Over 150 mathematicians issue Leiden Declaration warning about AI errors
More than 150 mathematicians are pushing back on the hype around AI in math, saying it's not as flawless as some claim.
Their Leiden Declaration, supported by Ulrike Tillmann from the International Mathematical Union, stresses that real human judgment and ethical standards are still essential.
They point out that AI can create convincing but wrong proofs, making it harder for people to check what's actually true.
Declaration warns AI risks peer review
The group worries that relying too much on AI could mess up peer reviews and even give credit to algorithms instead of actual researchers.
Columbia professor Michael Harris called out companies for using math mainly to impress investors.
The declaration urges mathematicians to think carefully about their projects and step away from work that could do more harm than good, especially when trends push for quick AI advances over solid science.