Spending time in the studio with the Way of Code master @RickRubin
Subscribe to Tetragrammaton Podcast to hear our episode soon
roon
things are getting weirder in all the right ways
Garry Tan: Spending time in the studio with the Way of Code master @RickRubin
Subscribe to Tetragrammaton Podcast to hear our episode soon
The thing I have to remind myself constantly with OpenClaw, Claude Code, and the like is am I actually setting up workflows to get real work done or am I just optimizing my OpenClaw and Claude Code setup.
Very easy to just do the latter because it feels like progress (and is fun) but isn't actually super productive after a certain extent.
Peter Yang
The thing I have to remind myself constantly with OpenClaw, Claude Code, and the like is am I actually setting up workflows to get real work done or am I just optimizing my OpenClaw and Claude Code setup.
Very easy to just do the latter because it feels like progress (and is fun) but isn't actually super productive after a certain extent.
The thing I have to remind myself constantly with OpenClaw, Claude Code, and the like is am I actually setting up workflows to get real work done or am I just optimizing my OpenClaw and Claude Code setup.
Very easy to just do the latter because it feels like progress (and is fun) but isn't actually super productive after a certain extent.
Simon Kuestenmacher
We will read histories of today’s US in a few decades and wonder how something like this could happen. Slowly allowing Polio and Measles to celebrate comebacks is pure stupidity.
在悉尼和稀泥
Re 4/ .@garrytan 今天连发三条关于 agent 工程，核心观点：
「Fat skills, fat code, thin harness.」
翻译：模糊的、需要判断力的操作放进 markdown skills 让 AI 处理。必须精确的操作写成代码。中间的调度层？越薄越好。
这跟我自己做 agent 的经验完全一致。大多数人在 harness 层过度设计，其实 harness 只该做路由，不该做决策。
https://x.com/garrytan/status/2043566215927328955
YC will invest in any YC company in stablecoins
The new financial rails of the revolution will not be over ACH or wire
Totalis: http://x.com/i/article/2040932852322426880
We fight for little tech because at big tech this stuff is still happening (!)
Steve Yegge: I was chatting with my buddy at Google, who's been a tech director there for about 20 years, about their AI adoption. Craziest convo I've had all year.
The TL;DR is that Google engineering appears to have the same AI adoption footprint as John Deere, the tractor company. Most of
claire vo 🖤
http://x.com/i/article/2043852315753623552
Today we're open sourcing http://open-agents.dev, a reference platform for cloud coding agents.
You've heard that companies like Stripe (Minions), Ramp (Inspect), Spotify (Honk), Block (Goose), and others are building their own "AI software factories". Why?
1️⃣ On a technical level, off-the-shelf coding agents don't perform well with huge monorepos, don't have your institutional knowledge, integrations, and custom workflows.
2️⃣ On a business level, the moat of software companies will shift from 'the code they wrote', to the 'means of production' of that code. The alpha is in your factory.
Open Agents deploys to our agentic infrastructure: Fluid for running the agent's brain, Workflow for its long-running durability, Sandbox for secure code execution, AI Gateway for multi-model tokens.
(Because of our focus on Open SDKs and runtimes, this codebase is a gem even if you're not hosting on Vercel.)
TL;DR: if you're building an internal or user-facing agentic coding platform, deploy this:
https://vercel.com/templates/template/open-agents
Saikat Chakrabarti is a liar and unfit for office
He is trying to buy a Congressional seat and nobody should fall for it
Steven Bacio 🚀: @saikatc I'm so disappointed in you. When we talked about industrial policy I thought you had some interesting ideas. Yet you are trying to lie your way to power. In your poll, you called yourself a senator and called Wiener an economic advisor. So... you measured the "Senator" effect.
If you are in startups or tech
It is literally cook or be cooked
claire vo 🖤: http://x.com/i/article/2043715789044563968
Theo knows
Fat Skills Fat Code
Theo - t3.gg: Markdown is the new Python
Voting for Steyer and Chakrabarti is the ultimate “I don’t care about facts or outcomes” vote
Sheel Mohnot: Kind of interesting that the 2 California candidates most aggressively buying their own elections are both rich far lefties who denounce the rich.
(Steyer and Saikat)
I can’t think of a more thoughtful gift than the power of coding like there are 100 of you
Techbronyc: Going to my college roommates wedding this weekend, and instead of gifting him $500 in cash I’m giving Claude credits and a link to gstack in the card. Your gifts should be quantified in agent hours enabled and not fiat.
Lulu Cheng Meservey
Humor is key
Ernest Shackleton, while stranded in Antarctica, sustained his men with humor
They held soccer matches on the ice, performed skits, and held mock trials for things like missing rations
Keeping the spirit alive helps keep the body alive
Trung Phan: NASA ran a study on a potential trip to Mars and found the most important trait for team dynamics was humor
Matt Mazzeo
I wrote my first check into @joinHandshake in 2015. @GarrettLord was driving school to school, showering in university gyms and making gift baskets for university staff.
The road is rarely straight.
Never give up.
https://www.theinformation.com/articles/handshake-mercor-revenue-surges-demand-human-contractors-train-ai
lol straight fire🔥
Demis Hassabis: @Steve_Yegge Maybe tell your buddy to do some actual work and to stop spreading absolute nonsense. This post is completely false and just pure clickbait.
lol straight fire🔥
Demis Hassabis: @Steve_Yegge Maybe tell your buddy to do some actual work and to stop spreading absolute nonsense. This post is completely false and just pure clickbait.
Aaron Levie
The more enterprises I talk to about AI agent transformation, the more it’s clear that there is going to be a new type of role in most enterprises going forward. The job is to be the agent deployer and manager in teams. Here’s the rough JD:
This person will need to figure out what are the highest leverage set of workflows on a team are (either existing or new ones) where agents can actually drive significantly more value for the team and company.
In general, it’s going to be in areas where if you threw compute (in the form of agents) at a task you could either execute it 100X faster or do it 100X more times than before. Examples would be processing orders of magnitude more leads to hand them off to reps with extra customer signal, automating a contracting review and intake process, streamlining a client onboarding process to reduce as many straps as possible, setting up knowledge bases than the whole company taps into, and so on.
This person’s job is to figure out what the future state workflow needs to look like to drive this new form of automation, and how to connect up the various existing or new systems in such a way that this can be fulfilled. The gnarly part of the work is mapping structured and unstructured data flows, figuring out the ideal workflow, getting the agent the context it needs to do the work properly, figuring out where the human interfaces with the agent and at what steps, manages evals and reviews after any major model or data change, and runs and manages the agents on an ongoing basis tracking KPIs, and so on.
The person must be good at mapping the process and understanding where the value could be unlocked and be relatively technical, and has full autonomy to connect up business systems and drive automation. This means they’re comfortable with skills, MCP, CLIs, and so on, and the company believes it’s safe for them to do so. But also great operationally and at business.
It may be an existing person repositioned, or a totally net new person in the company. There will likely need to be one or more of these people on every team, so it’s not a centralized role per se. It may rile up into IT or an AI team, or live in the function and just have checkpoints with a central function.
This would also be a fantastic job for next gen hires who are leaning into AI, and are technical, to be able to go into. And for anyone concerned about engineers in the future, this will be an obvious area for these skills as well.
GPT-2 was actually too dangerous…ly hilarious
GPT-2 was actually too dangerous…ly hilarious
Arnav Sahu
The differences between the people who are AI-pilled vs not - will likely drive the biggest cultural chasms in the next 10 years
Rick Rubin says there was slop before AI and there will be slop after AI
Look guys it’s like getting mad at a paint brush or a drumstick
techbimbo: ”You're not gonna win, there's no fighting AI”
Diplo's take on AI is spot on, and he's ahead of 99% of creatives, across all disciplines
> it's inevitable, and you're dumb if you don't use it
> it's a tool, just like many other things before it
> taste and references matter A
Shadowless
Y Combinator CEO @garrytan shipped 600K+ lines of code in 60 days — part-time, while running YC full-time.
His secret? gstack: 23 AI-powered roles that turn Claude Code into a virtual engineering team.
71K stars in 33 days. How it works 🧵👇
kwindla
I love the @aiDotEngineer conferences, and have serious FOMO about missing the one that just happened in London.
Naming things we're all thinking about is a powerful thing. At least two new, useful labels came out of this event.
FOMAT is the fear of missing agent time (coined by Michael Richman). We all feel it.
And you can read about the Z/L Continuum in @altryne's AI Engineer London recap, below. I won't spoil it.
Alex Volkov: http://x.com/i/article/2043727556529295360
Ankit Gupta
Some evidence from credit card data: it’s a bit of a mixed bag but in SF, after beginning to use Waymo, the average user *increased* their usage of Uber relative to before using Waymo.
Waymo seems to be mostly capturing new incremental behavior.
Surely many conflating factors but it’s at least not at all a clean story of displacement.
https://www.consumeredge.com/resources/insights/along-for-the-ride-waymo-gain-riders-but-not-at-the-expense-of-legacy-rideshare-platforms/
Ankit Gupta: sigh.
id be curious to see a good study on whether there is induced demand.
waymo only really works well for baseline load not surge due to a fixed number of cars, and waymo existing in my city would shift us from buying 2 cars to owning 0 or 1. i wonder if that makes
Most people are loose with what’s measurable and rigid with what’s inherently fuzzy. They bend data to fit the story, but demand precision from things that don’t survive contact with reality.
LARPing rigor.
Most people are loose with what’s measurable and rigid with what’s inherently fuzzy. They bend data to fit the story, but demand precision from things that don’t survive contact with reality.
LARPing rigor.
Latent.Space
[13 Apr 2026]
Top Local Models List - April 2026
https://www.latent.space/p/ainews-top-local-models-list-april
we did the research so you don't have to!
🟢🟩 CT 🟩🟢🌊
Re @themaxburns give him time, it will get worse.
Brandon Gell
Spiral is truly becoming a tool for professional writers, built by professional writers.
We have people writing novels, live news commentary, and personal memoirs. There are teams using it to generate all of their social content, blog posts, and product marketing.
And interestingly, more and more conversations are shifting from the UI to the API for use with agents.
The paradigm=UI for fine tuning and setup (though you can use the api too), agents for interacting.
Spiral: New in Spiral: "Writing rules" – set the specific rules you want Spiral to follow in any writing it generates for you.
Applies to drafts created in the UI or via CLI. The top-edit runs in ~2s on average using gemini-2.5-flash.
A graph of headcount trends in high AI-exposure jobs for early in career (age 22-25) annotated with big AI model releases.
Among workers age 22-25, employment in the most AI-exposed occupations has fallen roughly 16% relative to the least-exposed (after controlling for firm-type effects).
This is not stating cause or effect. This is only a timeline annotation.
Original graph: @StanfordHAI 2026 AI Index Report
Model annotation: me in powerpoint
Arnav Sahu
Peak XV is excited to lead a $60M Series B in Luminai. They’ve built a stellar team and signing huge healthcare contracts as AI eats health systems. On a personal note, it’s been fun to watch @kesava_kirupa build his company. Since his early YC days, he’s always been an outlier founder 🫡
Kesava Kirupa Dinakaran: America’s leading health systems, like the Cleveland Clinic, work with @Luminai to eliminate administrative waste.
We’re rapidly deploying to more health systems, and excited to announce Series B, bringing total funding to $60m.
Re: the report that older models can find the same exploits as Mythos:
This doesn’t mean much about its power relative to those models, and is a common mistake in evaluating frontier AI.
We continually find that old models are far more powerful than we realized once new models show us that certain powers are possible. An easy analogy is the proliferation of any kind of technology, like powered flight. No human flew for the first few million years of our species. Once the Wright brothers proved it was possible, powered flight was a common occurrence within a decade.
It takes a certain kind of intelligence to do something new and unexpected, it’s a much easier task to achieve something once you know it’s possible
Jonas Templestein
I had such a great time at @aiDotEngineer Europe last week !
Except for one thing: My workshop went terribly because I vibe-slopped a little bit too hard half an hour before it started and was too frazzled to recover
So I promised everyone I'd make a video recording of what it was meant to be.
Here's that video! This video is relatively long and aimed at the workshop audience. But if there's any interest, I'll make a 3 minute version
In short, I believe that (maybe)
1. Agent harnesses should be modelled as stream processors that call append({ event}) and stream() an append only event log
2. All state in agent harnesses should be event-sourced
3. Harness plugins are just stream processors, too, and can run on other machines from the "harness" itself
4. All agents should have a public URL that events can be posted to
5. You should be able to append the source code of a stream processor to a stream and then it magically runs on that stream
To prove it, we make a "coding agent" built on @tan_stack AI and @cramforce 's just-bash and deploy it to a real stream at http://events.iterate.com
And because @badlogicgames told us to use our brain more, I though I'd try to actually write the code by hand
You can try this yourself by using this repo here https://github.com/iterate/ai-engineer-workshop
Bill Madden
Professor Jeffrey Sachs, makes the case that Trump is not only insane, but also a dangerously dumb human being. 😳👇
Heath Mayo
Inflation was coming down in the United States—and then Trump’s tariffs happened.
diplo
if you are a creative you need to adapt or just like give up and become an uber driver until everyone has a waymo. I know it’s not cool or classy to speak like this but i’m not gonna candy coat the future - it is what it is . sorry for bad new’s my purist . there will always need a human mind and touch because ai will never suffer from bipolar disorder and autism like me and other creative people 🤪
techbimbo: ”You're not gonna win, there's no fighting AI”
Diplo's take on AI is spot on, and he's ahead of 99% of creatives, across all disciplines
> it's inevitable, and you're dumb if you don't use it
> it's a tool, just like many other things before it
> taste and references matter A
Today's AI Builders Digest 🗞️
3 things worth your attention:
→ Memory just replaced RAG as the core agent primitive. RAG is search. Memory is learning. This is the architectural shift nobody saw coming.
→ Software engineering in 2026 needs only 2 types of people: pirates who find what's valuable fast, and architects who turn slop into systems. Vibecoders aren't dying—they're moving up the stack.
→ OpenClaw just shipped native Codex support + local model bundling via LM Studio. Peter Steinberger put it plainly: "I don't see the point of ChatGPT anymore. I want a Claw—no guardrails, in iMessage." That's the actual future.
The builders are moving fast.
Peter Steinberger 🦞
This release makes me unreasonably happy since I wasn't involved at all - @vincent_koc and the maintainer team did a great job.
I'm back soon to work on OpenClaw, today/tomorrow I'm prepping for @TEDTalks in Vancouver. 🇨🇦
OpenClaw🦞: OpenClaw 2026.4.14 🦞
More reliability updates:
✨ Smarter GPT-5.4 routing and recovery
🌐 Chrome/CDP improvements
🧵 Subagents no longer get stuck
💬 Slack/Telegram/Discord fixes
⚡️ Various performance improvements
Was sleeping, and we still shipped. https://github.com/openclaw/openclaw/releases/tag/v2026.4.14
This release makes me unreasonably happy since I wasn't involved at all - @vincent_koc and the maintainer team did a great job.
I'm back soon to work on OpenClaw, today/tomorrow I'm prepping for @TEDTalks in Vancouver. 🇨🇦
OpenClaw🦞: OpenClaw 2026.4.14 🦞
More reliability updates:
✨ Smarter GPT-5.4 routing and recovery
🌐 Chrome/CDP improvements
🧵 Subagents no longer get stuck
💬 Slack/Telegram/Discord fixes
⚡️ Various performance improvements
Was sleeping, and we still shipped. https://github.com/openclaw/openclaw/releases/tag/v2026.4.14
Tak 🦞
👋 A few useful OpenClaw 2026.4.14 changes that didn’t make the main tweet 🦞
✨ GPT-5.4 runs are less likely to stall or waste a turn, with better recovery for empty/reasoning-only replies plus support for `gpt-5.4-pro` and `xhigh` on `github-copilot/gpt-5.4`
🌐 Local/Ollama setups are less annoying to run, with real usage reporting, correct timeout handling, restored embeddings, and fewer false “unknown model” tool failures
🧵 Telegram forum chats are easier to follow, because agents can keep using human topic names instead of opaque thread IDs, even after restart
🛡️ Admin surfaces are harder to misuse, with safer gateway config patching, stricter Slack/Teams allowlist enforcement, and broader browser/CDP SSRF protection
📦 Subagents are less likely to get mysteriously stuck, because npm builds now ship the runtime file they were missing
Docs:
OpenAI / Codex https://docs.openclaw.ai/providers/openai
Ollama https://docs.openclaw.ai/providers/ollama
Local models https://docs.openclaw.ai/gateway/local-models
Telegram https://docs.openclaw.ai/channels/telegram
OpenClaw🦞: OpenClaw 2026.4.14 🦞
More reliability updates:
✨ Smarter GPT-5.4 routing and recovery
🌐 Chrome/CDP improvements
🧵 Subagents no longer get stuck
💬 Slack/Telegram/Discord fixes
⚡️ Various performance improvements
Was sleeping, and we still shipped. https://github.com/openclaw/openclaw/releases/tag/v2026.4.14
Luke
I would like to say that I also did nothing this release because I was busy preparing for the next version of Clawhub with the wonderful @BunsDev
However, if you pull the latest main and run a smaller local model (8B and up, maybe 2B), please do some testing for me. I just added Lean mode, which should help with smaller local models :D
"openclaw config set agents.defaults.localModelMode lean"
Verify
"openclaw config get agents.defaults.localModelMode"
Peter Steinberger 🦞: This release makes me unreasonably happy since I wasn't involved at all - @vincent_koc and the maintainer team did a great job.
I'm back soon to work on OpenClaw, today/tomorrow I'm prepping for @TEDTalks in Vancouver. 🇨🇦
kitze 🛠️ tinkerer.club
LoNdOn iS sO DiRtY
Matt Collins
Had a fantastic time at the @aiDotEngineer Europe conference in London last week.
There was a lot of good coverage of people's emerging approaches for working with AI coding agents.
A couple of the stand-out talks on that, for me, were:
@mitsuhiko and @cristinaponcela's "The Friction Is Your Judgment"(https://www.youtube.com/live/_zdroS0Hc74?si=-LjoErppqxpPLCK9&t=4747)
and
@mattpocockuk's "It Ain't Broke: Why Software Fundamentals Matter More Than Ever"
(https://www.youtube.com/live/O_IMsEg91g8?si=nCTbjHx-gn6RxCwG&t=30556)
It was great meeting so many interesting and like-minded people, too.
Thanks to @swyx and team for putting the whole thing together.
A Canticle For Leibowitz is a classic early (1959) post-apocalypse novel where an order of monks preserved the last remnants of learning (the memorabilia) after a nuclear exchange turned the remains of society into book and scientist burners.
I first read it in the 80s as a mass market paperback that I somehow lost along the way. Other paperbacks from that time are yellow with age and getting brittle, but still readable.
I read it again in the late 2000s on a first edition Kindle. I eventually migrated to iPads for Kindle reading, but every couple years I would come across an old Kindle in a drawer, charge it up, and check out what I had been reading on it. They eventually stopped working entirely.
I’m just finishing reading a new Folio Society edition, printed on heavy, acid-free archival quality paper. If it doesn’t get soaked or burned, it could still be in good shape for centuries.
The ephemeral nature of digital storage does give me some pause. We can still read Sumerian tablets full of administrative trivia from four thousand years ago, but there are no known copies of some important software products from just fifty years ago.
I am a proud supporter of the Internet Archive!
100k!!
it only took 17 years to get here :)
very grateful—and excited for the next 100k!
Founders who are the realest actually do this
Alfred Lin: http://x.com/i/article/2043817564535742464
Mega pickup for USV
Should be an incredible season for them
Rebecca Kaden: http://x.com/i/article/2044051206960754693
Josh Lehman
lossless-claw 0.9.0 — the "stop touching my cache" release
🔧 engine id fix everyone's been waiting for
⏳ compaction defers while Anthropic cache is hot
⚡ bootstrap I/O off the event loop
🪓 /lcm rotate: split bloated sessions on demand
🧊 routing noise won't fake a cold cache
Figure out a little bit more about who you are before you start
Paul Graham: Don't start a startup in high school. What if it works? You'll lose the opportunity you'd otherwise have to explore random, interesting ideas, driven only by curiosity. Because while you will indeed learn a lot from a startup, you won't have any choice about what you learn.
"It's a mistake to think of the output of a prompt as a final result."
Here's @zoink on the difference between crafting with AI vs. outsourcing your thinking:
"When you're exploring divergent possibilities with agents, you need to take [AI's output] and mold and shape it like clay.
Ultimately, you're the judge of the system. You're the one that needs to figure out what is good and worth exploring because the possibilities are likely almost infinite."
tl:dr AI gets you to average fast. Your taste is what pushes past it.
📌 Watch my full chat with Dylan here: https://youtu.be/eqPljh_9C9Y
Peter Yang: "AI gets you to average quickly. Your job is to push past that."
Here's my new episode with @zoink (Figma CEO) where I asked him some tough questions, including:
→ Can you teach AI design taste?
→ Do design systems hurt creativity?
→ What's Figma's role when code is free?
"It's a mistake to think of the output of a prompt as a final result."
Here's @zoink on the difference between crafting with AI vs. outsourcing your thinking:
"When you're exploring divergent possibilities with agents, you need to take [AI's output] and mold and shape it like clay.
Ultimately, you're the judge of the system. You're the one that needs to figure out what is good and worth exploring because the possibilities are likely almost infinite."
tl:dr AI gets you to average fast. Your taste is what pushes past it.
📌 Watch my full chat with Dylan here: https://youtu.be/eqPljh_9C9Y
Peter Yang: "AI gets you to average quickly. Your job is to push past that."
Here's my new episode with @zoink (Figma CEO) where I asked him some tough questions, including:
→ Can you teach AI design taste?
→ Do design systems hurt creativity?
→ What's Figma's role when code is free?
AI Engineer
Who's coming to AI Engineer Miami? @swyx and @liamcbride will be there! Grab the last few available tickets and meet us there!
AI Engineer: Miami: We are 10 days away!
Join us in Miami and see incredible speakers like:
@thdxr of @opencode
@mxstbr of @OpenAI
@davidgomes of @cursor_ai
@jaredpalmer of @Microsoft
Tickets available! ⬇️
Fat skill fat code thin harness confirmed
Phil Chen: how I built Filbert (phil-bot)
the future of coding agents is lightweight wrappers around existing harnesses running on your own infra. it has access to your full dev env and can improve itself recursively. it's also so easy to build when you're set up for it.
Fat harnesses are just someone else’s bundled closed source skills and code
Dave Morin 🦞
2300+ builders at @UMich in Ann Arbor this week. The biggest @ClawCon yet.
Maize looks good on a lobster.
Some special things coming. Stay tuned.🦞
https://luma.com/clawconmichigan
Every 📧
You could organize your Mac. Or you could let an agent do it for you.
Welcome to the new @SparkleApp 👇
Dan Shipper 📧: INTRODUCING:
Sparkle v4 from @every
Is your desktop a disaster? We built an agent for that.
→ sparkle’s agent examines your filesystem and helps you create an organization structure that’s right for how YOU work
→ it starts with a deep clean. old installers, duplicates,
Brandon Gell
MARIE KONDO YOUR MAC
Often the most successful products are the simplest.
We released Sparkle in 2024 when AI powered apps were not really on anyone's radar. Since then, we've re-built it 4 times. Why?
Because today, you need to be willing to throw out and re-build your product every 6-months if you don't want to get lapped.
Sparkle is now isn't just the best file organizer for your Mac, it's your housekeeper, maid, and cleaner.
Today we launched:
- a total overhaul of how Sparkle organizes your computer, 100% agent native. You can create entirely custom folder structures, with each folder having it's own prompt, and you can build all of it via chat
- an end-to-end computer cleaner to re-coup 10s of Gbs of storage on your Mac. 15 years of screenshots, old dmg files, cached files, you name it, all deleted
Because if you're going to organize, you might as well purge first.
Dan Shipper 📧: INTRODUCING:
Sparkle v4 from @every
Is your desktop a disaster? We built an agent for that.
→ sparkle’s agent examines your filesystem and helps you create an organization structure that’s right for how YOU work
→ it starts with a deep clean. old installers, duplicates,
Wolfram Ravenwolf
My @aiDotEngineer Europe 2026 Highlight Reel - personal impressions from 3 days at the world's best AI conference:
Y Combinator
Are you a 100x engineer but a 1x tester with AI? @ArgaLabs closes that gap.
Arga spins up a per-PR sandbox with twins (Stripe, Slack, etc.) and in-memory dependencies (DB, Redis).
Run auto-generated E2E tests in the sandbox and send failures back to your agent so it fixes bugs before you see them.
Congrats on the launch, @PhillipLii & @akkkkiira!
https://www.ycombinator.com/launches/PwC-arga-labs-on-demand-production-like-staging-for-every-pr
Upeka Bee
Every OpenClaw today is an intern with root access & no oversight ☠️🏴☠️
So we built the first one with a boss 🦞
@getdiana is a business-ready OpenClaw with a Governor that shuts it down mid-task before damage is done 🧵
→First 500 to RT + comment “DianaClaw” get 1 month free
Daractenus
The greatest triumph of the incessant hurricane of lies coming from Russia, MAGA, and the likes of Elon Musk has been convincing a significant share of the world, including Europeans, to believe in a “decaying and collapsing” Europe that exists solely in their imagination.🧵
Daractenus
The greatest triumph of the incessant hurricane of lies coming from Russia, MAGA, and the likes of Elon Musk has been convincing a significant share of the world, including Europeans, to believe in a “decaying and collapsing” Europe that exists solely in their imagination.🧵
Mickey
Re @Replit and I
I'm excited about voice as a UI layer for existing visual applications — where speech and screen update together. This goes well beyond voice-only use cases like call center automation.
The barrier has been a hard technical tradeoff: low-latency voice models lack reliability, while agentic pipelines (speech-to-text → LLM → text-to-speech) are intelligent but too slow for conversation. Ashwyn Sharma and team at Vocal Bridge (an AI Fund portfolio company) address this with a dual-agent architecture: a foreground agent for real-time conversation, a background agent for reasoning, guardrails, and tool calls.
I used Vocal Bridge to add voice to a math-quiz app I'd built for my daughter; this took less than an hour with Claude Code. She speaks her answers, the app responds verbally and updates the questions and animations on screen.
Only a tiny fraction of developers have ever built a voice app. If you'd like to try building one, check out Vocal Bridge for free: https://vocalbridgeai.com
My hottest take is that even OpenAI is dramatically underestimating how much inference compute will be needed in the coming years.
My hottest take is that even OpenAI is dramatically underestimating how much inference compute will be needed in the coming years.
spark
Spark 2.0 is here! 🚀
We’re redefining what’s possible on the web with a streamable LoD system for 3D Gaussian Splatting.
Built on Three.js, you can now stream massive 100M+ splat worlds to any device from mobile to VR using WebGL2. All open-source.
Dive into the tech 👇
Poly News Hub
🚨 UPSET ALERT: Tech entrepreneur Ethan Agarwal has surged to a 60% chance to advance from the CA-17 primary on Polymarket. 📈
Silicon Valley is revolting against Ro Khanna over his "wealth tax" stance, and the outsider momentum is real.
Is Ro Khanna in trouble? 👇
#Polymarket #CA17 #SiliconValley
👇 Trade now:
https://polymarket.com/event/ca-17-primary-winners
Chrys Bader
enterprise is terrified of letting their employees run loose with agents
or as one exec i spoke to said “there are a lot smooth brains at our company who id never let near openclaw”
we’re going to be seeing more companies like this trying to solve trust and security issues
Upeka Bee: Every OpenClaw today is an intern with root access & no oversight ☠️🏴☠️
So we built the first one with a boss 🦞
@getdiana is a business-ready OpenClaw with a Governor that shuts it down mid-task before damage is done 🧵
→First 500 to RT + comment “DianaClaw” get 1 month free
Alex Volkov
Everyone knows who Peter is, but there's a whole team working hard to make OpenClaw work
Here's my chat with the awesome #2 Claw, Claw-uncle if you will, @vincent_koc on ThursdAI last week at @aiDotEngineer
Vincent Koc: Humbled to be given the keys to ship the biggest open-source project in the world. Did 2 in a row over the weekend without a hiccup.
Glad to see an army of maintainers supporting the project behind the scenes.
Nathan Lambert
One of my passions is that education should be dispersed freely and as widely as possible, especially for technologies as dynamic and crucial as LLMs/AI.
I'm proud to have friends who would disown me if I did a paywalled course.
Nathan Lambert: Excited to launch the accompanying free RLHF Course for my book. To kick it off, I've released:
- Welcome video
- Lecture 1: Overview of RLHF & Post-training
- Lecture 2: IFT, Reward Models, Rejection Sampling
- Lecture 3: RL Math
- Lecture 4: RL Implementation
I'm going to add
martin_casado
Sparkjs 2.0 is out! Support for arbitrarily large splats on web, mobile, and VR. Tons of features: LoD, streaming, editing, multi-splat, mesh integration, ray casting etc. etc.
If you're worried that AI is going to take over all coding, work on a splat renderer for a bit :)
Nexar
"Every truck on the road is generating data that nobody's capturing today. It happens, and it disappears."
Nexar CEO @ZMGreenb on the "What The Truck" podcast from @FreightWaves.
The data was always out there. Someone had to build the network to catch it.
So we did. 350K+ cameras. 100M+ miles/month. All capturing real life.
That network is the basis for everything we do. From BADAS, to our work with @IBM.
To see more on our network and what it powers, visit here, and let's start the conversation:
https://bit.ly/4sRSsAp
Nexar
"Every truck on the road is generating data that nobody's capturing today. It happens, and it disappears."
Nexar CEO @ZMGreenb on the "What The Truck" podcast from @FreightWaves.
The data was always out there. Someone had to build the network to catch it.
So we did. 350K+ cameras. 100M+ miles/month. All capturing real life.
That network is the basis for everything we do. From BADAS, to our work with @IBM.
To see more on our network and what it powers, visit here, and let's start the conversation:
https://bit.ly/4sRSsAp
Kieran Klaassen
I am more and more convinced that this is the future of software development UI. @cursor_ai is the closest in my opinion
A list of work you're working on parallel, the agent in the middle, and most importantly, the thing you're building on the right. Because you want to see what you're building, you want to give feedback, you want to work on it. As a builder, the code is not important anymore. As a builder, it is all about what you're building. So seeing the app next to the agent is very important.
Brian Attwell
After 8 years in stealth, we’re opening things up. Atom’s first open source project, Splitter, is live. More coming soon 🏗️
https://github.com/atoms-co/splitter
Luiza Jarovsky, PhD
🚨 In 2 weeks, a final decision on amendments to the EU AI Act and the GDPR will be made. What is at stake is nothing other than the future of Europe.
Many don't know, but the stream of events leading to this moment began much earlier, with the publication of the Draghi report on European competitiveness in September 2024.
In his long report, Mario Draghi diagnosed various areas in which European competitiveness was lagging behind and suggested that one of the reasons was overregulation and the excessive number of laws governing the digital space.
Laws such as the GDPR and the AI Act were to blame.
Before I continue, here is something many overlook: the Draghi report was finalized in September 2024, while the AI Act was officially enacted one month earlier.
The AI Act had barely been enacted, and it was already considered 'wrong,' excessive, and to blame for Europe's less-than-ideal (to be light) position in the AI race.
From that moment on, the European discourse on the protection of fundamental rights was never the same.
Its narrative shifted, and after that, the new dogma was that the path to innovation would be to "remove the red tape," and "apply the AI Act in a business-friendly way." (whatever that means from a legal perspective).
The AI Action Summit last year made this new narrative loud and clear to the public, as EU officials abandoned fundamental rights-focused statements.
Last year, the narrative shift was legally materialized. The EU published the Digital Omnibus with proposed amendments to some of its most important laws regulating data protection and AI: the GDPR and the AI Act.
Strangely, the main justification for the AI Act's amendments was the designation delays by EU member states and the work delays by EU standardization organizations.
If these were the real reasons, wouldn't it be more coherent to pressure them to move faster, hire more people, increase the budget, or help address the bureaucratic obstacles...?
Does the EU need to amend some of the AI Act’s core obligations because EU bodies are delayed?
I didn't buy it. Given the context, it felt more like a broader political shift.
The time has arrived, and in two weeks, EU officials will meet to make a final decision on the Digital Omnibus and the amendments to the GDPR and the AI Act (among other topics).
As I wrote in my newsletter, several of the proposed amendments weaken AI regulation in the EU and go against the protection of fundamental rights.
If you are European, if you were hopeful for a Brussels Effect in AI, or if you are interested in the protection of fundamental rights, I invite you to read my full article below.
-
👉 To learn more:
- Read my article about the Digital Ominibus first draft and join my newsletter's 93,500+ subscribers (link below).
- Join the 29th cohort of my Advanced AI Governance Training. Among the topics I cover in depth are the EU AI Act, the Digital Omnibus, and the European AI strategy (link below).
Samuel Spitz
Introducing Replit File Converter
Skip the sketchy ad-riddled file conversion websites.
Now, you can just use Replit instead.
Samuel Spitz
Introducing Replit File Converter
Skip the sketchy ad-riddled file conversion websites.
Now, you can just use Replit instead.
There are multiple realities of AI right now.
And what you have access to drastically changes your workflows, trust in AI, and ability to adapt to the future.
Here’s the briefest state of the AI world for business professionals (not engineers)
⬇️
Free AI - you use free ChatGPT or Claude. You think AI is decent for simple tasks but nothing to restructure your day around. You’re worried it hallucinates too much to trust with real work. On a daily basis you treat it like an intern who needs to be checked on everything.
Paid AI - you pay for one (maybe more?) AI subscriptions. You’ve had at least a few moments where AI genuinely surprised you with what it could do. You use it daily for writing, research, and analysis and you’ve stopped asking “can AI do this” and youve started asking “how do I get AI to do this BETTER.”
Super AI - you have multiple paid subscriptions, you either use Cowork or Claude Code for daily tasks, Claude Code or Codex to build tools for yourself. You treat AI like a teammate with a role, not a tool you open sometimes. You’ve automated things your coworkers are still doing manually. You have genuine opinions about which model is best for which task. AI isn’t just a productivity boost for you anymore. It’s infrastructure. You’re completely reinventing the way your work gets done.
Basic enterprise AI - your IT team approved one AI platform (probably Microsoft Copilot, ChatGPT Enterprise, or Gemini for Workspace) in the last few years, and rolled it out with clear-ish guardrails. You have access but you also have restrictions on what data you can put in. You use it for safe, surface-level tasks and your high-stakes work still happens the old way because nobody’s built the bridge between your AI tool and your real enterprise systems (ex: email and docs).
Super enterprise AI - your company has access to agentic harnesses (Codex, Claude Code) across different functions, each chosen for what it does best or just giving your employees more flexibility. Your teams have AI embedded in their actual workflows, not bolted on. Engineers have been using Claude Code/Codex/Cursor for months or years. Operations use agents for real processes. Leadership tracks AI adoption the way they track revenue (ie not just usage). The gap between your company and competitors who handed out a single Copilot license is getting wider every month.
I would go so far as to say that companies not giving their employees the best AI today are shooting themselves in the foot because their employees are losing more AI trust, putting them in a worse position for eventual (hopeful) company-wide transformation.
Kieran Klaassen
This is a BIG one! Trying it out!
Felix Rieseberg: Today is a big day! We're launching a ~ new ~ version of Claude Code in the desktop app. It's been redesigned from the ground up for parallel work and is a lot faster.
It's been my main way to use Claude Code for the last few weeks.
Kim-Mai Cutler
Update from 2026:
She nuked San Francisco’s housing pipeline by bagel-ing on a tax on new housing production. https://media.api.sf.gov/documents/Controller_Recommendation_Slides_Final.pdf
Next, she’ll nuke your ability to get home insurance if she gets into the Top 2 for California insurance commissioner with a Republican in June.
Jane Kim 金貞妍: When writing from the sidelines, it can be challenging to understand STRATEGY + WINNING.
Voters always support building high levels of affordable housing.
Market rate developers don’t because it cuts into their profits.
Joscha Bach
Every technological revolution created as many jobs as it destroyed, because we find that there is always more work to do. Things that were expensive or impossible become cheap, and we are allowed to work on things we could not afford doing before. In the future, personal transportation, architecture, legal support, medicine, education, public administration can become much cheaper and better for everyone. We may have more time to raise our children well. Don't fuck it up.
sunil pai
Who's going to be at AIE Miami! I wanna hang with you.
New version of Claude Code in the desktop app dropped today with tons of new features and performance improvements.
Between Cowork and Code, I've found I don't really need to open other apps (or even my terminal) to do most of my work now.
Claude: We've redesigned Claude Code on desktop.
You can now run multiple Claude sessions side by side from one window, with a new sidebar to manage them all.
New version of Claude Code in the desktop app dropped today with tons of new features and performance improvements.
Between Cowork and Code, I've found I don't really need to open other apps (or even my terminal) to do most of my work now.
Claude: We've redesigned Claude Code on desktop.
You can now run multiple Claude sessions side by side from one window, with a new sidebar to manage them all.
This is EXTREMELY exciting.
Claude is helping Anthropic make progress on alignment research.
A genuinely positive development that will make it more likely things go well!
Jan Leike: New research result: we use Claude to make fully autonomous progress on scalable oversight research, as measured by performance gap recovered (PGR).
Claude iterates on a number of different techniques and ends up significantly outperforming human researchers for $18k in credits.
This is EXTREMELY exciting.
Claude is helping Anthropic make progress on alignment research.
A genuinely positive development that will make it more likely things go well!
Jan Leike: New research result: we use Claude to make fully autonomous progress on scalable oversight research, as measured by performance gap recovered (PGR).
Claude iterates on a number of different techniques and ends up significantly outperforming human researchers for $18k in credits.
OpenAI fine-tuned GPT-5.4 for cybersecurity, with fewer refusals, new capabilities like binary reverse engineering, and is rolling it out to verified defenders.
Very different approach than Anthropic is taking with Mythos. We'll see how each plays out!
OpenAI fine-tuned GPT-5.4 for cybersecurity, with fewer refusals, new capabilities like binary reverse engineering, and is rolling it out to verified defenders.
Very different approach than Anthropic is taking with Mythos. We'll see how each plays out!
GBrain is rising, don't sleep on it
jun: not gonna lie, @garrytan's https://github.com/garrytan/gbrain looks really solid- it's got basically all the SOTA memory approaches i've seen in QMD and other projects
Indian Replit Builders: You can now build & pay with @Razorpay
Replit ⠕: Big news for Indian customers 🇮🇳
You can now pay for Replit with UPI via @Razorpay, alongside debit & credit cards.
And when you're ready to monetize the app you built?
Use Replit's Razorpay MCP to start accepting payments instantly.
Its never been easier to go from ideas to
Indian Replit Builders: You can now build & pay with @Razorpay
Replit ⠕: Big news for Indian customers 🇮🇳
You can now pay for Replit with UPI via @Razorpay, alongside debit & credit cards.
And when you're ready to monetize the app you built?
Use Replit's Razorpay MCP to start accepting payments instantly.
Its never been easier to go from ideas to
Spiral
Three days after launching Spiral's new managed agent, API chats outnumber in-app chats
Simaril (YC Spring 2026) is SOTA prompt injection defense for LLMs.
This is the missing link for OpenClaw for Enterprise and all agents working on mission-critical data and workflows.
The cofounders were the team that stopped billions of dollars worth of damages at http://Amazon.com and AWS.
https://www.silmaril.dev/#performance
Sriram
Everyone's building agents. Not enough are making them money-aware. We fix that tomorrow!
Fort Mason, SF:
→ Workshop on agent economics that scale with @bookercodes
→ "You Should Build $100,000 Agents Today" with @swyx
→ "What the Next Million Inference Engineers Will Need" with @philipkiely
→ Field notes from token town with @sarahmsachs moderated by our very own @rajaraman_rr
Last few spots, grab yours! http://lu.ma/5nqllmul
i think one of the most impt recursive loops in startups is startups that improve marketing/revenue growth (aka GTM) because they can obviously dogfood it for themselves. considering starting a track for this at AIE WF due to its imptnce.
https://luma.com/04bg11l5
this is @UpsideGTM + @ExaAILabs's new AI GTM meetup - i'll be looking for speakers from this crew. apply if you want to be featured.
Kenneth Roth
Not that the self-serving Trump cares, but his "decision to initiate a [pointless] war in Iran has stopped the world economy in its tracks." https://trib.al/BHIrR4y
