subsidizing additional dataset generation

]]> Thu, 07 Aug 2025 01:00:00 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 2788 Exhausting the data on the Internet by 2026, the value or non-value of transcripts, training on other languages, and subsidizing additional dataset generation

]]> No Models building other models 7 7 Models building other models full b6e662be-b747-4c71-9cf2-98fbe3bc02d1 https://share.transistor.fm/s/22a2adac ChatGPT's new study mode for students, Facebook's play to be the first with augmented reality LLMs, Cerebras is trying to compete with Claude Code, and are we really in an "alphago moment" for model architecture discovery?

Further reading:
https://openai.com/index/chatgpt-study-mode/
https://s21.q4cdn.com/399680738/files/doc_financials/2025/q2/META-Q2-2025-Earnings-Call-Transcript.pdf
https://www.cerebras.ai/blog/introducing-cerebras-code
https://arxiv.org/abs/2507.18074

]]> ChatGPT's new study mode for students, Facebook's play to be the first with augmented reality LLMs, Cerebras is trying to compete with Claude Code, and are we really in an "alphago moment" for model architecture discovery?

Further reading:
https://openai.com/index/chatgpt-study-mode/
https://s21.q4cdn.com/399680738/files/doc_financials/2025/q2/META-Q2-2025-Earnings-Call-Transcript.pdf
https://www.cerebras.ai/blog/introducing-cerebras-code
https://arxiv.org/abs/2507.18074

]]> Tue, 05 Aug 2025 01:00:00 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 3851 ChatGPT's new study mode for students, Facebook's play to be the first with augmented reality LLMs, Cerebras is trying to compete with Claude Code, and are we really in an "alphago moment" for model architecture discovery?

Further reading:
https://openai.com/index/chatgpt-study-mode/
https://s21.q4cdn.com/399680738/files/doc_financials/2025/q2/META-Q2-2025-Earnings-Call-Transcript.pdf
https://www.cerebras.ai/blog/introducing-cerebras-code
https://arxiv.org/abs/2507.18074

]]> No Evaluation metrics for reasoning models 6 6 Evaluation metrics for reasoning models full 268a2474-8779-4c26-ac7f-cd708ee1690b https://share.transistor.fm/s/3a00c6e0 Evaluating models on benchmarks, passing a model vibe check, formal reasoning to synthesize datasets, and what type of datasets researchers prefer

]]> Evaluating models on benchmarks, passing a model vibe check, formal reasoning to synthesize datasets, and what type of datasets researchers prefer

]]> Thu, 31 Jul 2025 01:00:00 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 1965 Evaluating models on benchmarks, passing a model vibe check, formal reasoning to synthesize datasets, and what type of datasets researchers prefer

]]> No The tea on vibe coding data leaks 5 5 The tea on vibe coding data leaks full 49ebf699-d444-4569-bdad-5641036695dd https://share.transistor.fm/s/9caf9d00 OpenAI introduces their new Agent with full computer use capabilities, Anthropic is now evaluating models with other models, what the data leak at Tea can and can't tell us about vibe coding security, and more reasoning time isn't always better

Further reading:
https://openai.com/index/introducing-chatgpt-agent/
https://alignment.anthropic.com/2025/automated-auditing/
https://www.nytimes.com/2025/07/26/us/tea-safety-dating-app-hack.html
https://safety-research.github.io/inverse-scaling-ttc/

]]> OpenAI introduces their new Agent with full computer use capabilities, Anthropic is now evaluating models with other models, what the data leak at Tea can and can't tell us about vibe coding security, and more reasoning time isn't always better

Further reading:
https://openai.com/index/introducing-chatgpt-agent/
https://alignment.anthropic.com/2025/automated-auditing/
https://www.nytimes.com/2025/07/26/us/tea-safety-dating-app-hack.html
https://safety-research.github.io/inverse-scaling-ttc/

]]> Tue, 29 Jul 2025 01:00:00 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 3457 OpenAI introduces their new Agent with full computer use capabilities, Anthropic is now evaluating models with other models, what the data leak at Tea can and can't tell us about vibe coding security, and more reasoning time isn't always better

Further reading:
https://openai.com/index/introducing-chatgpt-agent/
https://alignment.anthropic.com/2025/automated-auditing/
https://www.nytimes.com/2025/07/26/us/tea-safety-dating-app-hack.html
https://safety-research.github.io/inverse-scaling-ttc/

]]> No Choosing a college path in a post-AI world 4 4 Choosing a college path in a post-AI world full 1210a27d-fca6-4179-bbdb-91a1bd263dff https://share.transistor.fm/s/33cb6cac Allocating enough time to find your passions, Bell Labs vs Frontier Labs, project based classes, and getting involved in communities

]]> Allocating enough time to find your passions, Bell Labs vs Frontier Labs, project based classes, and getting involved in communities

]]> Thu, 17 Jul 2025 02:00:00 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 1613 Allocating enough time to find your passions, Bell Labs vs Frontier Labs, project based classes, and getting involved in communities

]]> No Surfing the agentic web 3 3 Surfing the agentic web full a232ad6b-46c4-45a6-adf4-29c4013ee4a4 https://share.transistor.fm/s/a010d57e Perplexity releases their AI browser Comet, Grok 4 and xAI's safety alignment strategy, the David vs Goliath of language models, and the Muon optimizer challenging Adam for its title

]]> Perplexity releases their AI browser Comet, Grok 4 and xAI's safety alignment strategy, the David vs Goliath of language models, and the Muon optimizer challenging Adam for its title

]]> Tue, 15 Jul 2025 02:00:00 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 3087 Perplexity releases their AI browser Comet, Grok 4 and xAI's safety alignment strategy, the David vs Goliath of language models, and the Muon optimizer challenging Adam for its title

]]> No The death of peer-reviewed research in AI 2 2 The death of peer-reviewed research in AI full 9199204d-7cd0-4c1d-a6bc-4556ac543e16 https://share.transistor.fm/s/c1c28eec Anthropic & OpenAI's new approach for announcing models, double vs single blind peer review, and distribution channels for researchers

]]> Anthropic & OpenAI's new approach for announcing models, double vs single blind peer review, and distribution channels for researchers

]]> Thu, 10 Jul 2025 02:00:00 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 2323 Anthropic & OpenAI's new approach for announcing models, double vs single blind peer review, and distribution channels for researchers

]]> No Party with the warehouse robots 1 1 Party with the warehouse robots full cbf3a27e-b7f5-4b25-a984-66b9e48d4de4 https://share.transistor.fm/s/3095c6b3 Inception Lab's new diffusion architecture Mercury, Google's emissions up 51% since 2019, DeepFleet warehouse scheduling robots, and Microsoft's delayed AI chip

]]> Inception Lab's new diffusion architecture Mercury, Google's emissions up 51% since 2019, DeepFleet warehouse scheduling robots, and Microsoft's delayed AI chip

]]> Wed, 09 Jul 2025 19:42:23 -0700 Pierce Freeman & Richard Diehl Martinez Pierce Freeman & Richard Diehl Martinez 3053 Inception Lab's new diffusion architecture Mercury, Google's emissions up 51% since 2019, DeepFleet warehouse scheduling robots, and Microsoft's delayed AI chip

]]> No