Our co-founder Terence Tao is announcing SAIR Foundation's inaugural competition: the Mathematics Distillation Challenge.
Co-organized by @damekdavis, Terence Tao, and SAIR Foundation.
competition.sair.foundation/…
Science x AI Summit comes to UC Riverside on June 29.
SAIR co-founders Terence Tao and Barry Barish headline a day of keynotes and panels with leaders across the science x AI frontier.
Agenda:
sair.foundation/events/scien…
SAIR Playground is an open space to test ideas, inspect outputs, and compare methods.
Use SAIR Playground to test your model for the Modular Arithmetic Challenge:
playground.sair.foundation/p…
Join the SAIR competition:
competition.sair.foundation/…
At our Science x AI Summit, I spoke on SAIR: Open by Design.
SAIR Competitions are our first step toward open Science x AI infrastructure: shared data, open models, transparent evaluation, and community collaboration around frontier scientific questions.
The 2nd SAIR Competition is now open: the Modular Arithmetic Challenge.
Can a neural network learn to compute (a × b) mod p for large numbers?
Easy for standard algorithms, but a hard test of whether neural networks can generalize the computation.
Join the SAIR competition:
competition.sair.foundation/…
To test your model:
Upload your model to @huggingface, then use SAIR Playground to run the evaluation pipeline.
You can also test locally with the tools in our GitHub repo.
Official repo:
github.com/SAIRcompetition/m…
John Hennessy: we've cleared the Turing test — but it was never a tough enough bar. AI can now write a symphony like Beethoven or paint like Picasso.
We need to start asking: what actually makes humans different from the machines imitating us?
Terence Tao: AI is creating a “traffic jam” in math
If AI generates more proofs than humans can verify, science needs new infrastructure. SAIR competitions build that infrastructure by surfacing high-quality results, so the best work is not lost in a flood of AI-generated math.
Tim Gowers: AI Is Already Producing PhD-Level Mathematics
Fields Medalist Tim Gowers reflects on how quickly AI mathematical reasoning has advanced — from solving isolated problems to generating research-level mathematics in hours.
Lenore Blum takes on the hard problem of consciousness — the subjective feeling philosophers call qualia, made famous by Thomas Nagel's question of what it's like to be a bat. Where many say this demands a new kind of science, Blum argues consciousness is computational, emerging from a world model and an inner language she calls "Brainish."
With CS enrollments reportedly dropping 50% and new graduates struggling to find jobs, Turing Award winner Jeff Ullman poses the question that matters: if vibe coding makes programmers twice as productive, do we need half as many — or does twice as much work suddenly become feasible?
His take: students will still need to learn to code, even if they mostly write it by prompting.
Randy Schekman tells the story of DeepMind founder @demishassabis — a chess champion at four who turned down a million-pound game-design salary, went to Cambridge, and completed a PhD in neuroscience before building the company that would go on to win him a Nobel Prize.
Barry Barish: How AI Helps Humanity Detect Gravitational Waves
Barry Barish explains how AI helps scientists rapidly identify gravitational wave signals, combine data across observatories, and guide telescopes toward cosmic events happening millions of light-years away.
Thank you to everyone who joined us for the Science x AI Summit. Sessions posting daily on YouTube — subscribe so you don't miss one. Event photos are now live on our events page.
youtube.com/@SAIRfoundationsair.foundation/events/scien…