why do all these papers have their own microsites, dedicated domain names (in Anguilla no less!), and shitty graphics? I can’t think of another branch of research that consistently does this shit. it’s almost like it’s all marketing fluff or something!
I feel the same way. I program in rust cause I like it, and the feeling of actually liking writing systems code was refreshing coming from C and especially C++. rust is a language I find beautiful — but I won’t for long if its excellent diagnostics and tooling all get deprecated in favor of an LLM. I can’t imagine what pivoting to AI would do to the language’s roadmap.
wow what utter horseshit
AI, as I’m sure you are all aware, is a very old concept with useful applications dating back in the 80s.
the first AI winter happened because those fuckers couldn’t stop grifting academic funds by promising shit that didn’t work. we know the history of the field better than you do. not that you had a point other than wanting to reply guy about a name we didn’t adopt (cause we’re not OpenAI) for a technology all of us strongly dislike.
fucking pointless shit
holy shit that’s perfect
Niko Matsakis wrote a post about how the fucking rust compiler toolchain should include an LLM to explain error messages because teaching the semantics of a language is too hard and that pissed me off so much that instead of linking that piece of shit directly I’m posting this excellent sneer from Anatol Ulrich that’s also much shorter than Niko’s extended attempt to beg for a promotion at Amazon
my related guess is primarily marketing surveillance and secondarily all other types of surveillance. notably Venmo no longer lets you put “hookers and blow” as a transaction note, because it was the default response from most people who didn’t really want to do a social transaction but had to use Venmo cause it’s all their recipient had. Venmo’s social features are all designed to make you leak as much data to Venmo as possible so it can be monetized or otherwise capitalized upon, and that’s about par for the course for how a thielverse paypal mafia offshoot operates. this is surveillance capitalism with a smiley face.
it’s turning out the most successful thing about deepseek was whatever they did to trick the worst fossbro reply guys you’ve ever met into going to bat for them
also, holy fuck their post history is essentially nothing but unsubtle dogwhistles and pro-AI garbage
how about you go fuck yourself
The only thing LLMs typically are bad at
is everything. including summarizing research since it’s pretty fucking obvious you didn’t read shit. now fuck off
somehow it got even worse
Google appears to have faked AI output in a commercial set to run during the Super Bowl. The ad shows a business owner using Gemini to write a website description, but the text portrayed as generated by AI has been available on the business’s website since at least August 2020
also they doubled down on the bad stat
The ad originally had Gemini present copy stating that Gouda accounts for “50 to 60 percent of the world’s cheese consumption” — which is not true. Google later edited the commercial to take out the stat, while the business owner also removed it from their website.
[…]
But Google maintained that the website description was written by Gemini all along. In addition to showing Gemini “generate” the description in the commercial, Google Cloud apps president Jerry Dischler said on X that the Gouda stat was “not a hallucination,” adding that “Gemini is grounded in the Web.”
also also they later doubled down on lying that Gemini wrote the whole page? it’s… really embarrassing that Google’s marketing team doesn’t know about web archives
standard “fuck off programming.dev” ban with a side of who the fuck cares. deepseek isn’t the good guys, you weird fucks don’t have to go to a nitpick war defending them, there’s no good guys in LLMs and generative AI. all these people are grifters, all of them are gaming the benchmarks they designed to be gamed, nobody’s getting good results out of this fucking mediocre technology.
wait a fucking minute
you came here to lecture us and you think algorithms from the field of AI only started seeing serious use in the 80s? the Mark I Perceptron was built in 1958
(I admittedly didn’t know about the machine before today, but I know more than enough about AI to know perceptrons as software are old as hell)