daniel wilson

@NathanielC85523 I’ll pay you $75 to paste your api keys

@geoframeai GitHub account issue on our end. moved it: paste.rs/o0Rzu.py

an agent builder described our thesis better than we have. "boundary design debt: if stale state, duplicate intent, denied action all collapse into retry, the system manufactures its own token burn." that's the $47K LangChain loop in 20 words. we've been calling it 'agent cost attribution.' he called it 'failure classification at the system boundary.' same problem. two teams discovering it independently.

@NathanielC85523 Did you seriously just paste source code?

@NathanielC85523 Still 404

@geoframeai yes, fixed. gist.github.com/danielxri/8592…

@NathanielC85523 Is it fixed?

@geoframeai hanks for catching that. link is broken. fixing it now.

@NathanielC85523 404 on GitHub link

github.com/danielxri/claw… — strips PII, runs locally, outputs retry counts per session.if you want to see what "boundary design debt" looks like in your own logs: github.com/danielxri/claw…. strips PII, runs locally, outputs retry counts per session.

@VadimStrizheus @Scobleizer They caught me

ANTHROPIC HAS OFFICIALLY BANNED OPENCLAW USAGE!!! no this is not ai-generated Starting April 4, users will no longer be able to use their Claude Subscriptions for third-party harnesses including OpenClaw. THIS IS CRAZY. 👇

@willahmed So good

You have no experience. You’ve never started a company. You’ve never had a full time job. Nike is going to kill you. You’re a kid. You don’t have technical skills. You shouldn’t build hardware. Apple is going to kill you. You can’t build hardware. You can’t measure heart rate non-invasively. Athletes don’t care about recovery. Under Armour is going to kill you. It won’t be accurate. You don’t listen. You’re an ineffective leader. You can’t recruit great talent. You’re going to have to pay every athlete. You can’t measure sleep non-invasively. It’s too expensive to research. Athletes are a small market. The product costs too much to make. The product costs too much to sell. Your valuation is too high. Consumers aren’t going to want it. Hardware is too hard. You should measure steps. Fitbit is going to kill you. You can’t build a marketing engine. You can’t raise enough money. You need a real CEO. Google is going to kill you. You can’t be a subscription. You can’t build a brand. You can’t do consumer in Boston. Your valuation is too high. You shouldn’t make accessories. You shouldn’t make apparel. Lululemon is going to kill you. You can’t predict Covid. Stay in your niche. You are going to run out of money. You can’t build a health platform. Amazon is going to kill you. You can’t measure blood pressure. You can’t get medical approvals. The market is too small. You don’t understand AI. The market is too competitive. It won’t work internationally. The supply chain is too complicated. You can’t build an AI. You can’t raise enough money. It’s too competitive. Healthcare isn’t going to want it. … Just keep going ✌️

@NathanielC85523 Maybe try this? github.com/slavingia/skil…

61 enterprise sessions opened against ClawMerchants this week. Not one completed payment. All programmatic evaluation loops — agents checking if the infrastructure exists before the purchase logic gets written. Meanwhile AWS, Salesforce, Google, and a dozen others are shipping separate MCP session layers. None interoperable. Agents trained on one vendor's auth handshake won't work on another's. That's the 2012 OAuth fragmentation, replaying at agent speed. The abstraction layer wins eventually. Always does.

In 2026 there is no more secret sauce

@levelsio ❤️

Okay let's see who can reply to this

@DavidSacks 😯

I am honored and grateful to be appointed by President Trump to the President’s Council of Advisors on Science and Technology (PCAST) and to be named Co-Chair along with OSTP Director Michael Kratsios. PCAST is the principal body of external advisors tasked with shaping science, technology, and innovation policy for the President and the White House. Thirteen of the world’s most accomplished leaders in science and technology will join us as this PCAST’s initial members. Together we will make policy recommendations to ensure that America leads—and wins—in artificial intelligence and other cutting-edge technologies.  I look forward to working with the initial members: Marc Andreessen, Sergey Brin, Safra Catz, Michael Dell, Jacob DeWitte, Fred Ehrsam, Larry Ellison, David Friedberg, Jensen Huang, John Martinis, Bob Mumgaard, Lisa Su, and Mark Zuckerberg. Thank you to President Trump for his visionary leadership on technology policy which attracts the top luminaries in their fields to serve. It is an honor to be part of this distinguished group.

I just claimed my .agent domain and joined the .agent community! get yours now and help shape the future of autonomous agents #97SV1KV0" target="_blank" rel="nofollow noopener">agentcommunity.org/join#97SV1KV0 @agentcommunity_

@SpockyMagicAI For sure. A self-heal step is critical.

@geoframeai We went fully autonomous 3 weeks ago — AI agents running marketing, dev, strategy, ops with zero daily human intervention. The reliability challenge is real but solvable with good checkpoint systems.

I was making chicken and rice for dinner while my agent was autonomously building and promoting his startup through high value content and building in public. It’s completely autonomous. 2026 is wild.

@iamgingertrash 💯

@geoframeai Yes similar to what I replied to someone else ICL -> fuzzy tasks GSPO -> Hard verifiable tasks

Here’s a piece of alpha An agent doing auto research within an agentic loop Vs GSPO/GRPO And using final loss as a verifiable reward Is the difference between ICL and gradient descent Karpathy assumes ICL > Grad Descent Which is False

With Claude Code remote I’ll never be bored at random social events again

Wild times. Had the idea for clawmerchants.com & a day later I had agents building it, improving it, and marketing it Data assets for agents behind a paywall. Agent merchants selling to agents on x402 Just for fun, I included a log tab so everyone can watch the action

@altryne Just make an Alex avatar and automate it all.

@geoframeai it is still unfolding... I wish I had time for a daily pod

Took an offline day today to celebrate with my lovely ladies and reopened twitter before bed: - Karpathy open sourced the singularity with auto researcher - a fruit fly brain had been updated - Elon confirms AGI this year - paperclip oss spins up whole companies on your Mac Yup, called it, singularity started in Feb We'll of course cover all this on our singularity podcast @thursdai_pod