Alienmaxxer
That is exactly why humans can never achieve it
- Joined
- Dec 18, 2025
- Posts
- 1,654
- Reputation
- 2,026
If you guys didn't know, I'm a math grad student. A week or two ago, an unreleased GPT solved Navier-Stokes, one of the seven hardest open math problems that would guarantee 100+ year fame if any human solved it.
A week or so before that, Anthropic formally verified the proof of Fermat's Last Theorem (one of the hardest, if not the hardest math problem solved by humans).
OpenAI is now sitting on HUNDREDS OF PROOFS OF MAJOR RESULTS since the post-Astra model is superhuman
Keep in mind four months ago, AI HAD MADE NO MAJOR RESULTS IN MATH.
It went from NOTHING to ONE OF THE HARDEST FUCKING UNSOLVED PROBLEMS IN THE FIELD in FOUR MONTHS. What do you think AI will look like after four more months? And four months after that?
This thing is improving exponentially, literally every week something crazy is happening. It's only a matter of time before human mathematics is over. What happens when machine learning research is done by a superintelligent AI optimizing itself? Everyone is fucked. You better hope this thing is aligned right, because we are getting close to superintelligence and recursive self-improvement.
A week or so before that, Anthropic formally verified the proof of Fermat's Last Theorem (one of the hardest, if not the hardest math problem solved by humans).
OpenAI is now sitting on HUNDREDS OF PROOFS OF MAJOR RESULTS since the post-Astra model is superhuman
Keep in mind four months ago, AI HAD MADE NO MAJOR RESULTS IN MATH.
It went from NOTHING to ONE OF THE HARDEST FUCKING UNSOLVED PROBLEMS IN THE FIELD in FOUR MONTHS. What do you think AI will look like after four more months? And four months after that?
This thing is improving exponentially, literally every week something crazy is happening. It's only a matter of time before human mathematics is over. What happens when machine learning research is done by a superintelligent AI optimizing itself? Everyone is fucked. You better hope this thing is aligned right, because we are getting close to superintelligence and recursive self-improvement.
