ChatGPT
While it is true that there is too much hype surrounding #ChatGPT—which is mathematically very weak—one should not forget the dramatic recent progress in #machinelearning and #deeplearningai that are right now doing some amazing things (1/6)
• solving the chaotic three-body problem (2/6) arxiv.org/abs/1910.07291
• solving physics problems generated by Minerva 540B (3/6) ai.googleblog.com/2022/06/minerv…
• solving differential equations and integrals (4/6) ai.facebook.com/blog/using-neu…
• discovering PDEs in integral form from sparse and noisy data (5/6) arxiv.org/abs/2011.11981
• obtaining universal DEs (6/6) arxiv.org/pdf/2001.04385…
And there is also this—which I find *very* impressive:
openai.com/blog/formal-ma…
Plus I’d forgotten about the Ramanujan machine. Interestingly, only *some* of its conjectures have been proven… ramanujanmachine.com
And one more: Predicting zeros of the Riemann zeta function: sci.sdsu.edu/math-reu/2018-…
Share this Scrolly Tale with your friends.
A Scrolly Tale is a new way to read Twitter threads with a more visually immersive experience.
Discover more beautiful Scrolly Tales like this.