Piotr Miłoś
@piotrmilos.bsky.social
AI/ML researcher. Prof@Polish Academy of Sciences and Team Leader@Ideas NCBR. ex-visiting prof@U. Oxford. Member of Ellis Society.
www.mimuw.edu.pl/~pmilos
www.mimuw.edu.pl/~pmilos
Sure. Probably, taking any benchmark too seriously is intellectually dangerous.
Still, the fact that modern AI systems can (sort of) pass it is amazing. Something that 10 years ago seemed to be completely elusive.
Still, the fact that modern AI systems can (sort of) pass it is amazing. Something that 10 years ago seemed to be completely elusive.
April 2, 2025 at 4:30 PM
Sure. Probably, taking any benchmark too seriously is intellectually dangerous.
Still, the fact that modern AI systems can (sort of) pass it is amazing. Something that 10 years ago seemed to be completely elusive.
Still, the fact that modern AI systems can (sort of) pass it is amazing. Something that 10 years ago seemed to be completely elusive.
Read details in an excellent thread by Jan Ludziejewski x.com/jahulas/stat...
February 12, 2025 at 9:35 AM
Read details in an excellent thread by Jan Ludziejewski x.com/jahulas/stat...