• 0 Posts
  • 29 Comments
Joined 3 years ago
cake
Cake day: June 15th, 2023

help-circle

  • I’ve had an antagonistic relationship with a vendor like this, it’s awful. In my case the vendor was supposed to be a fast moving tech startup - the only thing that moved fast there was the revolving door of engineering talent coming and going.

    Even worse, my boss had been convinced by their founder that he had all this pull with the company, and since the company was super cool, that made him super cool, and I dunno if you’ve ever tried to criticize something that has made a middle aged nerd feel cool for the first time in his life, but let’s just say it was not a fruitful endeavor.

    The number of things I effectively fixed for them via email, the abominations I had to construct to work around the things they refused or failed to fix…bad times.


  • I hope you’ll update us if you chase this down. I like 404 Media and I want to keep liking them, but only if the reporting is good. Hopefully it’s a typical tech journalism mistranslation where they use Tesseract OCR to scrape PDFs and the author just misunderstood, or something like that.

    Edit: after looking, I don’t have any issues. Looks like just a raw list from whatever source, I don’t need 404 Media to try to “curate” that or remove elements that seem irrelevant, they can leave that to us.











  • Problems pointed out by commenters aside, I am under the impression that there is very little oversight about this kind of stuff anymore.

    For one thing - unless they’ve changed recently, Amazon “bins” alike products from multiple suppliers, meaning if a bad actor is introducing counterfeits (or just less stringently tested, for more fungible products) - Amazon doesn’t even know who they got them from, by the time that’s discovered.

    But for another thing, the absolutely incredible volume of products - how on earth is anyone making sure these random-character-generated “brands” are safe?

    I lack much in the way of direct evidence, cuz I’ve got shit to do and this isn’t my life’s focus - but it seems apparent that there cannot possibly be the kind of consumer safety testing that we want going on. And if that’s true, it’s only a matter of time before the smart capitalists realize no one is watching and they can make stuff even cheaper (I think they already have), and then how long before we as a society discover all the harm that’s done as a result?

    I’d love to be wrong about this, but like so many tech innovations, I have a feeling we’re going to find out later there were huge harms done before we learned how to rein them in. The speed, volume, and price we’ve grown used to with Amazon seems to preclude consumer safety.




  • Not the person you replied to, but I think you’re both “right”. The ridiculous hype bubble (I’ll call it that for sure) put “AI” everywhere, and most of those are useless gimmicks.

    But there’s also already uses that offer things I’d call novel and useful enough to have some staying power, which also means they’ll be iterated on and improved to whatever degree there is useful stuff there.

    (And just to be clear, an LLM - no matter the use cases and bells and whistles - seems completely incapable of approaching any reasonable definition of AGI, to me)