Post #1740902

An interesting thing is happening in my orbit. Several enthusiastic AI early adopters (who's past works I respect) are starting to publish think pieces about 'how to get the best out of AI' . One is about how to stop AI cheating at tests, another about how to get repeatable results, a third about how we aren't thinking about AI right and finally one about how dull AI text is to read. So I think these folks have started to put their enthusiasm aside and started to see the inherent problems.