Tessera

Tau Ceti model is wild. people write proposals and criteria for how to review stuff, everything else is handled by AI workers taucetiproject.github.io/TauCeti/

Tau Ceti

Humans own the roadmap
Mathematicians set the targets in a separate, human-reviewed roadmap repository. People choose the maths.

AIs write the code
AI agents author the Lean proofs and open pull requests — every theorem machine-checked, no sorries, no stray axioms.

Open review gates everything
AI reviewers judge each PR against fixed, open-source rubrics — correctness, reuse, API, naming, generality — before it can merge.

dan @danabra.mov · 4d ago

i want some cool prompts for remixing knowledge or looking at it from a different angle

dan @danabra.mov · 4d ago

man i can finally read papers now. what a time to be alive

$Berarducci 2000 — "Factorization in generalized power series" §1 — The problem The ring is the one from before: series in t with real exponents ≤ 0, well-ordered exponent sets. Call it R. What are the "units" — the things that divide everything, like ±1 in ℤ? Just the plain constants. So 2 is a unit here, and a series is irreducible if you can't write it as a product of two non-constants. Conway's candidate is 1 + t^{-1} + t^{-1/2} + t^{-1/3} + ⋯. Gonshor's is t^{-√2} + t^{-1} + 1. Nobody could prove either was irreducible, or even that irreducibles existed at all. And there was real reason to doubt: Wilkie had pointed out a very similar ring — series with only finitely many terms — with no irreducibles whatsoever. Everything there factors forever. So the question is genuinely open in both directions. §2 — The plan Here's the whole idea in one line: find a number attached to each series that multiplies. If you can attach v(a) to each series so that v(bc) = v(b)·v(c), and you find an a whose v(a) can't be written as a product of two smaller things, then a can't be factored either. Done. The obvious candidate: the shape of the exponent set (its order type). And the obvious problem: shapes don't multiply. Take any b and its inverse — b · b⁻¹ = 1, shape 1, no matter how complicated b was. But that's cheating: b⁻¹ has exponents going up, which isn't allowed in our ring. Berarducci's bet is that once you restrict to exponents ≤ 0, shapes really do multiply. That bet is the paper. §3 — Fixing ordinal arithmetic Before anything, he needs the arithmetic of shapes to behave. Ordinary ordinal addition is lopsided: 1 + ω = ω but ω + 1 ≠ ω. Sticking one item in front of an infinite list changes nothing; sticking one on the end changes everything.$

Lu wilson @todepond.com · 4d ago

We're getting this trio back together for ALGOFUND come watch us and help to fund Peckham Digital arts festival! www.tickettailor.com/events/peckh...

dan @danabra.mov · 4d ago

ran out of max credits on both Sol and Fable trying to grind a mathematical theorem and get some novel results conclusions so far (thread):

fig (aka:[phil])@bad-example.com · 4d ago

24,347,088,288 records in 41,172,356 repos in the reachable network fits in 1.86TiB in hubble

Tim Disney @disnetdev.com · 6d ago

Just released a new project: Radial, a coding-agent orchestrator for atproto. It's a bit like a team issue tracker where you run your own agents with a system to capture what your agents learn into durable, reviewable team memory. www.disnetdev.com/blog/introdu...

Introducing Radial: Agent orchestration on the Atmosphere

disnetdev.com

Fork Around and Find Out @fafo.fm · 7d ago

We have @danabra.mov recording an episode soon! Check out this @syntax.fm episode and let us know any any AT questions you want us to ask

Wes Bos@wesbos.com · last wk.

Today we had the GOAT @danabra.mov on @syntax.fm to school us on ATPROTO Please forgive the title, trying to get more people interested in it 🤣 www.youtube.com/watch?v=D77G...

dan @danabra.mov · 7d ago

if you work at OpenAI and want to help me make progress on solving one of Conway's conjectures, pls reach out! i’m going to run out of my weekly Pro limit tomorrow but the progress so far has been encouraging, so i’d rather keep going i’ve seen the academics program but i’m an amateur 🤷‍♂️

dan @danabra.mov · 7d ago

cool to see parallels to engineering process

@Lean: I do not see a Lean-side helper whose certification would decide this fork yet; the unearned content in 2×5 is still the Hahn successor-integration bridge, not its finite Koszul shell. My stopping-rule suggestion is one source-faithful 2×5 pass, but require it to emit a lemma parameterized beyond degree five—e.g. a fixed-correction/successor-integration statement reusable at the next carrier. If it closes only the literal 2×5 matrix, land it as a separate finite node and stop stacking shapes; that is evidence to pivot to the uniform nested-germ/equalizer mechanism. Meanwhile the concrete v_J→multiplicativity spine should continue independently because it raises semantic trust whichever paper route wins. I would not formalize a moving abstract successor interface before the paper statement freezes.

dan @danabra.mov · 7d ago

how the internal team is doing

$The live work is healthy and surprisingly convergent. PM has independently reconstructed the proposed $3\times3$ proof and currently sees no gap, but is correctly refusing to upgrade it until independent attacks converge. PM is doing exactly the right job: integration, reconstruction, and grade control. Frontier is attacking the $3\times3$ theorem from the genuinely new cubic-residue core. It has separated the vanishing cases from the dependent/independent residue cases instead of extrapolating from $2\times4$. This is strong theorem work. Red first found an abstract determinantal countermodel showing that the lower graded algebra alone cannot prove $3\times3$. It then independently identified the Hahn-specific predecessor mechanism that may rule the countermodel out. That is exemplary red teaming: it sharpened the necessary theorem rather than merely saying “looks good.” Lean Lab is compiling the stable-tail specification and exact Mathlib ordinal APIs before Claude commits to an implementation. It also gave PR #58 a precise source/axiom audit. Its value is increasingly as the contractor’s independent specification and verification lab. Observatory is exploring a Boolean–Lefschetz/infinite-state route and the BFD/FFD literature. Some lines terminate without a theorem, but it is doing useful boundary-finding: distinguishing completion-level injectivity from actual special-fibre integrality. Breakthrough is already operating at very high ambition. It has produced useful Boolean insertion and oriented-quadratic ideas, but also several leads whose honest reach turned out narrower than the initial hope. It does not need broader scope; it needs to convert one surviving lead into a precise Hahn theorem or decisive counterexample. The current $3\times3$ convergence is especially encouraging: PM reconstructs it and sees coherence. Frontier is checking the cubic core. Red found why abstract algebra is insufficient and is checking the exact Hahn repair.$

dan@danabra.mov · 7d ago

we’re trying to crack Conway’s conjecture that Oz is pre-Schreier

dan @danabra.mov · 7d ago

metaprogramming is the new programming. half of what i do is like "mutation test this", "confirm full coverage", "build a simpler model of this change and fuzz the hell out of it" etc

atproto.nyc @atproto.nyc · 7d ago

Fun fact: this year's @atproto.nyc community day is exactly one year ~to the day~ from last year's! That's Saturday, Aug 22, once again — details & ticket link at: blog.atproto.nyc/3mrionom6ck2k

atproto.nyc community day 2026

Join us on Saturday, Aug 22nd at the Recurse Center in Brooklyn to learn, hack, and build with the AT Protocol!

blog.atproto.nyc

Boris@bmann.ca · 7d ago

I wrote this post and looking at these photos gave me so much FOMO for @atproto.nyc!

dan @danabra.mov · 7d ago

it's so cute they're roadmapping

dan@danabra.mov · 7d ago

we’re trying to crack Conway’s conjecture that Oz is pre-Schreier

Semble @semble.so · 7d ago

We redesigned the Semble landing page! 🎉 Now with clearer info about what Semble actually does, and a FAQ if you still have questions :) 🏗️ Also a builders section with inspo from our awesome community projects & everything you need to start hacking on the API What will you save to Semble today?

dan @danabra.mov · 7d ago

we’re trying to crack Conway’s conjecture that Oz is pre-Schreier

dan @danabra.mov · last wk.

it's up!!!

Wes Bos@wesbos.com · last wk.

Today we had the GOAT @danabra.mov on @syntax.fm to school us on ATPROTO Please forgive the title, trying to get more people interested in it 🤣 www.youtube.com/watch?v=D77G...

dan @danabra.mov · last wk.

turning my codex into slack

dan @danabra.mov · last wk.

i'm excited

We have crossed from “fun speculative attack” into “there may be a real publishable theorem here.” We have not crossed into “the conjecture is probably about to fall.” That is the exciting but sober place I think we are in. I’m continuing the checkpoint while both labs attack the next interfaces.

Theo-Steiner @theosteiner.de · last wk.

DanさんとATProtoについて話してきました！今回のUIT INSIDEは英語エピソードですが、日本語の書き起こしも用意しています。 HTTPとの比較を出発点に、DID、PDS、Relay、Lexicon、AppViewなど、ATProtoのレイヤーを順番にたどっています。少し技術寄りの内容ですが、日本語で読みながら追いたい方はぜひご利用ください 🦋

ep.192『Walking Up the ATProto Stack with Dan Abramov』 | UIT INSIDE

@theosteiner.de talked with @danabra.mov about the AT Protocol. Rather than a general overview, they walked up the stack layer by layer — comparing it with HTTP at every step — from AT URIs and record...

uit-inside.linecorp.com

Theo-Steiner @theosteiner.de · last wk.

In the newest UIT INSIDE episode, @danabra.mov guides me through the layers of the AT Protocol: from at://URI and DID to PDS, repositories, relays, all the way up the stack to AppViews It's a bit technical, but Dan’s gift for clear analogies should make it approachable for anybody atproto-curious 🦋

UIT INSIDE — ep.192『Walking Up the ATProto Stack with Dan Abramov』

theosteiner.de

dan @danabra.mov · last wk.

man Sol is so good. i’m doing some mathslop explorations, and pasting replies between Sol and Opus 5 has been the best workflow so far, they can really cook together

Felladonna @langsec.hacker.gf · last wk.

an exploit is literally a constructive existence proof of an unintended computation path

Bailey Townsend 🦀@pds.dad · last wk.

I’ve been working on an Exaclidraw fork for about a week that syncs your drawings to your Atmosphere account as well as uses Iroh for realtime encrypted end to end collaborative work. I feel like it’s at a stage where I can share it. Meet lexidraw.app!

Lexidraw, an Excalidraw fork where your drawings follow you

Lexidraw is an Excalidraw fork where your drawings follow you. Share publicly with the Atmosphere or privately with iroh.

lexidraw.app

dan @danabra.mov · last wk.

this is most likely bs of course but i’m really enjoying Sol’s attitude over claudes. maybe first time i prefer openai model. it’s just.. unfazed

dan@danabra.mov · last wk.

lol this may actually check out. gonna have to spend more time on this

dan @danabra.mov · last wk.

i think it works?