• 14 Posts
  • 624 Comments
Joined 2 years ago
cake
Cake day: June 14th, 2023

help-circle
  • The Internet boom didn’t have the weird you’re-holding-it-wrong vibe too. Legit “It doesn’t help with my use case concerns” seem to all too often get answered with choruses of “but have you tried this week’s model? Have you spent enough time trying to play with it and tweak it to get something more like you want?” Don’t admit limits to the tech, just keep hitting the gacha.

    I’ve had people say I’m not approaching AI in “good faith”. I say that you didn’t need “good faith” to see that Lotus 1-2-3 was more flexible and faster than tallying up inventory on paper, or that AltaVista was faster than browsing a card catalog.


  • I have to think that most people won’t want to do local training.

    It’s like Gentoo Linux. Yeah, you can compile everything with the exact optimal set of options for your kit, but at huge inefficiency when most use cases might be mostly served by two or three pre built options.

    If you’re just running pre-made models, plenty of them will run on a 6900XT or whatever.



  • It’s a quick way to say that the emperor still has no clothes.

    We’re boiling the oceans to train the models, and these are well-publicised failure modes. If they haven’t fixed it, it seems to suggest they CAN’T fix it with the tools and architecture they have. So what other problems is it whiffing on that aren’t trivially checkable?

    If marketing boxed in the product and said “it does these ten things well”, we might be willing to forgive limitations when we leave its wheelhouse. Nobody kvetches that Microsoft Word is an awful IDE, after all. But that would require a retreat from a public that’s been promised Lt. Cmdr. Data in your pocket, and investors that have priced it as such.









  • It’s being shoved at us.

    Most new tech starts with a narrow legit use csse or an enthusiast culture and gradually works to a breakout moment where everyone wants it. Think of cars in 1900 vs 1925 or home computers in 1976 vs 1999. Also note plenty of new tech fails to go mainstream no matter how much effort went into it. 3D TVs, turbine locomotives, non-photovoltaic solar: they tried but didn’t really make it.

    Capital has decided AI will be the next thing and they want it now, so they refuse to let the process run. They can’t wait for a product that solves real faults with the current designs (inefficiency, hallucinations) or does something people actually want (nobody asked for extra fingers) before stuffing it in everything.




  • TBH, I could see non-Israeli Jews getting worried that the longer the horror show continues in Gaza, the more likely it will be used to demonize them.

    This isn’t just a moral stand-- it’s sensible self-preservation.

    Personally, I hate that “support Israel” has been turned into “issue a blank cheque for genocide.” Real “support” includes recognizing when a terrible mistake with long term consequences is happening and trying to pull them out of it. Cutting off arms to Israel is the nation-state version of taking your buddy’s car keys at the pub so he can’t drive drunk. They’ll scream and curse your name, but they’ll get home alive.



  • I ordered a large keyboard enclosure from JLCPCB’s 3D-printing division recently. The tarriffs were like $48 on top of $45 postage and a $80 actual-goods price.

    When I fed the job into Craftcloud (probably not the cheapest but a quick way to read the market) trying to get a US-based supplier would have been like $800.

    They can’t tarrif these industries back on shore. At least not in any sort of useful timescale.

    But the most frustrating part is just the ever-changing aspect. If they said it was a specific amount eith a clear timetable, merchants could at least build prepayment and accurate prices into their checkout flows. Now there’s the risk that whatever amount you paid 2 weeks ago is wrong, and the couriers seem to be responsible for collection, who love to turn that into an excuse to add penalty fees and hold parcels hostage.



  • At some point they broke the compact. You come, you get a $30-per-night hotel and a $8 steak dinner because the rest of the money is going into the machines/tables. That’s why so many of the attractions used to be the gawkable buildings and public shows-- you could still enjoy them if you had blown your budget.

    I guess they pivoted away, but to what? There are whales who want a $5000-per-night suite but you can’t fill an entire 30-story building with them (especially when there are 50 such buildings within walking distance all chasing them)

    I went in May and even cheap meals were over $10, the low-mid priced Fremont Street hotel was around a hundred bucks a night, and the one show I went to was 1/3 full probably because it was $75 for an act that’s been running for decades. I budgeted $1000 to gamble but ended up only dropping 350 because it felt like it wasn’t much I couldn’t see in the local Native-reservation casino.

    I will say nothing but good things for the Pinball Hall of Fame though.