• potoo22@programming.dev
    link
    fedilink
    English
    arrow-up
    70
    ·
    22 days ago

    No publisher is going to pay a professional to narrate their audiobooks when they can have AI do a shitty job for much less.

    A shitty narrator can get me to hate a book I like. A great narrator can bring the characters to life, enhance the experience, and turn me from a listener to a fan. I’ve searched for books by narrators like Nick Podehl and Jeff Hayes and bought audiobooks I wouldn’t have otherwise.

    • lemonskate@lemmy.world
      link
      fedilink
      English
      arrow-up
      12
      ·
      22 days ago

      I tried, and failed, to get into audio books for years. Then I listened to Dungeon Crawler Carl narrated by Jeff Hayes and what an absolute delight it was. There’s no way I would’ve gotten even 10 minutes in if it was one of those soulless AI voices instead.

    • monkeyman512@lemmy.world
      link
      fedilink
      English
      arrow-up
      11
      ·
      22 days ago

      That depends entirely on how profitable it is and how much they can get authors onboard.

      I do agree that a good narrator delivers a performance that adds the work. James Marster will always be Harry Dresden in my head.

    • 48954246@lemmy.world
      link
      fedilink
      English
      arrow-up
      1
      ·
      21 days ago

      Nick Podehl is such an amazing narrator. The voices and performance are amazing.

      I’ve been slowly getting through the Kel Kade books and the narration just makes it for me

      • utopiah@lemmy.world
        link
        fedilink
        English
        arrow-up
        2
        ·
        21 days ago

        Because… the tool has no understanding of anything? It reads written words, yes, but no intention, no cultural context, no intonation. Unless everything is spelled out like a script, then it will not sound great, would it?

        • Kusimulkku@lemm.ee
          link
          fedilink
          English
          arrow-up
          1
          ·
          edit-2
          21 days ago

          Someone can manually go through it and correct and edit it, as one would a regular, human made recording. It’s not rocket science exactly. It wouldn’t be a story time for children but it would probably be alright for more plain stuff

          • SynopsisTantilize@lemm.ee
            link
            fedilink
            English
            arrow-up
            1
            ·
            21 days ago

            These people just want to hate AI. Read through and see how many times they complain about copywrited material stolen, but claim piracy is the solution.

          • utopiah@lemmy.world
            link
            fedilink
            English
            arrow-up
            1
            ·
            21 days ago

            If the “fix” for an AI implementation in a use case is, again, to manually correct it and find a less demanding audience then… yes, by definition it’s shitty.

            The point isn’t that it’s infeasible, just that it will be low quality.

            • Kusimulkku@lemm.ee
              link
              fedilink
              English
              arrow-up
              1
              ·
              edit-2
              21 days ago

              I mean you have to correct and edit human made stuff too, doesn’t mean it’s shit lol

              If you want the stuff read out and don’t care for the radio type stuff, I’d imagine the better voice AIs do a pretty good job. And I personally prefer the more neutral voices to the story time stuff, so works for me.

              • utopiah@lemmy.world
                link
                fedilink
                English
                arrow-up
                1
                ·
                21 days ago

                This is me just speculating here but if they follow the path of this CEO who fired his human staff to replace it by AI… then rollback admit it’s shit https://gizmodo.com/klarna-hiring-back-human-help-after-going-all-in-on-ai-2000600767 then my bet is that it’s not done to improve quality but rather margins.

                If AI is done alongside professionals, and done so ethically (not stolen training data, not ignoring ecological cost by pumping water in dry areas to cool down GPUs, etc) and economically (i.e. not having it “cheap” now but once a monopoly position is obtain, raise prices for a captive set of consumers) then yes it can be potentially empowering. This though is pretty much never the case.

                That being said, if one “just” want read aloud, there are plenty of FLOSS alternatives and I believe Mozilla even a TTS/STT system based solely on voluntary voices.

                • Kusimulkku@lemm.ee
                  link
                  fedilink
                  English
                  arrow-up
                  1
                  ·
                  21 days ago

                  It’s a company, of course it’s done to increase profits. I’m just saying it being AI doesn’t automatically mean it’s shit, it could be done just fine. AI is a tool, the end result depends on how that tool is used.

                  • utopiah@lemmy.world
                    link
                    fedilink
                    English
                    arrow-up
                    1
                    ·
                    edit-2
                    21 days ago

                    Like I try to highlight, in most cases it’s a shitty tool, doing a bad job, trained on stolen data, requiring a TON of energy and often used to put people out of work (and failing at it, cf news above).

                    So… sure, it’s “just” a tool and in theory, it can be made the right way and used in a good context.

                    It is rarely the case though. Here specifically we are talking about Amazon, a company that has from its inception been built to be a monopoly, relying on AWS a service that is basically destroying the Internet by removing its decentralized nature.

                    So… again even if the tool would in theory itself be used the right way, build the right way, the company using that tool is problematic.

                    TL;DR: in theory, yes, in practice here, no.

    • catloaf@lemm.ee
      link
      fedilink
      English
      arrow-up
      1
      ·
      22 days ago

      For fiction, yeah, that’s true. For nonfiction, this could work pretty well.

      I’m still generally opposed to it because it’s using the work of existing voice recording without compensation, though.

    • ssillyssadass@lemmy.world
      link
      fedilink
      English
      arrow-up
      1
      ·
      21 days ago

      The thing with this is that there won’t be shitty narrations any more. Hate it all you may, fact of the matter is that AI-powered voice generation is pretty good at what it does. So in the future you won’t have shitty narrations and great narrations. You’ll have decent narrations and great (human) narrations.

      • ExperiencedWinter@lemmy.world
        link
        fedilink
        English
        arrow-up
        1
        ·
        21 days ago

        And teslas will have full self driving tomorrow and crypto currency will replace normal currency within one year! Always believe in the hype!