4 ms·

The original tweet was clearly misunderstood... https://x.com/SebastienBubeck/status/1977181716457701775 https://x.com/SebastienBubeck/status/19771817164577017

by Palmik 1y ago

The original tweet was clearly misunderstood... https://x.com/SebastienBubeck/status/1977181716457701775 https://x.com/SebastienBubeck/status/1977181716457701775: > gpt5-pro is superhuman at literature search: > it just solved Erdos Problem #339 (listed as open in the official database https://erdosproblems.com/forum/thread/339 https://erdosproblems.com/forum/thread/339) by realizing that it had actually been solved 20 years ago https://x.com/MarkSellke/status/1979226538059931886 https://x.com/MarkSellke/status/1979226538059931886: > Update: Mehtaab and I pushed further on this. Using thousands of GPT5 queries, we found solutions to 10 Erdős problems that were listed as open: 223, 339, 494, 515, 621, 822, 883 (part 2/2), 903, 1043, 1079. It's clearly talking about finding existing solutions to "open" problems. The main mistake is by Kevin Weil, OpenAI CTO, who misunderstood the tweet: https://x.com/kevinweil/status/1979270343941591525 https://x.com/kevinweil/status/1979270343941591525: > you are totally right—I actually misunderstood @MarkSellke's original post, embarrassingly enough. Still very cool, but not the right words. Will delete this since I can't edit it any longer I think. Obviously embarassing, but completely overblown reaction. Just another way for people to dunk on OpenAI :)

Topfi 1y ago
If holding the CTO of OpenAI accountable for his wildly inaccurate statement constitutes "dunking on OpenAI", then I'd say dunk away. He, more than anyone else, should be able to for one parse the original statements correctly and for another maybe realize that if one of their models had accomplished what he seemed to think GPT-5 had, that may require some more scrutiny and research before posting it. That would have, after all, been a clear and incredibly massive development for the space, something the CTO of OpenAI should recognize instantly. The amount of people that told me this is clear and indisputable proof that AGI/ASI/whatever is either around the corner or already here is far more than zero and arguing against their misunderstanding was made all the more challenging because "the CTO of OpenAI knows more than you" is quite a solid appeal to authority. I'd recommend maybe a waiting period of 48h before any authority in any field can send a tweet, that might resolve some of the inaccuracies and the incredibly annoying need to just jump on wild bandwagons...
zozbot234 1y ago
"you are totally right—I actually misunderstood" ...like, seriously? Did an AI come up with this retraction, or are humans actually talking like robots now?
- Topfi 1y ago
  Guess even the CTO of OpenAI relies on Anthropic models in a pinch...

by Palmik 1y ago

Topfi 1y ago

If holding the CTO of OpenAI accountable for his wildly inaccurate statement constitutes "dunking on OpenAI", then I'd say dunk away. He, more than anyone else, should be able to for one parse the original statements correctly and for another maybe realize that if one of their models had accomplished what he seemed to think GPT-5 had, that may require some more scrutiny and research before posting it. That would have, after all, been a clear and incredibly massive development for the space, something the CTO of OpenAI should recognize instantly. The amount of people that told me this is clear and indisputable proof that AGI/ASI/whatever is either around the corner or already here is far more than zero and arguing against their misunderstanding was made all the more challenging because "the CTO of OpenAI knows more than you" is quite a solid appeal to authority. I'd recommend maybe a waiting period of 48h before any authority in any field can send a tweet, that might resolve some of the inaccuracies and the incredibly annoying need to just jump on wild bandwagons...

zozbot234 1y ago

"you are totally right—I actually misunderstood" ...like, seriously? Did an AI come up with this retraction, or are humans actually talking like robots now?

Topfi 1y ago
Guess even the CTO of OpenAI relies on Anthropic models in a pinch...