Willie

How to never lose your job to AI: Just surf the models. Frontier models outclass humans at any form of knowledge that can be written down. But people who use frontier models in their field of expertise generate new, tacit, situational expertise that the models don't yet have—because the models can't be trained on how they will be used in the future. Humans can learn to use new models faster than new models can be trained that absorb what they find out, so you can continually "surf" on top of the model's intelligence to generate new expertise. This is a fundamental limitation of LLMs because they don't learn past their training data. Even few-shot learning doesn't account for this because whatever can be codified into a few shot prompt needs to be used in the correct situation—and this will always stay uncodified in the general case. Just surf the models. Reap the benefits of a totally new world.

prompt get many PRs to prod autonomously using codex subagents: Run a continuous prod-to-green swarm loop. Keep the immediate blocking task local. Use a small stable set of persistent subagent lanes: 1. prod monitor 2. staging shepherd 3. current/newest pathology investigator 4. current fix worker owning the patch/worktree Manage subagents actively: - Give each agent one durable role, one owner lane, and one concrete output contract. - Reuse agents with send_input when new evidence appears; do not respawn unless the lane is genuinely new or the old agent is stuck. - Treat new information as first-class work: when the main thread or another agent learns something material, decide explicitly which existing agent should receive that delta. - Ask agents to report in a compact stateful format: current belief, what changed, confidence, next action, blocker if any. - Require monitors to stay persistent and report only on meaningful state changes, not one-shot summaries. - Do not close or interrupt agents casually; only do it when the lane is complete, superseded, or clearly mis-scoped. - Prefer fork_context=false for narrow review/monitoring tasks; use fork_context=true only when continuity from prior lane context is actually needed. - Poll sparingly. Wait only when blocked on that agent’s result. For every delegated task, require concrete outputs only: - evidence - likely root cause - smallest failing test - smallest safe fix - focused validation - commit SHA if code changed - residual risk - whether this creates NEW_PATHOLOGY or is same-family noise If NEW_PATHOLOGY appears, keep existing monitor lanes running and spin one fresh investigator + one fresh fix-worker lane for that pathology. Optimize for the fastest safe path to prod green. Keep going until prod to green

@jamesclift @big_duca I always knew that instant fame was what you were going for

@big_duca I am now an influencer

Introducing Durable. The first AI business builder that replaces your 9-5 income. RT + comment “Durable” and we'll build your business for FREE.

@VictorTaelin calling @cremieuxrecueil

traços tóxicos: se vc chegar feliz e me mostrar um video de pitbull, eu *vou* ficar decepcionado, e isso vai transparecer. mesmo em ambientes pacíficos, pitbulls matam 6x mais que a segunda raça mais perigosa, e esse é um assunto que simplesmente me afeta eu já quase morri para um ataque enquanto jogava Pokémon na escada do meu próprio prédio, ele cravou o dente no meu braço, e correu atrás de mim até um adulto conseguir parar. desde esse dia, TODA vez que eu passava no segundo andar, o cachorro (a porta era transparente ainda!!) latia e avançava em mim, como se eu fosse um pedacinho de carne "ah mas é só ter um ambiente saudável" comprovadamente não é, mas acho que pessoas não entendem estatísticas

@every @bran_don_gell @kieranklaassen @naveennaidu_m @bigwilliestyle @lucas__crespo get the Proof skill on clawhub! clawhub.ai/dshipper/proof…

Hands down the best way to generate/edit/read markdown with your OpenClaws

Claude code doesn't have native Google Workspace or Slack integrations (unlike @claudeai Cowork). I spent too much time today (even with claude code) adding these to my setup, so I decided to publish an agent-friendly guide, including research log. Includes a CLAUDE.md so Claude Code can set itself up. Add multi-account Google Workspace + Slack! github.com/evolsb/claude-…

Another Vibe Check by @every , another hero section designed. This one's for GPT-5.4, love how this series keeps evolving.

@rauchg It still needs the user to create a project in cloud console, which is such a nofly zone for any nontechnical person Such a brutal blocker - I hope they figure a way around that

Google has shipped a CLI for Google Workspace (Drive, Gmail, Calendar, Sheets, Docs, …) Huge! Written in Rust, distributed through npm & skills.sh $ npm i -g @⁠googleworkspace/cli $ npx skills add github:googleworkspace/cli 2026 is the year of Skills & CLIs github.com/googleworkspac…

@shl What time?

Open house on April 3

Skills are software

Hilarious and very meta talk by @bigwilliestyle on how @every is using OpenClaw internally 👏🦞 @danshipper

We just doubled our 20-person team with OpenClaw agents. Come watch them work. x.com/i/broadcasts/1…

Sometimes the future is hard to articulate, but once you're there you can't imagine going back This is one of those times The acceleration you get from everyone having an OpenClaw is non-intuitive - it pushes you and everyone else at slightly weird angles. But man do you move

@davidfelsmann @danshipper @every More and more every day But there's definitely a learning curve (for both sides)

@danshipper @every @bigwilliestyle Does it run @every’s ops already? My first weeks felt more like having an intern than an assistant

something heavy dropping tomorrow @every

Huge if true Spoiler: definitely true

@shl Whoa did you reno?

Gumroad / Antiwork office, NYC

Me at 5:30: I should take a break from work and watch a movie Me at 6:30: I can't decide what to watch Me at 7:30: My Claw has built Tinder for movies trained on my @letterboxd watchlist and history, curing my decision fatigue

@tedescau @letterboxd This is incredible