@JimDMiller@Neil_Jetter Or just playing this one out… you asked it to write 100 versions of a thing and to rank them and it has been trained to not blow tokens.
The best teacher I ever ever had just did lots of in class essays and graded hard. It sucked but you read the thing and everyone read the thing. It’s always been pretty simple. BUT the teacher has to care. I think the AI reckoning in education is more about a lot of teachers mailing it in being met, now, by students mailing it in in a way they can’t catch.
@NateSilver538 In fight sports, the little I know, this is indeed almost always a cope. If you are trying hard, it feels morally good but you are probably losing to someone with better techinque.
I used to run a bcorp tech co and have served on the board of a few tech nonprofits.
My (admittedly anecdotal) view is no institutional structure is itself going to achieve that goal. The good news is these companies, however they are structured, are creatures of US law.
If and when things get sufficiently serious, the lever for achieving these goals is through are our actual democratic institutions.
Same. Works for certain categories of software problems really well but in others it spirals into nonsense. I think that’s a good test for where AI is going to be most disruptive in knowledge work generally in the immediate term: places where definition of success is ultra clear and you can run lots of iterations eg stock trading, ranking songs on Spotify, ppc ad design, etc.
@datarade@BrianNorgard If it’s consumer use cases being saturated by local compute, question becomes which company captures the margin? Apple is good at that.