I mean just throw in a bunch of stuff you already liked, and then ask it to find similar stuff… Or just ask for a genre of stuff like “Space themed show with a lot of politics” (Assuming you didn’t already know that, then voila, it spits out: Expanse)

If it fucks up and give you something totally not what you were looking for, there’s not much harm besides a small bit of your time. Its fictional media anyways, just don’t go looking for news or medical advice lol, avoid treating it as a truth machine and its not that bad…

Okay sure you can hate the corporate that is gonna do evil shit with these tech, but the tech itself still kinda useful… (maybe we’ll get some very advanced and ethically made FOSS “AI”/LLM in the future, who knows)

    • endless_nameless@lemmy.world
      link
      fedilink
      English
      arrow-up
      1
      arrow-down
      8
      ·
      2 days ago

      Are wildly speculative “what its” the best you can do to disparage this technology/industry? Secretly advertising without disclosing it would probably open them up to a fuckton of liability. I’m not a lawyer but that seems to be the case in pretty much every context.

      Surely there is something real you can point to instead. AI already fucking sucks we don’t need to make stuff up when there are real things to point to.

  • nutbutter@discuss.tchncs.de
    link
    fedilink
    English
    arrow-up
    5
    ·
    2 days ago

    I have tried it, doesn’t really work for me. I have tried online LLMs and local LLMs using Camofox etc. All I get are names of popular media, which I can usually find with a simple search on duckduckgo. In my case LLMs do not provide much help. Directly searching on pirated media sites like YTS, 1337x, or even using TMDB helps a lot more.

    LLMs may perform better if they had direct API access of TMDB or something, like a skill on Hermes agent or open claw, that could also be better. But it still takes a lot more time and resources than just looking up stuff yourself. Also, searching yourself might show a thing or two more that you weren’t even looking for.

    Upvote for the actual unpopular opinion, though.

    • cronenthal@discuss.tchncs.de
      link
      fedilink
      English
      arrow-up
      2
      ·
      2 days ago

      This: just plain old “search” often yields better results. The information is out there and it’s not hard to find.

  • gurty@lemmy.world
    link
    fedilink
    English
    arrow-up
    2
    ·
    1 day ago

    Basically a search engine with a chatbot-face, which isn’t terrible but certainly isn’t the gamechanger the AI industry likes to imply.

  • hendrik@palaver.p3x.de
    link
    fedilink
    English
    arrow-up
    1
    ·
    edit-2
    1 day ago

    I do that as well. Like tell it I have fetish for shag piles and space ships. And it’ll recommend the movie Barbarella to me 😅

    And one time, after reading how people publish AI written books, I tried to write a few short stories with AI myself. And at some point, I’d often end with the feeling “wait a minute, I know that movie”. And after querying it about “similar” movies and TV shows, it’d tell me what it just ripped off. And recommend some more. Nothing fancy, I could have googled that. But nevertheless, I learned about some nice TV shows that way.

    But fact-check the output. It also hallucinated some. And recommended some unwatchable tearjerker movies to me.

    And nothing beats human word of mouth. If you have friends who get to learn you didn’t watch The Expanse, or The Orville… they’ll proactively recommend that to you and it’ll make your day. Or week. And no amount of AI can do that. Same goes for your next favorite niche Italian metal band and their music.

  • Iconoclast@feddit.uk
    link
    fedilink
    English
    arrow-up
    2
    arrow-down
    1
    ·
    2 days ago

    There’s plenty of good use cases for LLMs. The most devout haters simply refuse to acknowledge that because giving even an inch suggests they might need to give another, and it starts to chip away at their binary worldview where things can be neatly labeled good or bad with very little nuance in between.

    As is with most technology, it has its limits. It can be used for both good and bad. Some people find them genuinely useful tools and they’ve helped a ton of people, but the opposite is also true. There’s also valid criticism about the energy demands, environmental damage, and so on. It’s a messy subject that can be approached from many perspectives. Our current online environments simply don’t reward nuance, so you only hear from the extremes.

  • newton@feddit.online
    link
    fedilink
    English
    arrow-up
    1
    arrow-down
    1
    ·
    1 day ago

    Do your homework, and search the movie databases yourself, so you come across unexpected beautiful movies.

    But yes, in 2026 everything has to be served right now, on a plate in front of you.

    I watched movies from the early 1900s to the present, and some even before they hit the cinema