• 0 Posts
  • 37 Comments
Joined 18 days ago
cake
Cake day: September 19th, 2026

help-circle




  • This goes for a lot of genres. Metalheads are like, “dude listen to this ultra black metaldeathcore album my favorite bad just dropped!” and it’s nails on a chalkboard, but also incredibly technically impressive with shifting time signatures and somehow if you play it backwards through a filter it sounds like Flight of the Valkyries. If you enjoy that, power to you, but most don’t.


  • A lot of all entertainment media is garbage. Maybe anime has a lower barrier to entry, so the signal to noise ratio is worse, but then again, the big movie studios are so risk averse that they mostly pump out boring, focus-group-tested sequels and reboots. Wait maybe there’s a barrier-to-entry vs financial-risk sweet spot that produces the greatest share of good stuff? Hmmm…









  • Depending on where the figure is sourced, the water usage may include the electricity generation, in which case you might be using tons of water with a closed loop (data center) cooling system, because the power plant uses an open loop system. Open loop is wasteful, but if the climate is dry, it’s more effective because you get to reject the latent heat of vaporization of water, which is a shitload of energy. (That’s the technical term forgive my jargon.)





  • I think it’s more that the way the chatbots work is the same. They assign a bunch of numbers to tokens, and the numbers determine how interrelated different tokens are. You put in a bunch of tokens and then it spits out what it thinks the next one should be based on all the interrelatedness numbers and some math things like transformers, and the math things have knobs on them to determine how strong they are. It doesn’t matter if the tokens are words in a sentence or peptides in a protein. The underlying technology is the same.

    This is an exciting use for a tool that I mostly despise for probably all the same reasons you do.


  • I want to know where they think this will go, because there is an obvious and looming problem: the content–the library–wasn’t made or at least hosted for free.

    Somebody set up a website to host the interactions between people who knew things and people who wanted to learn things. The people exchanging information often got as much as they gave, and the person hosting the forum typically sold ads or information about the people participating in the discussions. The result was a virtuous cycle where content gets made and everyone is happy.

    Now all those sites are scraped, and their information is regurgitated by slop bots into a “zero click search” where the creators of the information have no role, the seekers of information never encountered the source, and the hosts never get their slice of the pie. OK, well that works for now, but what about when something new is made and there’s no forum to talk about it because all forums get scraped into the slop machine? How will the slop center learn? Are we supposed to expect the AI companies will spend time meticulously testing how best to DIY a mass spectrometer or whatever?


  • Not to “My generation good other generations bad,” but Millennials and Gen X really grew up with technology as it was becoming impressively capable, but required a lot of technical know-how to use. It’s really apparent in the workplace: our fresh-out-of-college guys and ready-for-retirement guys alike come to people around my age when stuff isn’t working. Everything’s too easy now, and it’s hard to look under the hood. (Of course, there are still exceptions for enthusiasts, but you’re never forced to tinker with a raspberry pi.)