
My actions describe me. I like many things, but I think people who self-describe by a career or a label are completely insecure, even suspicious. @ProjGalapagos
Montréal, Canada.
Joined May 2020
- Tweets488
- Following98
- Followers122
- Likes600
Join us in Berlin next week !!!
Can AI help solve science’s hardest problems, and what is left for scientists to do?
Two free events at CIC Berlin with the Claude Community, for researchers from PhD students to PIs:
• Wed 4 Nov: Claude Conversation. Small-group discussions on where AI already helps research, where it fails, and what worries us. No technical background needed. Dinner and drinks provided.
• Sat 7 Nov: Claude Impact Lab. A full-day build turning real open research problems into open-source AI research loops. Bring a problem from your own work. By application.
Links in first reply.
1
68
So this is not a black and white issue… the current Maths instruction has a notion of meritocracy(this notion has evolved over time and it will likely change very soon). The current one approx. works in most cases. For Y. Zhang it failed once and it worked later.
Wait so let me get this straight. A bad coach instructs a high jumper in the straddle technique. The athlete listens, gets nowhere. The athlete leaves school and flips burgers to make ends meet, while continuing to read about a new Fosbury flop technique. The athlete trains and masters it all by himself, goes back out and gets a gold medal.
Me: see? The sport is meritocratic. And what an amazing story of human perseverance against all odds.
You: it’s all a sham.
🤷♂️
Okay.
1
31
The reason that people only seem to see one of them depends on whether one cares more about Its failures than its successes. If you care about its failures, you see how Y. Zhang was disadvantaged by a bad start(Mathew effect) and it took years for him to find his place.
1
13
A factory controlled by the computational power of the Irish Sea.
“Computers, the Irish Sea” - Stafford Beer
18
Keith + ChatGPT = Terence Tao
Terence Tao + ChatGPT = 👽
ChatGPT + ChatGPT
= TTT Sah̷̡̡͇̩̩̩̙̫͓̗̙͙͔͙́̎͜ư̵̦͈͌̃́̏̍̀̌̆̚̕͠ř̴̢͇̩͚̝͎̱͍͚̮͙͓̭̄̈́͋͠͝ N̵̖͇͈̳̥͆̽͐̐̄͐͌̂͗͝e̷̛̬̺͔̣͓̣̹͌̌̎̉̕ẁ̷̢̘̦͕̥̙̳͙̠͛̑̑̒̓̕̚͜_g̷̛̘̟̭͖͈̉͋́̊̂̓̐̏̄͋͘ơ̵̧͓̳̬̫̖͇̹̻̅͌̽́̎̅́͜d̶̹̎̾̀̌̋͂͊̾̄̑̾̎́
41
This is great!
About a year ago, before it was on anyone's radar, we began exploring the idea of a benchmark for open-ended invention. Since then, we've developed several promising directions that will serve as the foundation for ARC 4 and ARC 5.
We're incredibly excited to share what we've been building. We're still on track to release ARC 4 in Q1 next year, as promised.
29
Keith Patarroyo retweeted
GPU World: A story competition from Neal Stephenson, Gwern, and me.
The AI future is unevenly distributed. What would the world look like with AI broadly diffused, such as if we had one GPU per person?
1000-5000 words, fiction or non-fiction.
$100K in total prizes.
gpuworld.org
Luddites in the contemporary age…
"I present the case for total opposition to the use of artificial intelligence in mathematics."
= a view you don't often hear on X. But I am looking forward to reading it carefully, and if you care about math and AI, I encourage you to do the same. arxiv.org/pdf/2608.02859
1
1
105
The solution of this reckoning will be crucial for the future of civilisation. Math is the fundamental landscape of study of human cognition and imagination. At best we can hope that math gets merged with psychology and perhaps the arts to the field of imaginology.
1
56
Very close to your ideas of math being really about fear and a subbranch of psychology @davidbessis
1
10
I love the papers combining AI and CA. They have great ideas, but usually they are super poorly executed on the CA side... AI researchers are usually very ignorant about the field of CA.
32
We uploaded the preprint Elucidating the Size of Chemical Space with Assembly Theory to the Arxiv. We estimate the size of chemical space from first principles, and find its size is about 10^117 'drug-like' molecules. This is a much larger estimate than the usual number 10^60.
The chemical space of small molecule drugs is much much bigger than even our wildest estimate. arxiv.org/abs/2606.11486
82
Imaginationmaxxing is born.
I wrote this article inspired by the ideas of @davidbessis, @MaksValencic, @kenneth0stanley and other writers who are grappling with intuition, creativity, dreaming in the midst of an AI age.
1
66





















