• 5 Posts
  • 38 Comments
Joined 1 year ago
cake
Cake day: July 30th, 2025

help-circle



  • Reading about this from what sources I can find it’s actually more ambiguous whether or not they are profitable . It looks like they are profitable per token if you ignore the continuing R&D and capex for the next model . So they will probably be able to operate unsubsidized if they get their costs low enough some will survive. Chatgpt sol just dropped with relatively high efficiency per unit of intelligence while they were getting smoked by anthropic in real world utility but by anthropic really getting intelligence through much more brute force and thus much more extreme cost per token , which shows people will pay high costs per token for the utility they get from it , but at some point you have to compete on efficiency and input costs . The Chinese models are really efficient. If they keep dropping open source weights by 2030 we will surely have open source models as good as mythos which I would say is good enough for 99% of people. After that its pure commodity inference purchase x efficiency of the model . Maybe not a capitalist goldmine, but the end result is useful and if we get enough capacity expansion in chips and memory the deficit will turn into a glut and mythos grade home systems will be viable for power users a few years after that



  • its crazy how fast the end of the internet has accelerated. It got to the point i started learning to code last september now that i can get taught directly by AI. Hopefully i can build some useful tools for keeping a wild west alive in cyberspace. Have to use the AI tools available now because they are already leaning towards a 2 tiered system where good coding AI is no longer available to the plebians and i suspect the “subsidized by investor” plans will not last much longer. there are open source models that can do syntax well but they still have crazy hardware requirements. so best to use the shit out of it for making the software we need now before it becomes cost prohibitive or outright removed from public access. the new anthropic model is fully incapable of coding now without triggering the government mandated shutdown. can’t get it to even do a code review without triggering the terrorist suspect shutdown. the screeching about 4 horseman of the infocalypse are upon us again.



  • “I didn’t want to waste another minute of my life writing academic articles that nobody, not even the peer reviewers, really wanted to read. I didn’t want to spend another minute writing reports for deans who would rather go to wine tastings in the middle of the afternoon. I didn’t want to spend another Friday stuck in an endless department meeting, talking about problems nobody was ever going to do anything about. I didn’t want to spend another weekend grading papers that students didn’t want to write, and I didn’t want to read. Not because they were bad papers. Because there were other things we needed to be doing. There were other things my students needed to be learning.”

    This is probably like 80% of all jobs. its all just a boondoggle hierarchy of cruelty and waste.






  • There is shitloads of demand, the companies like anthropic were so compute constrained their models turned retarded for a couple months until Elon let them use his data centers.

    The companies with functional models are maxed out and growing extremely fast.

    Just the latent demand for coding alone could probably support the entire current build out globally and outstrip it soon enough.

    Tons of white collar work is demanding AI to stay competitive with other firms using it to increase productivity.

    I’ve done easily a years worth of coding in the last two weeks alone .

    With each gens capability increase it opens up more things where AI can take over. I wouldn’t call this lack of demand for at least the next 5 years when the build out capacity for chip production and other infrastructure makes a glut, though even that may be a byproduct of energy limitations more than limitations in the demand for AI







  • I’m glad my digital hygiene is good enough that I didn’t even know about tick conspiracies.

    These stats tho 😱

    Positive tests for alpha-gal syndrome have increased 100-fold since 2013; nearly half a million people in the U.S. now carry an allergy to red meat. Cases of anaplasmosis, a disease carried by black-legged ticks that hospitalizes roughly 30 percent of the people who contract it, increased 16-fold between 2000 and 2017. Babesiosis, a malaria-like illness also carried by black-legged ticks, has risen roughly 10 percent year-over-year since 2015. It’s not uncommon now for a single tick to carry two or more diseases.

    I’m in tick country now, not as bad as ozarks but I got a permethrin coated tyvek jumpsuit with permethrin coated tyvek overboots.

    If you get the 10% animal dip you can make suit spray down to .5% then let it dry before wearing. Doesn’t smell great but I think that’s more from whatever petroleum distillates carrier and emulsifiers more than the permethrin itself . Gross they use this on animals we eat though. Better than catching permanent diseases .

    Need to figure out extraction from pyrethrum daisies