• 5 Posts
  • 21 Comments
Joined 1 year ago
cake
Cake day: June 6th, 2023

help-circle
  • The last time I went to a doctor, they read a list of questions from a form, entered my answers into their system, and then said they’d get back to me in a couple weeks to tell me if my insurance company would allow a follow-up. That appointment should have been a web page.

    Most doctor’s appointments I’ve had recently have followed the same pattern. A good doctor is invaluable. A burnt-out noob doctor following strict procedure is like a worse GPT that your have to meet in a building full of every conceivable virus, and that costs $500 instead of $0.05. A motivated layman with GPT4 and a prescription pad would have beaten 3 out of 4 doctors I’ve seen since covid.

    This is just my experience in the US mind you. Maybe I’ve had bad luck with humans, but I haven’t been impressed since all of the experienced ones retired.









  • I have a similar build for similar reasons. It works great, though I use Windows, so no driver issues (VR introduces too much jank with Linux). Notes below.

    CUDA is essential. Definitely the right call paying the Nvidia tax.

    My Gigabyte 4090 works for LLM stuff without a second card, and has no coil whine I can hear. I use Alpaca 4-bit entirely in VRAM, and SDXL runs like a dream. I only have 48GB of RAM total, but VRAM is pretty much always the limiting factor (if I understand correctly, it works best when you have at least enough spare RAM as you have VRAM when you’re loading the model, but after that the computation is on the GPU if you have a 4090. Moving layers to the CPU/RAM drops performance fast). I have an A4000 in another machine that I was planning to add with a riser cable, and I just haven’t bothered because I didn’t end up needing it.

    Leaving the upgrade path open is a solid choice. The space is so volatile that it’s impossible to predict what the requirements will be like in six months. They could even go down like they did when 4-bit happened.

    I use an external DAC, so can’t speak to the whine there. They’re not that expensive though.


  • If I were you, I’d look into something like the HP Reverb G2 that has inside-out tracking (no external cameras or lighthouses). The immersion you get with VR is way beyond anything you can get from a screen.

    My full setup a Vive Pro 2 with a VKB HOSAS setup and a YawVR 2, and it feels spectacular, though getting interdicted the first time almost made me piss myself.




  • I’m for defederation at least until instance blocking is available. I’ve already blocked most of their communities from my feed, but comments have been really unpleasant since they federated.

    It’s not really about the ideology as much as not wanting to have to scroll past endless political bickering. Rage addiction is real and contagious.



  • I’m having a great time, but I also love FO4 and No Man’s Sky. The toe-dip I’ve done into colony building shows that they put real thought into Astroneer-like automated manufacturing stuff, which is my crack, and something I missed in NMS and FO4. It’s also clear from the first city that they know how depressing FO4 is, and wanted to add more variety.

    Story and characters are a cut above any other Bethesda game so far, but that’s not saying much. My wife is replaying BG3 next to me, and it makes Starfield’s writing look amateurish by comparison. It’s not the core of the game though, so eh.

    Downsides so far have been that the minor planets/moons don’t have much to do, and that inventory management is annoying with how much crafting components weigh.

    Ship combat is… Fine. It’s not as intricate as Elite: Dangerous or SW:Squadrons (for sim gamers, weapons are all on REALLY forgiving gimbals, which makes precision unnecessary), but not actively bad like NMS VR. I think it’s a good compromise, because not everyone wants to deal with a realistic sim in what is essentially a minigame.

    It’s also complex, which is good, but adds some awkwardness to the beginning.






  • There’s a difference between allowing speech about a thing and embracing the thing. This is a classic case of embrace, extend, extinguish.

    If you’re interested, I’d look into what happened with XMPP and Google talk. XMPP was a federated chat service. Google Talk became compatible with it, and instantly became the most popular client for it.

    It then broke compatibility slowly, pushing more people from other XMPP clients onto Google talk.

    They finally removed it completely, and because they were the most popular client, XMPP users moved to Google talk to maintain their connections to other users. The protocol basically ceased to exist.

    People are broadly assuming that’s Meta’s plan with threads and Mastodon, because it’s an extremely common way for corporations to get rid of open systems.