Max Pavlov

Building @steelmanlabs

[Max Pavlov @mind_user]

Jev vs Titanic, we’re safe for now

[Max Pavlov @mind_user]

@tshevl Congrats 🎉

[Max Pavlov @mind_user]

Curious if this approach would improve interpretability further than reasoning traces audit. Sound promising!

[Sasha Malysheva @aimalysheva]

meet @mostik_ai!

what happens when you put 12 PhDs in one room for four months? first place on the ARC-AGI leaderboard, which I can't say much about while the competition is still running. and this, which I can.

everyone's arguing about whether open models will catch up to frontier models. we think it's the wrong question. here's the one we pose: why does a frontier model have to generate your answer at all, when the only thing you need from it is the reasoning?

we do this by enabling models to communicate in latent space. through our protocol, hidden states pass straight from a frontier model into a small one running on your infrastructure -- no text between them, and neither model is fine-tuned. two models from different families, sharing reasoning, both left untouched.

how do we know it works? we tested it on a setup where a 753B model reads the problem, and a 4B edge-class model writes the answer. with this approach, we get results 80% as accurate as the frontier model, but at 20x faster performance.

we're committed to preventing frontier model lock-in and are already partnering with inference providers to accelerate open-weight adoption. we've done this between 15 of us, in four months, 12 PhDs and a Fields medalist, backed by @generalcatalyst

WIRED has the first external account of the company and the work: https://t.co/tP8nItCsDl
full writeup, the setup, and all the numbers: https://t.co/C9NZ5vtV1V

[Max Pavlov @mind_user]

@k_grajeda are you raising?

[Max Pavlov @mind_user]

Time machine feature is really cool. Will definitely check it out. Congrats to the team!

[Andrey Styskin @styskin]

Today we are announcing @KeenableAI, an AI-native index of the best human knowledge we have, starting with the open web.

AI seems to know just about everything until you ask it about something you know deeply. The answer isn’t wrong, but it's just very average.

We started Keenable to fix this problem: every model and every agent should be able to query, reason over, and continuously learn from the living web.

Backed by a $26M Seed from @Accel and @conviction. Built by the team that took on Google at Yandex Search, with researchers and engineers from Amazon AGI, X, and Perplexity.

The Web Search API and Web Query Language for AI are live now. Together they make the web queryable at AI scale. Both are free until end of September.

Seven days of benchmarks, stories, and launches ahead.

Let your agents search.

[Max Pavlov @mind_user]

@rubenhume I know it’s just a coincidence…

[Max Pavlov @mind_user]

@kanjun @_MathAcademy_ They’re also thoroughly calibrating where on this graph the new student is standing — it takes an hour long test, if I recall it right, I took it two years ago

[Max Pavlov @mind_user]

@fabrisera2000 Great analysis 👏
We published something similar recently — there's still a gap in what computer-use models handle natively, so teams compensate with harness complexity that shouldn't need to exist.

https://t.co/D7T2LP4Hla

[Max Pavlov @mind_user]

Here's what we're building at @SteelmanLabs @SteelmanLabs

[Steelman Labs @SteelmanLabs]

We started @SteelmanLabs to allow agents use GUIs at a human level. Here's why it's the way to solve agentic computer use.

[Max Pavlov @mind_user]

@AdamMarblestone Does it imply that it will be a non-profit?

[Max Pavlov @mind_user]

@kepano @patrickc any way to become an early tester?

[Max Pavlov @mind_user]

@alienorg @kirillzzy Let’s go!

[Alien @alienorg]

The Continuous Human Verification Protocol (CHVP) - Part I

If you're interested in how Alien ID verification works under the hood and have wondered why it is privacy-preserving, we explain all the details in our new blog post

https://t.co/8xC4WjrT5c

[kirill avery @kirillzzy]

at 21 i sold a piece of my next 15 years for capital

when you're young you have energy, ambition, and ideas, but little capital. when you're older you may have resources, but less stamina or willingness to take risks

this structure allowed me to explore ideas that are too early for vc and invest in my friends' startups

if you think about it, it's what the vc market already does: betting on founders. what's the chance of someone ambitious succeeding with their N startup rather than any particular one? it's a great deal for investors

i'm certain i will be extremely successful and i don't want to settle for a less ambitious idea just to make a little bit of money right now. artists, musicians, and athletes already do something similar

because of this structure, i was able to get Grenadian citizenship, start Alien and invest in SpaceX at a $140b valuation

technically how it works is you create a holding company and sign a pledge agreement to conduct all activities through that holding company. investors invest in the holding company (in my case, Kirill Co)

i think this is a really cool model, and i wanted to show through my own example that it can work, that more young people could do it. we'll see by the time i'm 36 :)

glad to be sharing this with the world today❤️

[Max Pavlov @mind_user]

👀

[kirill avery @kirillzzy]

introducing alien. verifying humanity in the age of ai.

mainnet alpha is live at alien[dot]org

[Elon Musk @elonmusk]

Know when to hold ‘em

[Max Pavlov @mind_user]

Wrote a blog post on how it feels to see a new beta AI-powered feature in your favorite product
https://t.co/jKxnoDwXes

[Max Pavlov @mind_user]

@prestonattebery It’s close https://t.co/IL6iEQAani

[Max Pavlov @mind_user]

@valeranazzarov Вон, там целых 5 табов внизу