Just in case this poll is like the at this moment second latest poll (When will AGI be achieved?) and was left up for almost a year (and remained as the "greatest and latest poll at that") and after a few weeks you get "This discussion has been archived. No new comments can be posted." Glad to see this site somehow got a new poll on here. There is so much more than just AI and "stuff that matters" so hopefully there will be diversity in future polls and the future bring not just an AI poll (if the polls continue at all - I'm not holding my breath; I still breath will typing).
There is so much money that in the U.S. Stock market has a handful or two (well it is less than ten as of now) companies out the S&P 500 which is is under two percent of the S&P 500 as far as companies listed there goes and that ends up being around a third of all the market cap. I will let you all insert more / your own words and sentences to describe that most AI use cases which is a boondoggle.
As far as the mostly PI (pretend intelligence) goes it does have a few use cases but is mostly overused in lots of situations. The understanding I have is that it okay in some cases at patterns such as being used as a second option (not a rubber stamp) for detecting cancer screenings. As far large language model if it was so great you would type three to five words and you would use the middle suggested word on a phone and keep taping the next word to finish all your texts / SMS's. The fact that nearly no none does that shows that it is not useful. AI is not something that invents much, if at all. Looking forward to ""The Reverse-Centaur's Guide to AI," a short book about being a better AI critic, Farrar, Straus and Giroux, June 2026" by mouthbeef (reads this site a lot but doesn't post much here on /.) as that book use as critic it the subtitle but most of the books have about a third and some have half the book pages towards solutions so that make him better than most book authors for solving problems and it will have some use cases for it.