Why Mathematical Research Still Requires Human Involvement

Written by

in

The author argues that mathematics will not end with the involvement of artificial intelligence, but will shift toward research directions that rely more heavily on human judgment, taste, and understanding. The article mentions that an unpublished internal OpenAI model once solved the Navier-Stokes equations in September, with a swarm of agents made up of approximately 10,000 model copies approaching the derivation from different angles, after which other computer systems filtered the results and followed up on them. Mathematicians currently cannot confirm whether the proof is valid or what reasoning process the AI used, while the selection of problems, interpretation of results, and value judgments still require human involvement.

Source:Scientific American · Read Original

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *