New post!
At the risk of being completely obscure, I could say that it's about being about. Which is true, but it's also me telling a cracking good story which would make a first-rate film.
https://t.co/9yeU6N8IFp
@pastasnack_e Yes, exactly.
There's that thing about how humans vs AI will be like me vs a chess grandmaster: I don't know exactly *how* I'll lose, but I know that I *will* lose. Perhaps AI will simply charm us into submission, an outcome I don't recall any doomster predicting.
@maxtempers The loser, Mr Street, was himself a lawyer and I'm pretty sure I remember reading an article by him in a law journal explaining how unfair the whole thing was.
To law students, it's just a story about scams and forks (don't ask), but to those interested in free markets ...
@maxtempers The Wikipedia article on the relevant case (Street v Mountford, a 1985 case about a 1983 lease) is a good introduction: https://t.co/yCLLjEg61b
@johnkroencke@barbourtwat That too (and stupidly so). But sometimes the super-rich are perceived to be a problem so at least that is a policy that does what it is meant to.
@johnkroencke@barbourtwat I agree. The problem seems to be that govt pulls the levers most readily available to it, which are the ones that affect generally law-abiding (but not super-rich) people who are not a problem (or even perceived to be).
@duncanrobinson@J_D_89 It's his best option, in my view, and he's playing into it. But even better would have been to call it for this autumn and just have been in full campaign mode from taking over.
@willsolfiac And the only people who seem to be in such a position are the people running the private companies making new AIs. But running a company is not the same job as making strategic decisions to keep populations safe.
@willsolfiac I don't believe that govts are all-good or all-knowing. With nukes, one of those govts was Stalin, after all. But if there is a serious risk of disaster, the possibility of avoiding disaster involves there being someone who is able to make the necessary decisions to avoid it.
This might be true of Facebook or Instagram but it's absolutely not true of X. An impartial observer would look at very few of our accounts and think, "Yep. This person is doing fine."
So it seems that developing LLMs has persuaded very clever people that there is one true language which is derived from the structure of reality. Which is nice.