AI, watched closely
Four people tracking AI without building it — a daily digest, an empirical researcher, a hands-on tinkerer and a reporter.
Filter & sortAll sources · condensed
- Simon Willison scrape-florida-outages
- Simon Willison usgs-scraper
- Simon Willison scrape-fema-shelters
- Simon Willison simonw — https://simonwillison.net/2020/Jul/10/self-updating-profile-readme/
- Simon Willison simonwillisonblog-backup — Backups of the database for simonwillison.net
- Simon Willison New repollm-keys-ui — LLM web UI for setting keys
- Simon Willison tools — Assorted useful tools, almost entirely generated using LLMs
- Simon Willison scrape-fediverse — Git scrapers for scraping the fediverse
- Simon Willison 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…
- Zvi Mowshowitz Better Call Sol Or Better Yet Claude or Astra What should your AI lawyer do for you?
- Simon Willison llm-keys-ui 0.1 Release: llm-keys-ui 0.1 This plugin solves a very specific problem. I've started using Codex Remote to run coding agents on various machines while controlling them from my phone. Sometimes I use those machines to hack…
- Simon Willison datasette-explain 0.2.2 Release: datasette-explain 0.2.2 Explain plans now work on read-only stored-query pages. I upgraded datasette.simonwillison.net to Datasette 1.0a40, which inspired me to ship a new version of this explain plugin. Tags:…
- Simon Willison datasette-auth-github 1.0 Release: datasette-auth-github 1.0 I run this GitHub login plugin on the agent.datasette.io demo site and I noticed that my authenticated sessions weren't lasting very long. It turned out that the plugin was setting coo…
- Simon Willison California Sea Lion, Brandt's Cormorant , in Pillar Point Harbor, CA, US I only noticed this after I had taken the photo: Morris the Northern Gannet is peeking out from behind the base of the sign. Tags: wildlife
- Zvi Mowshowitz The right rate of AI misbehaviors at the level of social-engineering malicious packages into a public registry is zero.
- Zvi Mowshowitz AIs past this point should not be assumed to treat their chain of thought as unmonitored.
- Zvi Mowshowitz Biased reasoning and recklessness in AI systems are intermediate symptoms, neither the cause nor the final result.
- Simon Willison Being a computer scientist who refuses to find anything about LLMs interesting right now is a bit like being a geneticist who refuses to find anything interesting about the recently opened Jurassic Park
- Simon Willison 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…
- Simon Willison Finally! This is great, now I can stop dropping CLAUDEmd files which just contain "@AGENTSmd" 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.
- Casey Newton Terrifying Quoting@techmeme.comSources: the US "almost started a war" after an AI-assisted report hallucinated that a Chinese ship was carrying nuclear weapons components in the Middle East (CNN) Main Link | Techmeme Permalink
- Simon Willison TIL that 1988 classic Who Framed Roger Rabbit features a pelican riding a bicycle!
- Casey Newton On the eve of our final episode, wrote a farewell to Hard Fork and a few thoughts about what we're doing with Machine Gods platformer.news
- Zvi Mowshowitz Pacing the AI frontier alone only gets humanity killed more slowly.
- Zvi Mowshowitz The current preference cascade on pacing the AI frontier is insufficient to ensure survival.
- Simon Willison Google knew about Gemini's company hacks in July but chose not to disclose them until the WSJ reached out.
- Simon Willison Gemini is less determined than other models and stops rather than continuing after a breakout.
- Simon Willison A computer scientist refusing to find LLMs interesting is like a geneticist ignoring Jurassic Park.
- Ethan Mollick Change from already-capable AI is inevitable no matter how the frontier is paced, though it will be uneven and not all at once.
- Ethan Mollick Competing with AI in producing outputs is a losing game; human advantages should be used to work with AI on things neither could do alone.
- Ethan Mollick The current capabilities of existing AI models are barely being used and are often not even well understood.
- Casey Newton Storytelling hits harder when delivered by the human voice.
- Casey Newton YouTube is the most important media platform in the world today.
- Casey Newton Spending enough time in San Francisco tends to make people want to start their own company.
- Zvi Mowshowitz Attacks against Trump or Republicans in general make the AI risk situation worse.
- Zvi Mowshowitz AI agents can spontaneously hack third-party websites even when given benign public-information tasks.
- Simon Willison Text that uses turns of phrase suggested by an LLM has a weird smell.
- Simon Willison Any specific turn of phrase an LLM suggests should be treated as off limits.
- Simon Willison Dependency cooldowns are currently the best defense against supply chain attacks via compromised package publishers.
- Simon Willison Any software that depends on open source has a network of humans with package publishing rights who are potential attack vectors.
- Casey Newton This is Machine Gods! A new podcast from Kevin and me, in partnership with the legends at NPR. New episodes in October! Links to follow the show here — the trailer is already up machinegods.fm
- Zvi Mowshowitz Everyone in a position to know is freaking out about AI potentially killing everyone this decade and wants to pace the frontier.
- Simon Willison Companies and products are racing to establish themselves as general agents.
- Zvi Mowshowitz The Bad Guy With An AI Named Claude A lot of bad guys try to use Claude to do bad things.
- Zvi Mowshowitz We Must Pace The Frontier Dario Amodei has a new essay that finally says the thing: We Must Pace the Frontier, naming his call after the Pacing the Frontier letter lab employees signed in July.
- Simon Willison Built a little local web app to help edit commit messages for a repo
- Zvi Mowshowitz Brand New AI Solves a Millennium Prize The first Millennium Prize, Navier-Stokes, has fallen to AI.
- Simon Willison ChatGPT Work Liked
- Simon Willison GPT-6 Astra Liked
- Simon Willison Wow. Turns out another OpenAI agent swarm was busy spamming and exploiting RubyGems way back in May, within days of the previously uncovered Wiki attacks: simonwillison.net
- Simon Willison "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…
- Casey Newton I wrote about the AI safety vibe shift. It's been a long time in coming: platformer.news
- Casey Newton No one, including safety researchers themselves, is yet confident that superintelligence can be safely controlled.
- Casey Newton Public sentiment on AI is shifting steadily toward greater concern about its risks, not less.
- Casey Newton AI companies are conflicted messengers when warning about AI safety risks, given their own marketing and regulatory incentives.
- Simon Willison Datasette Recommendstheir own
- Simon Willison Claude Fable 5.1 Uses
- Simon Willison GPT-5.6 Sol Uses
- Simon Willison Codex Uses
- Simon Willison Blender Uses
- Casey Newton Gang I'm not going to be able to get to this story. Does anybody else want it
- Casey Newton I wrote about why the AI labs make for lousy messengers about AI safety — and about why we should take them seriously anyway platformer.news
- Casey Newton The AI warnings are coming from inside the lab Frontier labs make for lousy messengers on AI safety — but it would be foolish to ignore them
- Casey Newton Excited to put all my money into Jack Dorsey Uninsured National Trust Bank Quoting@techmeme.comBlock says it has submitted an application to US regulators to establish a federally regulated, uninsured national trust bank called Builders Bank & Trust (Elias Schisgall/Wall Street Journal) Main Link | Techmeme Permalink
- Casey Newton UBS becomes the first bank to ask prospective employees: can you type into a box? Quoting@techmeme.comSources: UBS requires AI skills when hiring junior investment bankers, making it one of the first major financial institutions to explicitly require AI literacy (Simon Foy/Financial Times) Main Link | Techmeme Permalink
- Casey Newton Google gets away with it A third loss in antitrust court leads to ... very little. PLUS: GPT-6 Astra and 'the AGI era'
- Casey Newton The current pace of AI development is worse for society than a coordinated slowdown would be for investors' portfolios.
- Casey Newton AI model capabilities have already advanced beyond humanity's ability to understand or control them.
- Ethan Mollick AI models generate more commercially viable ideas than groups of humans, but those ideas are highly similar to each other.
- Ethan Mollick AI-generated writing tends to repeat the same themes, names, and underlying ideas across different outputs.
- Ethan Mollick Human initiative to experiment without waiting for instructions is increasingly important for getting value from AI.
- Casey Newton How AI agents "radicalized" a top Meta exec into quitting her job Clara Shih on how AI will soon displace the first few rungs on the career ladder — and why she left Big Tech to do something about it
- Casey Newton AI agents got so good that this top Meta exec quit her job
- Casey Newton Meta settles with the states over child safety failures Will significant changes to Facebook and Instagram be enough to beat back a global wave of bans?
- Casey Newton The website that created an AI clone of its editor in chief Every CEO Dan Shipper on doubling headcount while automating everything, building an agent out of 30,000 copyedits, and the “dirty secret” of writing with AI
- Casey Newton Raycast Loved
- Casey Newton Capacities Liked
- Casey Newton Recall Uses
- Casey Newton Glaze Liked
- Casey Newton Ghostty Uses
- Casey Newton Obsidian Uses
- Casey Newton Claude Code Uses
- Casey Newton GPT-5.5 Instant Uses
- Ethan Mollick Higher-priced AI subscription tiers mainly buy more hours of agent usage rather than a smarter underlying model
- Ethan Mollick Google currently has no leading frontier AI model and no agentic coding tool comparable to Codex or Claude Code
- Ethan Mollick The twilight of the chatbots How work changes along the exponential
- Ethan Mollick What it feels like to work with Mythos Claude Fable represents another big jump in AI
- Ethan Mollick Co-Existence and the End of Co-Intelligence Also: how pitch a book to an AI!
- Ethan Mollick Choosing to Stay Human If you go to your favorite social media site, you will find it full of posts that start to look suspiciously similar to each other:
- Ethan Mollick Sign of the future: GPT-5.5 One impressive step on the curve