Columnist @QuantaMagazine and @NYTScience. Curiosity, kindness, integrity.

New York, New York
📕Some personal news!📕 This will be my first book, an attempt to weave together the ideas and controversies of fundamental physics into a coherent narrative. Wish me luck!
134
48
931
171,912
Natalie Wolchover retweeted
1/ We discovered something totally unexpected about the neural structure of consciousness that overturns decades of thinking on its biological basis and brings us significantly closer to the recipe for generating conscious awareness. New preprint: biorxiv.org/content/10.64898… (led by @HesseJanis, with @FrankLanfranchi, @_conor_mcgrory, @Yuelin_Shi, and @EngelTatiana)
387
1,686
11,293
2,149,501
Natalie Wolchover retweeted
We want to support the mathematics community in turning progress in AI into deeper human understanding and new discoveries. To support that goal, we are contributing $5 million to a new Simons Foundation initiative to help mathematicians examine, understand, and build on results produced by AI, including the results we recently published. The funding, managed and awarded by the @SimonsFdn will support workshops, conferences, and special programs. We will also provide access to our models and computing resources to Simons Foundation-supported researchers.
How can the math community forge the best path forward in the AI age? We're committing $15M to workshops and initiatives — plus free AI computing resources for foundation-supported researchers — through new partnerships with @AnthropicAI and @OpenAI: bit.ly/4zpfIbW #math #science #AI
514
484
6,204
955,165
One fascinating question I scratch the surface of here is how the ability to process and produce language relates to intelligence. LLMs are of course fueling new research on this age-old question.
I wrote about attention, the mathematical operation that enables LLMs to process language at scale. Devised by Dzmitry Bahdanau in 2014, the trick was supercharged in the famous paper "Attention Is All You Need." I ask: 𝘞𝘩𝘺 is it all you need? Is it? quantamagazine.org/is-attent…
1
3
30
1,608
The X post that prompted my column:
I wrote about attention, the mathematical operation that enables LLMs to process language at scale. Devised by Dzmitry Bahdanau in 2014, the trick was supercharged in the famous paper "Attention Is All You Need." I ask: 𝘞𝘩𝘺 is it all you need? Is it? quantamagazine.org/is-attent…
5
15
265
21,867
It's not a stretch to credit Bahdanau with sparking the LLM revolution, yet so few know his name! He was a 23yo intern at U of Montreal at the time who'd done some competitive coding as a kid in Belarus.
I wrote about attention, the mathematical operation that enables LLMs to process language at scale. Devised by Dzmitry Bahdanau in 2014, the trick was supercharged in the famous paper "Attention Is All You Need." I ask: 𝘞𝘩𝘺 is it all you need? Is it? quantamagazine.org/is-attent…
8
43
2,964
I wrote about attention, the mathematical operation that enables LLMs to process language at scale. Devised by Dzmitry Bahdanau in 2014, the trick was supercharged in the famous paper "Attention Is All You Need." I ask: 𝘞𝘩𝘺 is it all you need? Is it? quantamagazine.org/is-attent…
8
79
543
49,890
Incredible visualization.
How do bacteria build their flagella? We’ve put together an animation of Salmonella flagellum assembly based on cryo-EM and X-ray structures. From the membrane-embedded machinery to the growing filament, here’s how the parts come together, step-by-step.
2
19
202
13,897
Natalie Wolchover retweeted
We’re releasing a broad range of new mathematical results produced by an internal frontier model. We’ve been consulting with the independent Advisory Group on Mathematics and Artificial Intelligence at the Institute for Advanced Study, and we have drawn on their advice and public recommendations to inform how we release these results. github.com/openai/math
1,899
6,307
40,745
22,680,393
Natalie Wolchover retweeted
From the guest post by Matt von Hippel who issued the challenge: 'Things definitely seem to be moving fast. In March, AI was accomplishing physics projects like a student: smaller-scale tasks with a lot of hand-holding and mistakes. In contrast, this is a real frontier calculation, the kind of thing normally tackled by the top experts in amplitudes.'
New on the Science Blog: Yes, Claude can do Nine Loops. Theoretical physicists predict how particles behave using formulas called scattering amplitudes. These are notoriously hard to compute, so researchers work with layers of increasingly fine corrections called “loops”—each added loop makes the answer more precise but takes exponentially more computation. Most calculations stop at two or three loops. Eight loops was the previous record in a simplified model physicists use as a testing ground (planar N=4 super-Yang-Mills), set by SLAC's Lance Dixon and collaborators. Last month, physicist and science writer @4gravitons issued a challenge: could an AI push past eight loops in this model, using only the compute budget an academic could reasonably access? Given a single prompt describing the nine-loop problem, Claude ran largely unsupervised for days in Claude Science and solved it using methods developed by Dixon and his colleagues, at a total cost of a few thousand dollars. Dixon independently verified the result, and von Hippel wrote about the experience for our blog. Read more: anthropic.com/research/yes-c…
10
29
377
26,386
Natalie Wolchover retweeted
This is how the model tiles to fill space.
2
2
14
1,591
Natalie Wolchover retweeted
Replying to @nattyover @kkakaes
sentences never heard before 2026 "Mathematicians have de-slopped the proof"
2
12
211
9,421
Natalie Wolchover retweeted
We are hiring, for two editor jobs that I think will be dream jobs. One will focus on the politics of science and health, and one will focus on frontiers of discovery.
11
39
178
31,448
Natalie Wolchover retweeted
Article Twitter is back! A whole week of ppl talking about articles! LETS GO I MIGHT STILL HAVE A JOB IN 2 YEARS
2
60
1,417
28,575
Natalie Wolchover retweeted
Replying to @nattyover @kkakaes
Thank you for sharing this. I am the "author" of the paper. Here is the ChatGPT thread that led to the discovery. chatgpt.com/share/6ab36b6c-2…
8
12
103
6,716