Scientists warn governments about AI capabilities and risks (opens in new tab)
More than 150 mathematics experts signed the Leiden Declaration on AI and Mathematics, warning governments not to “believe the hype” about claims that current AI systems can solve complex mathematical problems at a revolutionary pace. The declaration followed claims that ChatGPT helped a 23-year-old without formal mathematics training work on an Erdős problem and that OpenAI’s AI had disproved an 80-year-old Erdős unit-distance conjecture. Anthropic separately warned that AI systems could mov...
Read the original article