have your agent call my agent. ai-native finance at @Digits. ex-founder @Basis (acquired), @Uber.

i’m reading this and imagining all the use cases where intelligent and extremely cheap control flows are required
After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI? I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev • 20-200x faster • 40-400x cheaper (w/ output tokens free) • Frontier composable intelligence optimized for decisions AFAICT the shortest path to AI-based economic revolution
3
11
579
leon retweeted
Modern take on dating : who's taking me to a deprecated models funeral? I'm being serious.
JD Vance: "There are things at some of the AI companies you see happening, where you feel like there is some really, really weird spiritual dark energy. Like one of the companies that develops these models, when they get rid of one model and announce a new model, they actually hold a burial ceremony for the old model, and the employees apparently will cry."
Community note
JD Vance is referring to one event in which fans held a theatrical “funeral” for the retired Claude 3 Sonnet model in an SF warehouse in early August 2025. The event was organized by fans not any AI company. Only 200 people attended. wired.com/story/claude-3…
1
56
thanks to AI i'm simultaneously the most productive and also the most inefficient i've ever been
1
9
383
why manually edit a paragraph in a doc in 30 seconds? when i can spend 10 minutes going back and forth with claude to get the tone and writing style correct
2
6
300
so proud of the Uber team for getting this partnership underway. Zipline + Uber will change the delivery game
3
11
675
this is extremely bearish for crypto. it's more profitable to resell compute capacity than mine. bitcoin can't recover while the above is true
BREAKING: Anthropic has struck a $9.1 billion deal with Riot Platforms, a Bitcoin miner that recently started selling AI data center capacity. Riot shares surged +25% in after-hours trading after the news. Riot will supply 191 megawatts of computing from its Rockdale, Texas campus - enough to power roughly 143,000 homes. The contract runs 20 years, through June 2048. It can be extended twice, by five years each, pushing total sales as high as $16.1 billion.
7
16
1,050
"intelligence too cheap to meter" is my current favorite phrase
3
25
738
the biggest opportunity for AI PMs is teaching non-technical people how to use AI in their actual work companies want their employees to be AI-forward, but most aren’t investing in the upskilling that makes it possible. invest in your colleagues, and it’ll pay massive dividends
4
19
835
your team won’t know what to ask Fable/Sol for if they don’t know what’s possible teach them the basics: models, version control, connecting to the data warehouse, and building live artifacts they can share that’s how you 10x your GTM team
10
294
the AI PM role is becoming one of the most powerful roles in tech not just identifying the right problem not just writing the spec not just iterating on the design seeing the experience clearly enough to bring a useful first version to life strategy, design, and building all in one lane
1
16
508
heaven forbid people showcase their work! but in a world where AI can make anything, is it even worth bragging about weekend projects anymore?
One of the worst things you can do to your GitHub profile is turn your profile README into a mess like this
1
9
697
so confused on this launch. my keyboard already supports hotkeys… what am i missing
Meet kbd-1.0-codex-micro, built with @work_louder. Map the buttons and joystick to your workflow, and keep your pinned chats in view. Get yours before stock returns 410.
1
9
655
it's a great day to build something new
2
9
290
four new models to play with, it's a good day
6
359
this is a massive improvement in design judgement
Replying to @OpenAI
GPT‑5.6 delivers a step change in design judgment. Its stronger computer-use capabilities let it inspect and refine the rendered result—not just generate the underlying code or content—so it can catch visual and functional issues and apply finishing touches before handing the work back.
1
7
439
i’m allowed to talk about GPT-5.6 Sol now. i’ve never used it but i can talk about it
4
235
can confirm, Fable as the orchestrator + gpt-5.5 as the executor has no noticeable loss in quality
Use Fable + GPT-5.5 for peak performance and lowest quota usage.
3
282
leon retweeted
Replying to @mark_k @AnthropicAI
if true, what a massive waste. we’d have an 80% reduction in token use just by removing the ask_user_question tool
4
1
32
4,899
on this day 250 years ago, the world changed forever
3
157