Wasn't that like his last post ever though?
Him not being an overt eugenics enthusiast while also not being the popular face of AI scientology probably helps ingratiate him to people here. Additionally, even though admittedly I haven't really bothered to revisit since he stopped posting like a decade ago, whatever overall sociopolitical agenda he might have had can't have been as glaringly obvious as siskind's, which can make for some inconsequential reading.
I see, thank you.
this isn’t really even related to GenAI at all
Besides the ocr there appears to be all sorts of image-to-text metadata recorded, the nadella demo had the journalist supposedly doing a search and getting results with terms that were neither typed at the time nor appearing in the stored screenshots.
Also, I thought they might be doing something image-to-text-to-image-again related (which - I read somewhere - was what bing copilot did when you asked it to edit an image) to save space, instead of storing eleventy billion multimonitor screenshots forever.
edit - in the demo the results included screens.
Echoing the audience's fawning of heavyweight boxers is probably the least objectionable thing in this racist shitheap of an article, I like how it ends by basically saying people should shut up about the judges possibly favoring Usyk for being Ukrainian, not because that's just Tyson fans coping but because the current notable russian heavyweights are either icky muslims or not full whites by parentage.
P4P is mostly a marketing term anyway, size aside the meta is different enough between distant weight classes to really strain comparison.
lemmy thread here: https://awful.systems/post/1596726 which spun off from here: https://awful.systems/post/1587716
Either that or he let his performative contrarianism get out of hand, he did delete the post after all.
Still, it's just like an HBD enthusiast heavy into eugenic optimisation to think that there might be something to measuring skulls, even if it didn't pan out the first time, maybe if they had known about IQ it would have been different, it's a shame the woke mob has made using calipers on school children a crime, etc.
This is almost the plot of The Fifth Season, a hugo winner from a while back.
Yet AI researcher Pablo Villalobos told the Journal that he believes that GPT-5 (OpenAI's next model) will require at least five times the training data of GPT-4.
I tried finding the non-layman's version of the reasoning for this assertion and it appears to be a very black box assessment, based on historical trends and some other similarly abstracted attempts at modelling dataset size vs model size.
This is EpochAI's whole thing apparently, not that there's necessarily anything wrong with that. I was just hoping for some insight into dataset length vs architecture and maybe the gossip on what's going on with the next batch of LLMs, like how it eventually came out that gpt4.x is mostly several gpt3.xs in a trench coat.
Now I really wonder how much EA donation money went into making this.
It would have to be something specific but recurring, like tescreal guru accidentally-but-not-really advocates eugenics on main, or yudkowski running afoul of actual AI researchers on twitter and getting roundly mocked.
But where's the fun in canned sneering?
Maybe we should have a weekly thread for meta discussion and sneers that are too low effort to be their own post?
...with a huge chip on his shoulder about how the system caters primarily to normies instead of specifically to him, thinks he has fat-no-matter-what genes and is really into rape play.