JoshuaZ comments on Progress on automated mathematical theorem proving? - Less Wrong
You are viewing a comment permalink. View the original post to see all comments and the full post content.
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (65)
More explicitly, there are a lot of things that humans can do easily that seem to be extremely difficult for AI (vision and language are the most obvious), and there are many things that humans have trouble with that we can easily make AI do a decent job. At some level, this shouldn't be that surprising because a lot of the things humans find really easy are things we have millions of years of evolution optimizing those procedures.