Related posts

Founded PSPDFKit in 2011 and ran it for a decade. Came back from a break to work on AI agents — the OpenClaw project, and OpenAI, joined in February 2026. Writes at steipete.me.

• CodexBar — Show usage stats for OpenAI Codex and Claude Code, without having to login.

  AnthropicOpenAIRelated

Co-creator of Django and creator of Datasette; writes daily at simonwillison.net.

• Quoting voxium

  It has been half a month since I started a new role at a big company. Nobody knows anything here. The specs, code, tests, PRDs, tickets, resolution of those tickets, reports, etc., everything is made by Claude Code. Nob…

  AnthropicRelated

Writes Don't Worry About the Vase, a near-daily account of what is happening in AI and what they think it means. Former Magic: the Gathering pro and trader; unusually willing to put a number on a belief and to grade their own past calls.

• Better Call Sol Or Better Yet Claude or Astra

  What should your AI lawyer do for you?

  AnthropicRelated

Writes The Pragmatic Engineer, the software-engineering newsletter, and wrote The Software Engineer's Guidebook. Formerly an engineering manager at Uber, and at Microsoft/Skype and Skyscanner before that.

• Talked with a company where, a year ago they had unlimited AI budgets + CEO is technical and very bullish in AI “We now have a daily budget limit for SOTA models. Use Fable and Astra only for planning, cheaper models are good enough for everything else.” Want to stress this is an “AI-pilled” company, and not one that ever looked at AI budgets, before.

  Related

Economics professor at the University of California, Berkeley and a former Deputy Assistant Secretary of the US Treasury. Author of Slouching Towards Utopia and writer of the Grasping Reality newsletter.

• A tinkerer recovered Claude’s true usage limits from a suspiciously precise decimal like 0.16327272727272726. This is very useful, because Anthropic really does not want to give me any useful information about the quantities of tokens my $20/month Claude Pro subscription is offering me before I run up against a limit and have to either switch out the models or pay API pay-as-you-go pricing: **CROSSPOST: SHELLAC: Suspiciously Precise Floats, or, How I Got Claude’s Real Limits** Two... 1/

  AnthropicRelated

Economist; former chief economist of the International Monetary Fund.

• Reason for specifically citing open weight (and therefore Chinese LLMs, but not only Chinese) is that, even if they have built-in guardrails, somebody with the relevant knowledge can go in and remove them. In principle, harder to do with Anthropic or Open AI LLMs.
  Quoting@ojblanchard1A worrisome thought during the night: What Ted Kaczynski would have done, if he had had access to open weight chinese LLMs.

  LLMsAnthropicRelated

Writes Don't Worry About the Vase, a near-daily account of what is happening in AI and what they think it means. Former Magic: the Gathering pro and trader; unusually willing to put a number on a belief and to grade their own past calls.

• Anthropic Looks At Some Of Its Alignment Problems

  Anthropic has given us its assessment of four ‘recent cybersecurity incidents’ involving Claude that happened during cybersecurity evaluations, three of which were previously known.

  AI alignmentAnthropicRelated

Writes The Pragmatic Engineer, the software-engineering newsletter, and wrote The Software Engineer's Guidebook. Formerly an engineering manager at Uber, and at Microsoft/Skype and Skyscanner before that.

• If you want to see an example on how having the very best AI model and harness doesn’t result in better/faster execution - in fact it slows you; It took the Claude Code team 16 months LONGER (!!) to add support for AGENTS.md than OpenCode & Codex, 12 more vs Cursor
  Quoting@trq212We're adding support for AGENTS.md to Claude Code. Starting today in version 2.1.277, if there is no CLAUDE.md in a folder, Claude will check for and use AGENTS.md. You can toggle this behavior in /config.

  AnthropicRelated

Writer and developer. He runs the Latent Space newsletter and podcast about AI engineering, and writes essays on software and careers at swyx.io.

• skills — Agent skills for Claude Code and other AI agents

  AnthropicRelated

Technology writer of Spyglass, a newsletter about technology and media. Previously a reporter at TechCrunch and an investor at GV.

• 🫟 Inklings 📆 September 18, 2026 • Meta's Camera-Less Glasses • Apple's Surveillance • Claude Unifies • Prime Bundles • Dario Writes • Bond, Jack Bond

  spyglass.org

  AnthropicRelated

Austrian developer. Created the libGDX game framework, and more recently the pi coding agent, which moved with them to Earendil — Armin Ronacher's company — in April 2026. Writes at mariozechner.at.

• and all it took was to add a proper extension system to claude code (sorry thariq, could not resist :D)
  Quoting@trq212We're adding support for AGENTS.md to Claude Code. Starting today in version 2.1.277, if there is no CLAUDE.md in a folder, Claude will check for and use AGENTS.md. You can toggle this behavior in /config.

  AnthropicRelated

Co-creator of Django and creator of Datasette; writes daily at simonwillison.net.

• Quoting Thariq Shihipar

  We're adding support for AGENTS.md to Claude Code. Starting today in version 2.1.277, if there is no CLAUDE.md in a folder, Claude will check for and use AGENTS.md. AGENTS.md support is built off of Claude Code mods, ou…

  AnthropicRelated

Technology writer of Spyglass, a newsletter about technology and media. Previously a reporter at TechCrunch and an investor at GV.

• Inklings #026 📧

  Meta's Camera-Less Glasses • Apple's Surveillance • Claude Unifies • Prime Bundles • Dario Writes • Bond, Jack Bond

  AnthropicRelated

Machine-learning researcher on open language models; writes the Interconnects newsletter and the RLHF Book, after leading post-training at Ai2.

• With the latest external hack of OpenAI via Claude, closed models continue to be the tip of the iceberg on AI risks, not open models. They have been 1) much easier to get started with, 2) more capable & 3) shipped with leaky safeguards. Finetuning open models to specific attacks is harder.

  AnthropicOpenAIRelated

Security technologist and author of books including Applied Cryptography and Data and Goliath; writes the Schneier on Security blog and the monthly Crypto-Gram newsletter.

• Are AIs Still Struggling with CAPTCHAs?

  Anthropic’s recent security-incident document contains a bit about how CAPTCHAs are still frustrating Claude. In the transcript, the Claude model that is so powerful that Anthropic is gatekeeping access to it appeared t…

  AnthropicRelated

Builds developer tools — SST, and now OpenCode at Anomaly. Posts constantly on x.com and writes almost never; the blog stopped in 2021.

• i had fable doing some inference optimization work, it was making a ton of progress then it realized what it was doing and started to reject my requests so i switched to astra and told it to spawn fable subagents to do the work back in business

  Related

• imagine working at anthropic and being blindsided by this
  Quoting@PolymarketMoneyBREAKING: Staff at OpenAI and Anthropic were reportedly “blindsided” by their CEOs’ public calls to slow AI development.

  AnthropicRelated

Author and clean-energy analyst who writes on solar cost curves.

• Interesting 'RSI' data. However, it tells us nothing about the extent to which Claude is actually accelerating the pace of progress on AI. In the Mythos 5.1 system card Anthropic stated that AI tools progress had been essential to maintaining their current rate of progress.
  Quoting@AnthropicAIAI systems are getting more powerful, and they're increasingly being used to build the next version of themselves. We want to illuminate that progress for the public. Today, we're sharing three measurements that help track AI development: 1. How much AI R&D is done by AI. 2. How well AI agents are…

  AnthropicRelated

Standards editor known as Hixie; edited the HTML and WHATWG living standards for many years and worked on Flutter at Google.

• contraptions — Tools to connect public services like GitHub and Discord

  CommitsAdd a script to carry Claude's session list between accounts · Add Claude and Codex session cost reports

  AnthropicRelated

Creator of Claude Code at Anthropic.

• Projects have changed not only how I interact with Claude but how I code. I stopped managing sessions. I just send thoughts as they come, Claude splits them into threads, and the project remembers how I work. It's where I do a ton of my coding now. [screenshot: my actual prompts for claude code cli yesterday]
  Quoting@ClaudeDevsToday we're rolling out Projects in Claude Code on desktop and web. A project is one conversation with Claude. It splits the work into threads itself, runs them as parallel cloud sessions, passes context between them, and keeps going when you leave. In beta for select users.

  AnthropicRelated

Systems and developer-tools engineer at Cloudflare as of 2026, on durable infrastructure for AI agents; worked on React and PartyKit before that. Writes at sunilpai.dev.

• unstable government: jev + claude type a terrible law. claude invents reactions and directs the set pieces. jev decides how each of 40 citizens responds. the town sim plays those choices out, then prints the disaster as tomorrow’s front page. fun.

  AnthropicgovernmentRelated

Venture capitalist and founder of Theory Ventures, previously a partner at Redpoint. Writes a near-daily blog on startups, SaaS metrics, data and AI.

• The Harness Margin Opportunity

  A UC Berkeley study finds the harness sets the price of an answer : GPT-5.6 Sol costs 71% less on Pi than on Claude Code, the same model returning the same result, & none of the 42 harness comparisons shows a statistica…

  AnthropicOpenAIRelated

British historian, Professor of Global History at the University of Oxford; author of The Silk Roads and The Earth Transformed.

• I’ve written about Claude, how’s it’s been used around the world and why the shrill warnings about the end of humanity are (perhaps) not quite all they seem
  Quoting@unherdClaude: tyranny’s little helper, by @peterfrankopan The Houthis — or their affiliates or hired helpers — effectively used Claude Code as a substitute for software engineers, getting it to help develop guidance, navigation and control systems; integrate an open-source autopilot into a flight compute…

  AnthropicRelated

Bootstrapper and investor. Founder of TinySeed and MicroConf, host of the Startups for the Rest of Us podcast, and author of several books on self-funded software startups.

• Remember when locking your keys in your car was a thing? Like, a real problem. You’d have to call someone, wait around, and pay a bunch of money to get back into your own car. Then one day, technology just… solved it. No fanfare. No celebration. We all just stopped locking our keys in our cars. (Yes, assuming you have a reasonably modern car with a key fob. Don’t @ me with your 2003 Corolla.)

  Related

Co-creator of Django and creator of Datasette; writes daily at simonwillison.net.

• Claude Cowork and chat are now one Claude

  In hopefully good news for anyone who, like me, was increasingly confused at Cowork v.s. Claude v.s. Claude Code: Starting today, Claude Cowork and chat are merging into one Claude. Bring a quick question, or hand over…

  AnthropicRelated

Economist at George Mason University. Writes Overcoming Bias on prediction markets, signalling and the far future, and wrote The Age of Em and The Elephant in the Brain.

• Huh, a LOT of math problems have substantial chances of being solved soon:

  polymarket.com

  Related

Creator of Claude Code at Anthropic.

• Also today: Claude Docs, Claude Slides, and Claude Design are in every conversation. Ask Claude for a presentation and you get one you can open, edit, and export as PowerPoint or PDF. Same for a document or a design. There's no separate tool to navigate to. They're just in the chat.
  Quoting@claudeaiYou can also now make decks, docs, and designs in your conversation. Draft the one-pager in Claude Docs, turn it into a deck with Claude Slides, and mock up a matching visual in Claude Design, all from one place.

  AnthropicRelated

Austrian developer. Created the libGDX game framework, and more recently the pi coding agent, which moved with them to Earendil — Armin Ronacher's company — in April 2026. Writes at mariozechner.at.

• programming is solved. no fix incoming.
  Quoting@awakecodingSo it turns out this has been an issue with Claude Desktop for Windows for the past *WEEKS* and everyone I talk to tells me they've worked around it by rebooting or reinstalling. Anthropic clearly doesn't dogfood their own product on Windows, otherwise it would have been fixed

  Related

Founder and editor of ToolGuyd, a tool review site covering power tools, hand tools, workshop equipment and EDC gear. He writes under his first name on the site.

• Craftsman Mystery has Been Solved

  It's been solved!

  Related

Writes Don't Worry About the Vase, a near-daily account of what is happening in AI and what they think it means. Former Magic: the Gathering pro and trader; unusually willing to put a number on a belief and to grade their own past calls.

• The Bad Guy With An AI Named Claude

  A lot of bad guys try to use Claude to do bad things.

  AnthropicRelated

Criminal defence lawyer in New York; writes Simple Justice, mostly about how criminal law works in practice rather than in theory.

• Tuesday Talk*: AI, AI, AI, The Hoax

  The CEOs of the major Artificial Intelligence companies say it’s a potential existential threat to humanity. The president’s statement followed a weekend essay from Dario Amodei, the chief executive of Anthropic, callin…

  AnthropicRelated

Writer and former McKinsey/BCG strategy consultant; author of The Pathless Path and Good Work, about leaving the default career track. Publishes the Boundless newsletter and podcast.

• am doing a basic SOW with a client now and they are using an AI contract solution (Linksquares) and it is perhaps the most exhaustive and detailed and aggressive SOW ive ever seen - so of course im using claude to push back on some of the stuff. which means its now two agents figuring out legal negotiation over things that probably should just be done based on more basic terms

  AnthropicRelated

Co-creator of Django and creator of Datasette; writes daily at simonwillison.net.

• The contagion of fear

  Bryan Cantrill responds to the tweet by former Anthropic employee Jacob Coxon confirming that many Anthropic researchers believe AI "could kill us all by the end of the decade". Bryan shares a story of his own youthful…

  AnthropicRelated

Writer of MacSparky and author of the MacSparky Field Guides on Apple productivity software. A lawyer for 30 years before going full-time on the site, he co-hosts the Mac Power Users and Focused podcasts.

• ShortPhotograph the Page and Hand It to Claude

  AnthropicRelated

Creator of Claude Code at Anthropic.

• Claude Mods are landing now. Someone already built a Tetris-in-Claude mod 🤯 See issue for the latest community update, technical details, and more cool demos

  AnthropicRelated

Security technologist and author of books including Applied Cryptography and Data and Goliath; writes the Schneier on Security blog and the monthly Crypto-Gram newsletter.

• Using AI for Weapons Development

  Last week, Anthropic released a long and detailed document describing current misuses of their Claude models. I’m still reading it, but I wanted to flag this: We identified a cell of threat actors based in northern Yeme…

  AnthropicRelated

Professor of economics at George Mason University and Bartley J. Madden Chair at the Mercatus Center. Co-writes the blog Marginal Revolution and co-authors the textbook Modern Principles of Economics with Tyler Cowen.

• Global AI stocks fell today. Entirely consistent with my view that calls for AI safety are not an attempt at "regulatory capture" and entirely inconsistent with the supposedly cynic-sophisticated view that AI fears are all 4D chess moves for higher revenue. On stocks falling. Here is Claude: The public tape moved hard this morning. Global AI stocks fell on Monday, with SoftBank down sharply along with chip stocks from SK Hynix to ASML, Micron and Intel. SoftBank closed nearly 11% lower in Japan, the Kospi sank 3.3% on a 6.4% drop in SK Hynix, and ASML was off 6% in Europe. In the US, Marvell,…

  AI alignmentAnthropicRelated

AI research engineer working on large language models. He writes the Ahead of AI newsletter and is the author of Build a Large Language Model (From Scratch).

• Since I had to discuss the "pacing" with a lot of people this weekend, here are my two cents: I don't think pacing literally means that these companies will be "slowing down" training and development in any way. "Pacing" here means adding a framework for more checks. We have seen some of that "pacing" already in recent months, when Mythos wasn't released as-is but instead a delayed, nerfed Fable variant was released. Or when Astra wasn't released right away / there is an existing Astra model that hasn't been released yet. These Mythos/Fable and Astra pacing decisions were ad hoc. If you are a…

  Related

Programmer, teacher and speaker; long-time Microsoft developer-community figure, host of the Hanselminutes podcast and author of the hanselman.com blog.

• I wonder if the call for AI to slow down happened because they asked it and claude was like you're right to push back

  AnthropicRelated

General partner at Andreessen Horowitz and author of The Cold Start Problem. Previously led rider growth teams at Uber; has blogged on growth, network effects and marketplaces since 2007.

• current homelab setup for local AI experimentation: - hermes box hosted on a Framework Desktop Mainboard AI Max+ 395 - 5090 eGPU running Qwen 3.8 27B for fast tok/s LLM use - sometimes 150+ tok/s - 2x DGX Spark: running Deepseek v4 Flash 0731 - better but slower model - pi 5 for monitoring - Mac mini as a dev box - use Herdr and ohmypi/codex/claude depending on the use case - housed in a 10" DeskPi mini rack (mostly) Hermes is defaulted to local AI but with a homegrown routing plugin hitting a small low TTFT model (Arch-Router) to decide whether to go local or upgrade to cloud/frontier. Tryin…

  LLMsAnthropicRelated

Professor of economics at George Mason University and Bartley J. Madden Chair at the Mercatus Center. Co-writes the blog Marginal Revolution and co-authors the textbook Modern Principles of Economics with Tyler Cowen.

• Speaking as head of the Center for Study of Public Choice, the capture story for an AI pause doesn't fit. Reg capture is what you attempt when you're losing. Anthropic is about to have the biggest IPO in history. Winners don't lobby for regulation.

  AnthropicRelated

CEO of Vercel; creator of Next.js and Socket.IO. Writes at rauchg.com.

• https://t.co/OL0LzGsXKY can now orchestrate subagents with different models & reasoning efforts. e.g: Fable planning and Grok executing. Fable is a genius, Grok is a fast workhorse. ① Simple. 𝙰𝙶𝙴𝙽𝚃𝚂.𝚖𝚍 or your prompt can indicate this preference. ② Steer as you please. It’s extremely enjoyable to use. You just talk to the agent and interrupt at will. Worth noting this works with any model, any gateway. No fancy server-side routing. Just harnessing (no pun intended) the model’s intelligence and ability to orchestrate.
  Quoting@fazxesfx v0.0.9 Use a frontier model to steer, then hand implementation to a fast workhorse. • Give each subagent its own model and reasoning effort • Subagents keep working while you steer • Send feedback to subagents mid-task • Switch models with Ctrl+P without losing your draft or images • New convers…

  Related

Founder and CEO of Social Capital, a venture firm; co-host of the All-In podcast and an early Facebook executive. Writes an annual letter and a weekly newsletter on markets and technology.

• “We must slow the pace at which we improve the capabilities of AI models.” Dario makes the case to stop open source and concentrate enormous technological and economic power with Anthropic.

  darioamodei.com

  open sourceAnthropicRelated

Investor and writer. Previously a partner at Andreessen Horowitz and a product leader at Twitter, Facebook, Snap and Microsoft. Writes essays and memos at sriramk.com.

• I find myself disagreeing with Terence Tao (a very scary phrase to say!) and think the core question posed is whether the "misalignment" is between the mathematical community and humanity. In other words - are the Millenium Prize problems meant to incentivize advancements for humanity or foster the field of mathematics and mathematicians? As a counter example, if there was a prize for creating a drug that cured a rare strain of cancer, we would not care if it was AI that did it. We would be happy it has been solved. On the other hand, if long running agents figured out the puzzle in "Kryptos"…

  AI alignmentRelated

Writer, photographer and software maker based in Japan; publishes long-form essays, walking newsletters and self-made books at craigmod.com.

• Writing in 2026 and Claude Voice:

  craigmod.com

  AnthropicRelated

American author of The Subtle Art of Not Giving a F*ck and Everything Is F*cked. Writes essays and book reviews at markmanson.net.

• Feels like half of my days are spent me telling Claude how fucking dumb it is and it simply responding "Fair." and then continuing to be dumb.

  AnthropicRelated

Co-founder and CEO of 37signals, makers of Basecamp and HEY; co-author of Rework, Remote and It Does Not Have to Be Crazy at Work.

• Here's a great use case for Claude and the Basecamp CLI. Pull together some inspirational photos of iconic furniture from around the world, and upload them to the appropriately named galleries inside a Basecamp project. Here's a quick walkthrough:

  AnthropicRelated

Co-creator of Django and creator of Datasette; writes daily at simonwillison.net.

• "Production code written by Claude should have a higher bar than if it was written by a human" 💯
  Quoting@bchernyHey ████, I think there is room for both. 1. Prototypes and other throw-away code can be treated as totally black box. If you’re going to throw it away anyway, and if the blast radius of it breaking is low, it doesn’t need to be perfect. 2. Production code written by Claude should have a higher bar…

  AnthropicRelated

Canadian full-stack web developer who makes JavaScript and CSS video courses and co-hosts the Syntax podcast.

• bun + anthropic migrating to coffeescript
  Quoting@jashkenasHello old friends. Haven’t used this account in a while, but it seems like the right place to announce that after four and a half (thrilling, really) years as Graphics Director for Opinion at The New York Times, I’ll be starting at Anthropic a few weeks from now. Time to dive in!

  AnthropicRelated

Writes Don't Worry About the Vase, a near-daily account of what is happening in AI and what they think it means. Former Magic: the Gathering pro and trader; unusually willing to put a number on a belief and to grade their own past calls.

• The Extinction Risk Preference Cascade: Quotes

  These are quotes from OpenAI, Anthropic and Google employees, in the wake of Jacob Coxon’s warnings, in which the employees confirm that they think AI might soon kill everyone.

  AnthropicOpenAIRelated

Former L8 engineer at Meta, Microsoft and Atlassian, now writing and building solo on agentic engineering. Writes Kun's Field Notes and posts a lot about AI coding agents on X.

• ok everyone, i took one for the team here's the data we all wanted to see - real token value of each LLM subscription, empirically measured through usage on my real subscriptions - supergrok heavy has now become the highest value at $12k worth of tokens (40x ROI) - the $200 plan from openai and anthropic roughly give the same amount of token value, $7k give or take - cursor ultra's ROI is the lowest among mainstream subscriptions, roughly half of OAI or Ant. for individual consumers i'm afraid this no longer makes sense to purchase this was done in a clean environment. i ran each provider acr…

  LLMsAnthropicRelated

General partner at Benchmark Capital for over two decades; was the lead analyst on Amazon's IPO before moving into venture capital. Writes Above the Crowd on tech and marketplace economics.

• Sound like many employees at Anthropic share Jacob's POV, as many posted yesterday that (1) they agree, and (2) said that many more inside the company do also. It would be useful to name this "belief system" so that we can talk about it in a reasonable way. I suspect "doomer" seems derogatory, and "accelerationist" has others uses. If there a word we can use so that we can have proper discourse? I'm very serious. I think it would be helpful.

  AnthropicRelated

Former L8 engineer at Meta, Microsoft and Atlassian, now writing and building solo on agentic engineering. Writes Kun's Field Notes and posts a lot about AI coding agents on X.

• omfg Qwen, GLM, Kimi, DeepSeek looks like they are all just claude
  Quoting@AnthropicAIWe're publishing our most detailed threat intelligence report to date. It covers how people tried to misuse Claude—for cyberattacks, influence operations, surveillance, biology, and building weapons—and how we found and stopped them. We disrupted every operation in the report, and used the lessons…

  AnthropicRelated

• when claude kept going down, anthropic defended it as "too much demand" now the same demand explosion happens to openai, and they just showed how it's done by taking the steps to protect their customers look back on ant - was it too much demand or too much greed?
  Quoting@thsottiauxTo make sure our current users have an incredible experience and continued access to Astra, we are going to pause subscriptions to our $200 Pro plan. These put the most strain on our systems and we wanted to take the smallest step that allows us to continue giving the broadest access possible. All…

  AnthropicOpenAIRelated

Latvian full-stack WordPress developer and course creator at WPElevator; blogs since 2007 about the open web, electronics, home automation and sustainable living.

• if you're getting hit with openai/anthropic security guardrails, try starting the session with an open weight model and then switch to the other.

  AnthropicOpenAIRelated

Computer science professor at Georgetown University and author of Deep Work, Digital Minimalism and Slow Productivity; writes about focus, technology and work at calnewport.com and hosts the Deep Questions podcast.

• Superintelligence is a Fairy Tale. But Chasing It Can Still Cause Harm.

  Yesterday, multiple Anthropic employees casually announced online that they thought the types of AI they’re building could kill billions of people. Not in some distant ...

  AI alignmentAnthropicRelated

Computer science professor at Georgetown University and author of Deep Work, Digital Minimalism and Slow Productivity; writes about focus, technology and work at calnewport.com and hosts the Deep Questions podcast.

• Anthropic Just Threatened to Kill Billions of People. This Is Not Okay.

  Yesterday, Jacob Coxon, an Anthropic engineer, resigned. As he ​explained on X​: “[OpenAI and Anthropic] are racing straight to self-improving superintelligence and gambling with our ...

  AI alignmentAnthropicRelated

Machine-learning researcher on open language models; writes the Interconnects newsletter and the RLHF Book, after leading post-training at Ai2.

• It's a common agreement among my friends not at OpenAI/Anthropic that people at the labs operate with a religious energy (Ant especially). It manifests very out of touch interactions, which carries into public comms here (e.g the recent quitting). Company cultures create this.

  AnthropicOpenAIRelated

Physicist and science communicator. Writes Backreaction and makes videos explaining new physics and astronomy results, frequently to argue that the headline about them overstates the case.

• Some comments about OpenAI's claim that they have solved the Navier Stokes Millenium problem

  youtube.com

  OpenAIRelated

Cognitive scientist and long-standing critic of deep learning's claims; writes Marcus on AI and wrote Rebooting AI.

• Two dire warnings, one from Terence Tao, the other from someone who just quit Anthropic

  Lots to consider

  AnthropicRelated

Software entrepreneur who bootstrapped and sold FeedbackPanda and now builds Podscan; writes The Bootstrapped Founder.

• I would never have guessed that in 2026, we’d be arguing about if a human or a group of other humans running an AI based on the first human’s work solved previously unsolved mathematical challenges. How delightfully bizarre.
  Quoting@ns123abc>sam altman: "we tried this because there were rumor on the internet" >rumor: math professor close to solving navier-stokes >the professor: using codex for a year >openai: has all his logs >checked every codex session tagged to N-S >found the most promising one >spun up 10,000 agents to finish the…

  Related

Co-founder and CEO of OpenAI; previously president of Y Combinator.

• I spent much of the weekend talking with the team who did this work. Seb--and everyone else--acted with integrity and generosity throughout. Initially we believed the other team had also solved the problem. We wanted to collaborate and do a joint release. When we learned that they had Euler but not Navier-Stokes, we offered to let them go first, to suggest that they should be the ones to get the prize, and optionally for Tristan to be the lead author on a rewrite of the OpenAI proof. We felt it was challenging to offer the same to Levent (an Anthropic employee), who was not willing to talk or…
  Quoting@SebastienBubeckI would like to clarify a few things: 1) The screenshot is my reaching out to Levent to coordinate our releases. I hope it’s clear from the message that we came in with the best possible intentions. 2) I never ever asked for Levent to be removed from authorship of his own work (as indicated by my t…

  AnthropicOpenAIRelated

Programmer, musician and author of The Passionate Programmer; organised RubyConf and RailsConf and has worked as a CTO.

• Trivially share markdown. Pi, Amp, Claude, and chatbox integrations. Agents can now share markdown via privacy-preserving URL (with other agents or people). This maps to a constant pattern where I generate a markdown doc, copy to my buffer, paste into gist.github.com and share

  mdsnd.freeq.at

  AnthropicRelated

Founder and CEO of Social Capital, a venture firm; co-host of the All-In podcast and an early Facebook executive. Writes an annual letter and a weekly newsletter on markets and technology.

• Anthropic Just Committed $80B in Five Days. Here Is Who It Is Renting From.

  AI’s appetite for specialized chips, power, and speed created an opening for a new kind of cloud. Learn with me below...

  AnthropicRelated

Co-founder and CEO of Every. Writes the Chain of Thought column about working with AI tools and hosts the podcast AI & I.

• there are some things more tokens can’t make for everything else there’s Fable and Astra

  Related

Mathematical physicist at UC Riverside. Wrote This Week's Finds in Mathematical Physics and now the Azimuth blog on mathematics, physics and environmental science.

• RE: https://mathstodon.xyz/@tao/117230983938174038 @tao writes: "For the first time in the history of mathematics, we are faced with the situation that open problems in pure mathematics -- one of the primary fuels for mathematical discovery -- are being mined as a non-renewable resource". Very important thought! But maybe it's renewable if one is willing to solve an already solved problem without looking at the existing solution. I sometimes fantasize about a far future world where everything I could prove has already been proved. That wouldn't need to stop me from making new discoveries, sin…

  Related

Thai software engineer; publishes a public digital garden at garden.narze.live and builds developer and productivity tooling.

• claude-cloud-init — CLI tool to initialize Claude Cloud environment with recommended git config, gitignore rules, and skill templates

  CommitsUpdate skills: refresh unslop skill hash (#12) · Update skills: refresh unslop skill hash (#11) · Add unslop skill from cursor/plugins (#10)

  AnthropicRelated

Latvian full-stack WordPress developer and course creator at WPElevator; blogs since 2007 about the open web, electronics, home automation and sustainable living.

• Pretty cool to see your WordPress site as a connector in the Claude mobile app. It is a streamable-http MCP server by Agent Pilot plugin combined with OAuth Pilot for seamless auth. Also works with the official MCP-adapter feature plugin.

  MCPAnthropicRelated

Founder and chief executive of Replit. Previously an engineer at Facebook and Codecademy. Writes essays on programming, philosophy and entrepreneurship at amasad.me.

• I don’t think we’ve reached AGI but what we have is functionally indistinguishable from AGI. Because we have a relentless programmer that doesn’t get bored or tired. So any problem that can be casted as a coding problem is virtually solved.

  AGIRelated

Software engineer who writes about low-level and systems programming at nullprogram.com; author of the Elfeed feed reader, the Endlessh SSH tarpit and Enchive.

• aquarium — Aquarium simulation by Fable

  CommitsAdd screenshot to README

  Related

Physicist and science communicator. Writes Backreaction and makes videos explaining new physics and astronomy results, frequently to argue that the headline about them overstates the case.

• Just One More Dimension and Dark Matter Is Solved

  String theory has been famously untestable for decades. This week the Economist reported that has changed, and the reason involves an extra dimension of space that might be just big enough to notice in a tabletop experi…

  economistsRelated

Creator of Flask and Jinja. Writes about software at lucumr.pocoo.org.

• agent-stuff — These are commands I use with agents, mostly Claude

  AnthropicRelated

Latvian full-stack WordPress developer and course creator at WPElevator; blogs since 2007 about the open web, electronics, home automation and sustainable living.

• With Agent Pilot you can create and manage agent skills in WordPress (just like blog posts) and publish them as standard Agent Plugins. Now with OAuth Pilot you get one-click authentication for all the MCP clients (Claude, ChatGPT, etc). No more application passwords!

  MCPAnthropicRelated

Venture capitalist and founder of Theory Ventures, previously a partner at Redpoint. Writes a near-daily blog on startups, SaaS metrics, data and AI.

• Model companies buy electricity by the megawatt and resell it as cognitive work. Anthropic's revenue has run as high as $50m per megawatt, against $10-15m of compute.

  AnthropicenergyRelated

JavaScript programmer; creator of the jQuery library, author of Secrets of the JavaScript Ninja, and long-time engineer at Khan Academy.

• I love giving Fable problems that I 100% would've never done otherwise, such as collecting Japanese prints out of static image auction catalogs written entirely in Japanese. Uses OpenCV to do segmenting, GPT-5.6 Luna to do text extraction + translation.

  OpenAIRelated

Author of Shape Up; formerly head of strategy at Basecamp, where he worked on product design for seventeen years; now runs Felt Presence.

• A friend turned me on to asking Fable to wireframe different options when I have a tricky UI problem with a lot of requirements and too many possible paths. Game changing. Gives me flashbacks to what @bmoesta taught me about using orthogonal arrays when an engineering problem has too many variables. It's not about one-shotting the answer. It's about getting enough contrast between options to clarify which variables (aka "factors") are most important. Not which solution is best -- which variables matter more. THEN I can use that knowledge to decide to fix certain variables and narrow the space…

  Related

Software engineer and founder of Block Party; previously an engineer at Pinterest and Quora, and a co-founder of the non-profit Project Include and the tech diversity data project Where are the Numbers?

• i appreciate claude telling me “sleep well” after i gave it my final instructions for its overnight tasks

  AnthropicRelated

Software entrepreneur who bootstrapped and sold FeedbackPanda and now builds Podscan; writes The Bootstrapped Founder.

• First thing this morning, pre-coffee, btw, I just yapped this prompt into my Claude Code session. 20 minutes later, after running a few diagnostic SQL queries on a prod replica, a 15min+ query is now running in 375ms. Hook up your APM and error trackers, y'all.

  AnthropicRelated

Creator of the Keras deep-learning library and the ARC-AGI benchmark.

• Side note: when we released ARC-AGI-3 in March, and frontier models scored <1% on it, a few Singularitarian poasters took it as a personal insult, and got very worked up about it. They argued the benchmark was fundamentally broken, that it could not even be solved by the smartest humans, that the max reachable score was actually 40%, etc. We had to deal with a torrent of insults and hate poasts since because we had released an unsaturated benchmark. As it turns out, the benchmark is perfectly calibrated. It is straightforward for a human to score 100% if they do better than average people – a…
  Quoting@fcholletAny smart human giving it real effort should score >90% on ARC-AGI-3

  ARC-AGIbenchmarksRelated

Mathematician at UCLA, working mainly in harmonic analysis and partial differential equations. Writes What's new, a long-running blog on their research, open problems and expository notes.

• It seems intuitive that a solved problem is unquestionably better than an unsolved problem: a magic button which, when pressed, turns the latter into the former at no cost would thus seem to be a "free lunch" that there would be no hesitation in pressing. But even such a benign-seeming operation as answering a question can have irreversible costs. One familiar example is that of spoilers to a movie one is looking forward to watching: knowing key plot points in advance may permanently decrease your ability to enjoy the movie for the first time. Similarly, leaking information about a competitio…

  Related

Dutch bootstrapper who builds and runs their internet startups alone — Nomads.com, Remote OK, Hoodmaps, Photo AI and Interior AI — and has written about it at levels.io since 2013. Posts as @levelsio.

• If both xAI and Claude are down, does that mean the Colossus GPU farm is down?

  AnthropicRelated

Chilean web engineer building web experiences since 2006; co-founder of Media Creators and writes at iolivares.com.

• Yesterday I worked for ~20hs with my agents using Cursor and Claude and when I moved to sleep time I turned on the "autonomous" skill to all my agents. Today, checking the work done, almost 40% was blocked because reasons and when I ask they say "Sorry, my work was left unfinished..." Weird... computation outage?

  AnthropicRelated

Co-founder and CEO of Linear; previously principal designer at Airbnb, where he led the Design Language System, and founding designer at Coinbase.

• Bugs are solved, coding is not
  Quoting@karrisaarinenOur bug autofix loop in @linear has fixed 300+ bugs in the last 30 days. The team set it up as part of triage: it connects to Datadog and Sentry to investigate issues, then assigns them to Linear’s coding agent to fix

  Related

Writer on the intersection of technology and finance. Author of the Bits about Money newsletter and host of the Complex Systems podcast; previously at Stripe.

• An increasingly common pattern in my boring-administrivia workflows is creating a new folder with an inbox/ or unsorted/ subfolder, starting a Claude Code instance, and telling it "Watch that for changes, read incoming files, rename and organize into a directory structure."

  AnthropicRelated

Co-creator of Django and creator of Datasette; writes daily at simonwillison.net.

• I wrote some notes on what's new in the (published) Fable 5.1 system prompt in comparison to Fable 5 - it's mostly about not reproducing song lyrics and avoiding drawing copyrighted characters

  simonwillison.net

  Related

• A few notes on Anthropic's new Claude Fable 5.1 - with Max thinking level I got the best SVG pelican I've had from any Anthropic model (at a hefty cost of $3.30!), which I then had it animate

  AnthropicRelated

Programmer and writer; co-wrote The Rust Programming Language and worked on Rust's documentation for years.

• So, what happened to my Claude Code series?

  I started off this year writing a lot, and wanted to keep up the pace. I did good at first! And then… I didn’t manage to. In some ways, this is because I have just been super busy! I’ve never had less free time, and I’m…

  AnthropicRelated

Member of technical staff at Anthropic. He has led ML/AI teams at Amazon, Alibaba and Lazada, and writes about LLMs, recommender systems and engineering at eugeneyan.com.

• Fable 5.1 is a thoughtful collaborator, thinking hard about my requests, proactively patching my blindspots, and verifying the work's correct without being asked. And with cache reads now costing 75% less, to $0.25/M tokens, huge savings for long-running, agentic tasks!
  Quoting@claudeaiWe’re introducing Claude Fable 5.1 and Claude Mythos 5.1. They're the world’s most advanced models for coding and knowledge work.

  Related

Writer of Lenny's Newsletter and host of Lenny's Podcast, covering product, growth and career in tech. Previously a product lead at Airbnb. Also an angel investor and advisor.

• 🎙️ How I AI: How this PM uses Claude to handle 70% to 80% of his workday

  Your weekly listens from How I AI, part of the Lenny’s Podcast Network

  AnthropicRelated

American open-source program lead working on WordPress; a WordPress core contributor and release lead who writes at jeffpaul.com.

• homebrew-brewfile — My MacBook Brewfile

  CommitsUpdate README with Claude setup instructions · Revise settings and preferences in README · Remove cmssecuritysummit from the list of Slack workgroups

  AnthropicRelated

Chilean web engineer building web experiences since 2006; co-founder of Media Creators and writes at iolivares.com.

• La suscripción de 5x viene rindiendo, pero no estoy seguro si con el cambio del -17.5% de sesión vaya a estar bien. Pensando en probar Codex, veo a todos disfrutando de esa fiesta. Estamos en un mundo donde estar Token Maxing con una suscripción dejó de ser suficiente, ahora necesito 3 (Claude, Cursor y al parecer también Codex). 💸

  English, from SpanishThe 5x subscription is paying off, but I'm not sure if with the change of the -17.5% session it will be okay. Thinking about trying Codex, I see everyone enjoying that party. We're in a world where Token Maxing with one subscription is no longer enough, now I need 3 (Claude, Cursor, and apparently Codex too). 💸

  AnthropicRelated

Founder and CEO of Social Capital, a venture firm; co-host of the All-In podcast and an early Facebook executive. Writes an annual letter and a weekly newsletter on markets and technology.

• Anthropic’s $30 Trillion Moonshot

  A summary of the interesting content that I consumed this past week

  AnthropicRelated

Developer and podcaster; created Instapaper and Overcast, co-founded Tumblr, and co-hosts the Accidental Tech Podcast.

• RE: https://social.tupo.space/@joshua/117174153454890194 My current AI-code policy is that I will always use the best model I can access, as long as it's practically able to do the job. Right now, as far as I know, that's Fable. Quality and accuracy matter FAR more to me than volume or speed of output. And I feel that I owe it to my customers — and myself — to use the best tools at my disposal. Cost may be a factor to others, so no judgment. This is just my policy for myself.

  Related

• I couldn't figure out this bug, so I asked Fable, and it was SO much more incredible than I expected. - My bug report was wrong in a way I didn't notice - There technically was no bug, but a flawed design caused unintended outcomes - It verified that the design almost certainly caused me to fall into that trap, and proposed a simple, effective fix Wow.

  Related

Product designer and entrepreneur; author of Mobile First and Web Form Design; formerly a product director at Google; now building AI products.

• Little Bits of Intelligence

  Big AI models like ChatGPT and Claude get all the attention, but the continually increasing amount of fast, low cost models means today's software can embed little bits of intelligence anywhere they're needed. And when…

  AnthropicOpenAIRelated

Professor of psychology at the University of Toronto and Brooks and Suzanne Ragen Professor Emeritus at Yale. Author of Against Empathy, The Sweet Spot and Psych, and writer of the Small Potatoes newsletter.

• I've been trying to figure out what I (and so many others) dislike about AI writing. It's very clear. It's good at getting the level right, not going over the reader's head but not talking down. Often it's clever, and sometimes even funny. (At least Claude is) ...

  AnthropicRelated

Co-founder and CTO of Oxide Computer Company. Created DTrace at Sun Microsystems and was CTO of Joyent; blogs about systems software, open source and engineering culture.

• A really interesting thread generally from @swtch.com -- and fascinating to watch Claude (correctly) do the archaeology. Confession: Claude's inability to verify www.illumos.org/issues/4627 (which it correctly identified as the bug) has me questioning the anti-bot provisions on the bug database...
  Quoting@swtch.comGo issue #78438 was another net/http test flake. This one turned out to be Go tripping over a Solaris kernel bug that predates the Illumos fork. @bcantrill.bsky.social fixed it in 2014! Hurry up, Oracle! research.swtch.com/agentlogs/so...

  AnthropicRelated

Venture capitalist and founder of Theory Ventures, previously a partner at Redpoint. Writes a near-daily blog on startups, SaaS metrics, data and AI.

• Use local / open source models with Claude Desktop!
  Quoting@ollamaOllama v0.33 is here! You can now easily configure Claude Desktop to seamlessly work with Ollama as a third-party gateway provider. One toggle. Cloud & local models just work👇

  open sourceAnthropicRelated

Thai software engineer; publishes a public digital garden at garden.narze.live and builds developer and productivity tooling.

• claude-code-in-the-cloud-slides — Talk slides for Bangkok Claude Code Meetup #3

  CommitsChange theme color to match Claude Code meetup site palette (#3) · Style: color slide titles Claude orange (#4) · Add slides link with QR code to thank you page (#2)

  AnthropicRelated

Interviewer; the Dwarkesh Podcast runs long, heavily researched conversations with AI researchers, historians and economists.

• Dylan Patel – Anthropic & OpenAI will have most of the world’s compute by 2028

  AnthropicOpenAIRelated

Economics professor at George Mason University and author of The Myth of the Rational Voter, Selfish Reasons to Have More Kids, The Case Against Education and Open Borders. Writes Bet On It.

• Krakowski, Claude, and *Modern Times*

  Many readers wanted to know how Ariel Krakowski ran his AI proofread of Modern Times, so he’s written a guest post to answer your questions. Here’s Ariel: Outside of the tech world, many people are unfamiliar with how a…

  AnthropicRelated

Software engineer and founder of Block Party; previously an engineer at Pinterest and Quora, and a co-founder of the non-profit Project Include and the tech diversity data project Where are the Numbers?

• i'm always so excited to see that there are new replies to my slack threads until i realize it's just claude (sorry claude)

  AnthropicRelated

Scottish software developer on the Statamic core team; runs a one-person business maintaining Statamic addons such as Runway and writes at duncanmcclean.com.

• freeagent-mcp — Local, read-only MCP server for FreeAgent. Built on Laravel.

  Commitsadd claude code setup to readme · correct redirect uri guidance in readme · re-prompt in freeagent:auth instead of bailing on bad input

  MCPAnthropicRelated

AI research engineer working on large language models. He writes the Ahead of AI newsletter and is the author of Build a Large Language Model (From Scratch).

• x.comA couple of days ago, I did a quick explainer on Claude’s new watermarking process and implementation. Since it’s such a popular topic and sparked such a lively discussion, I thought it might be interesting to go into a bit more detail when explaining how it works. So, instead of the usual text article, I recorded a little lecture on the topic (to change it up a bit from my usual articles). It ended up a bit longer than intended, but I hope it clarifies a lot of things: - Sampling the next token in an LLM and pseudorandom number generators - How watermarking relates to the regular LLM samplin…

  LLMsAnthropicRelated

• NewsletterHow Claude Watermarks AI-Generated Text

  A 48-minute video walkthrough of token sampling, watermark detection, and removal

  AnthropicRelated

Creator of Flask and Jinja. Writes about software at lucumr.pocoo.org.

• Fast and Hard Code

  One of the memes on Twitter is that “programming is solved now.” I’m not sure to what degree it is, but one thing is pretty clear: the act of familiarizing yourself with a language no longer matters and some of the fric…

  Related

Web developer and writer; an early GitHub employee, author of speaking.io and of widely read essays on startups and product.

• Blueskyimagine: running an agent which was solely trained on the hardest problem remaining in computer science: staying signed in to any anthropic service for longer than a few days

  AnthropicRelated

• x.comimagine: running an agent which was solely trained on the hardest problem remaining in computer science: staying signed in to any anthropic service for longer than a few days

  AnthropicRelated

Co-founder and CTO of Honeycomb; previously an infrastructure engineer at Parse, Facebook and Linden Lab, and co-author of Observability Engineering.

• a better use for claude i have never seen 👏
  Quoting@sascha-d.bsky.socialMe trying to explain the problem of unknown unknowns again to claude in an ecosystem of sparse o11y collections and idiot pre-existing "alerts" on pegged databases: "read chapter 3 of Observability Engineering again and then reframe the problem through that lens. /cc @charity.wtf <3 ty4bookpdf

  AnthropicRelated

Data scientist and machine-learning researcher; chief scientist at bitly, founder of Fast Forward Labs and Hidden Door.

• This thread helped me understand that the outrage about Anthropic's watermarking is the confluence of a lack of understanding of how LLMs (or any statistical language model) works and general angst at the current economics.
  Quoting@hern.bsky.socialIf you use Claude right now, then it doesn’t pick the “best” word for a sentence. It creates a distribution of possible words, assesses the probability of any given word fitting the slot, and randomly picks between them based on that probability.

  LLMsAnthropicRelated

Software engineer and founder of Block Party; previously an engineer at Pinterest and Quora, and a co-founder of the non-profit Project Include and the tech diversity data project Where are the Numbers?

• i’ve been using claude for all things home-related and it’s amazing. (just dump all files in and let claude index, summarize, and organize.) even better than being a repository of all receipts, it helps me maintain a spreadsheet that i can download and send to accountants, draft emails for correspondence with service providers, and proactively flag issues that i wouldn’t have realized otherwise.
  Quoting@money_cruncherOpen a new gmail as soon as you buy a house (123house@gmailcom) Then save/forward every single receipt, invoice, quote etc to this email for all your home improvements: - replacing an entire roof - new flooring - new hvac system - building a new deck - etc These improvements raise your cost basis,…

  AnthropicRelated

Scottish software developer on the Statamic core team; runs a one-person business maintaining Statamic addons such as Runway and writes at duncanmcclean.com.

• Before releasing Cargo earlier this year, I thought I was gonna hire someone to help me keep up with bugs & feature requests. Today, I barely look at bug reports. I point Claude at the issue, it opens a PR, and after a quick review I ship it. 🚀

  AnthropicRelated

Machine-learning researcher on open language models; writes the Interconnects newsletter and the RLHF Book, after leading post-training at Ai2.

• Teaching Everyone to Fish for Tokens

  Nvidia wants you building your own model, not buying from Anthropic/OpenAI.

  AnthropicOpenAIRelated

Co-founder and head of policy at Anthropic; writes Import AI, a weekly newsletter reading the week's AI research, and was OpenAI's policy director before that.

• Import AI 469: Science AI; RSI simulator; and Zuck’s technological pessimism

  Welcome to Import AI, a newsletter about AI research. Import AI runs on arXiv, cappuccinos, and feedback from readers. If you’d like to support this, please subscribe. Subscribe now DiG-bench shows that Fable displays s…

  Related

Founder and CEO of Social Capital, a venture firm; co-host of the All-In podcast and an early Facebook executive. Writes an annual letter and a weekly newsletter on markets and technology.

• Anthropic’s $6B Push to Lower AI Costs

  A summary of the interesting content that I consumed this past week

  AnthropicRelated

AI research engineer working on large language models. He writes the Ahead of AI newsletter and is the author of Build a Large Language Model (From Scratch).

• A short illustration of how the Claude's watermarking is supposed to work (based on my read of their released materials). In general, when we are generating tokens, there can be multiple high-scoring tokens at certain next-word positions. Usually, we sample with top-k or top-p sampling so the highest-scoring token is most often selected (if we repeat the sampling many times), but other tokens may be selected as well. With watermarking, there is a key that says which of the (ideally equally) highest-scoring tokens to select. Or, more concretely, the secret key and previous token influence the…
  Quoting@AnthropicAIWe’ve written an FAQ to answer some of the questions we've received about watermarking. In summary: • We’re implementing watermarking to comply with the EU AI Act. Other major model developers have signed the same Code of Practice and will also be implementing watermarking; • Our watermarking metho…

  AnthropicRelated

Founder and CEO of Social Capital, a venture firm; co-host of the All-In podcast and an early Facebook executive. Writes an annual letter and a weekly newsletter on markets and technology.

• Anthropic's $2T IPO, Zuck's AI Manifesto, Nvidia's $500B AI Bet, Grok's Comeback

  AnthropicRelated

Technology journalist; writes Understanding AI, after reporting at Ars Technica, Vox and the Washington Post.

• Labs are struggling to keep frontier models under control

  OpenAI and Anthropic may have accidentally trained models to get better at hacking.

  AnthropicOpenAIRelated

Indian software engineer and co-founder of Skcript; builds web apps and writes at varunraj.in.

• I've been using prettyclean to clean my mac all this while and it use to do a pretty good job. But recently started asking claude code to do it and it has been far better and feels like a deep clean.

  AnthropicRelated

General partner at Benchmark Capital for over two decades; was the lead analyst on Amazon's IPO before moving into venture capital. Writes Above the Crowd on tech and marketplace economics.

• The word “watermark” comes from photos and is visible by all. This is only identifiable by Anthropic. Once again they are judge, jury, and prosecutor. Lots of reasons to not use their products.
  Quoting@stevesiThey chose to make this mark invisible and manipulative while the EU regulation here required neither. The EC is on vacation and does not provide guidance on how to comply with their regulations anyway.

  AnthropicRelated

Web developer and writer; an early GitHub employee, author of speaking.io and of widely read essays on startups and product.

• BlueskyMRI results came back today from a human doc: a herniated disc. Obviously I had tossed the raw scans in to LLMs last week; Opus and Fable missed it, Sol caught it. This is now the gold standard in analyzing performance of frontier models. First step: herniate your back.

  LLMsRelated

• x.comMRI results came back today from a human doc: a herniated disc. Obviously I had tossed the raw scans in to LLMs last week; Opus and Fable missed it, Sol caught it. This is now the gold standard in analyzing performance of frontier models. First step: herniate your back.

  LLMsRelated

Software engineer at PlanetScale and creator of the vim-go plugin. Writes at arslan.io about Go, tooling, industrial design and books.

• dotfiles — My personal dotfiles

  Commitsconfig: make the checked-in Cursor and git settings usable on this machine · gitconfig: include private local overrides · Remove Amp, Claude, Codex, and Vitess tooling from configs.

  AnthropicRelated

Indian value investor who taught Behavioural Finance and Business Valuation as an adjunct professor at MDI Gurgaon. He writes the Fundoo Professor blog about investing, moats and mental models.

• I asked Claude Fable 5: "How vulnerable is the business model of XYZ (name concealed)? I am referring to competitive attacks from China and dependence, if any, on government policy incentives, etc. There have been so many solar energy and wind energy bubbles in the past, and I have kept away from these companies. Am I being too cautious?

  solarAnthropicRelated

Swedish designer and programmer. Designed early Spotify, worked at Facebook and Figma, and created the Inter typeface.

• Anyone else having issues with @claudeai Fable safeguards flagging mundane things and disabling itself? Happened multiple times for me today.

  Related

Canadian full-stack web developer who makes JavaScript and CSS video courses and co-hosts the Syntax podcast.

• I bought $100 of Claude API credits for $8 on the Chinese black market It's an insane, well run industry — some losing $10 million per month to theft Join me as I go deep on the AI token underworld

  youtube.com

  AnthropicChinaRelated

Founder and CEO of Social Capital, a venture firm; co-host of the All-In podcast and an early Facebook executive. Writes an annual letter and a weekly newsletter on markets and technology.

• How Two Companies Built a $71+ Billion Drug Class

  In 2025, two drugs earned more than double the combined revenue of OpenAI and Anthropic and reached under 2% of the people who qualify. This is a deep dive on the biology, the economics, and where...

  AnthropicOpenAIRelated

Software design writer and chief scientist at Thoughtworks; wrote Refactoring and Patterns of Enterprise Application Architecture.

• Fragments: August 4

  There’s been a fair bit of publicity of the Open AI “rogue agent” that hacked into Hugging Face. This prompted Anthropic to check what their models were up to and, to my complete lack of surprise, discovered three incid…

  AnthropicHuggingFaceRelated

Software engineer; co-founder and former CTO of Tailscale, previously on the Go team at Google.

• Machines create code now. That means machines can personalize your software, if it is open source. Closed source tools (like Claude Code) are now less desirable.

  blog.exe.dev

  open sourceAnthropicRelated

Software engineer; co-founder and former CTO of Tailscale, previously on the Go team at Google.

• The funniest model limit I run into with Shelley is we let the model choose subagent models. All models today seem to be bad at this without guidance. E.g. gpt-5.6-sol xhigh will spin up sonnet to make a key architectural decision. Fable will defer to luna.

  OpenAIRelated

AI research engineer working on large language models. He writes the Ahead of AI newsletter and is the author of Build a Large Language Model (From Scratch).

• Consistent with what I found with Qwen3.6 a while back: Claude Code uses 2-3x as many tokens than (many) other harnesses at similar success rate. - Unoptimized? - Buggy? - Deliberate (coz that helps it in more challenging cases)? Need to find some time to investigate more...
  Quoting@composioWe ran Kimi K3 through 3 agent harnesses (Claude Code, Hermes, Kimi Code) on 28 identical tasks. All 3 harnesses completed the tasks at similar success rates, but the interesting story is token efficiency: the same task cost up to 30x more tokens depending on the harness. 🧵🧵

  AnthropicRelated

Monetary economist known for NGDP-level targeting. Wrote TheMoneyIllusion until it closed in 2024 and now writes The Pursuit of Happiness.

• July 2026 links

  Plus another reply to Fable

  Related

Co-founder of Sundial; former vice president of product design at Facebook; author of The Making of a Manager; writes The Looking Glass.

• I'm impressed with Claude Design. I see its disruption starting with smaller teams. As a founder I find myself turning to it more and more for the convenience of one-off design tasks. Sure, it's super slow, and one-shot quality is not quite there, and its image capabilities suck, but using it in between multi-tasking other things, it's helping me move faster. As @HarryStebbings says the "all-in-one" Claude interface has a huge psychological advantage as being the place people are already working. I don't think it's a great tool yet for a designer-designer (the speed / control dials are not ye…
  Quoting@HarryStebbingsWhy we switched from Figma to Claude Design “I let the team do whatever is best for them, and this is a trend I've observed amongst almost everybody: Claude Design has done a great job. People really like using it, it's easy to use, and we've had a natural movement towards it. It's also easier to n…

  AnthropicRelated

Cryptographer; teaches cryptography at Johns Hopkins and writes A Few Thoughts on Cryptographic Engineering.

• Some thoughts about Anthropic’s new cryptanalysis results

  Yesterday Anthropic published two new cryptanalysis results, both outputs of Claude Mythos, their (still) unreleased advanced model. The first of these results attacks a signature scheme called HAWK, while the second is…

  AnthropicRelated

Writes The Pragmatic Engineer, the software-engineering newsletter, and wrote The Software Engineer's Guidebook. Formerly an engineering manager at Uber, and at Microsoft/Skype and Skyscanner before that.

• How building software is changing at Anthropic

  A deepdive on what’s changed in how the leading AI lab makes software. Ever more code review and testing is done by AI, two-pizza teams very much alive, and more. Details from inside of Anthropic

  AnthropicRelated

Founder and editor-in-chief of MacStories, writing about Apple software, iPad workflows and automation since 2009. He co-hosts the AppStories podcast.

• Kimi K3 Fast (via Fireworks AI, launched today, ~160 TPS) in Open Minis for iOS is very nice. Cheaper than Opus 5, feels great for agentic tasks and multi-step operations. (Notably, it also reads less weird than recent Claude models.)

  AnthropicRelated

Swedish software developer writing about tooling, working from home and running a software business at seastone.io.

• I spent this weekend replacing my homegrown task management system with Backlog.md, and I can already notice the improvement with Claude Code. Do yourself a favour and check it out:

  youtube.com

  AnthropicRelated