@cstross Charlie, IMHO the only palusible use for LLMs is accessibility. To allow for example, dyslectics, to be able to "compete" or have an equal share in "democracy" and/or society. But to use 10% of GDP, fucking up all business domains, indigenous areas and nature just for a glorified automated prostethis is just - - - not for real. Of course all LLM idiocracy is driven by greed. Then some "nice" applications arise and by the time the vulture capitalists want their moneys back... poof! 1/2
@cstross Ha ha!! It's interesting... to receive my "welfare" i have to attend AI-job-seeking-seminars every now and then. Instead of personal meetings, employers and employees engage a meta-circus of LLM-processing. Increasing the number of plausible connections by orders of multitude. More LLM use. Begets even more.
Why not see to it that people can live on a wage (so scamming need not scale) and we could use all that energy and nature for better things. Like enjoying it IRL. Meh!!!!
@cstross Love the optimism! Thank you, of course you are right. The other "smolk" clouding the outset is my own estimated lifespan. Hope to remain to see that future.
@thomasfuchs OMG! Saw you previous posts. Only now do i realize that the box fits into a homecomputer, the TI-4A (as i remember it). But it's the same! I had one. I used it. Never knew about the expansion box :( It's in a box somewhere. It sat in my parents attic for some years where lots of pollen got stuck to it and ruined it's appearance.
I got it on a computer-fair when a "dude" realized i could actually write software. I was 13. Ha!
@thomasfuchs Not too disagree, but would new (and faster) hardware been developed of new (and more complex, sometimes more inefficient) software would not?
A life full of non-profit engagements. Also ran a one-man shop developing for non-profits for 30+ years. Avid SF reader, mulling over development patterns and... the jackpot.