• @hendrik@lemmy.ml
    link
    fedilink
    105
    edit-2
    9 months ago

    For what it’s worth… I think there are useful AI tools. For example the offline translation feature that doesn’t send your content to google is something they recently introduced. I’d also like to see someone compete with a decent and open text-to-speech solution that gets wide adoption… And the idea of flagging fake reviews doesn’t sound too bad (I haven’t tried it.) I mean people are complaining about SEO making google unusable and fake news only ever getting more. I can see some benefit there - if done right.

    But we definitely don’t need a Clippy 2.0 or another smart assistant. And I don’t think everything has to be embedded in a browser and make it yet more complicated and bigger, or implemented in the operating system. An add-on will probably do.

    (Edit: And I sometimes don’t understand Mozilla. Why not focus on their core product and make that exceptionally great? If they’re already struggling… What’s with all these side-projects and dabbling in AI anyways?)

    • m-p{3}
      link
      fedilink
      29 months ago

      One feature that could be neat is having a locally-generated summary of a page, as well as suggested tags when bookmarking.

      • @rufus@discuss.tchncs.de
        link
        fedilink
        4
        edit-2
        9 months ago

        Uh yeah, I’m not sure. I’ve tried summarizing with AI tools. And there is the bot here on Lemmy that summarizes stuff… I never liked any of that. It’s really a mixed bag, from pretty okay summaries to entirely missing the point of the original article to bordering on false information. I think we’re far from there yet. However, it’s a common use-case for AI. Maybe in 1-2 years I can stop being afraid of misinformation being fed to me. Currently, I think the incorrectness of the information still outweighs any potential benefit. The more complicated it gets, thus making you in need of a summary in the first place, the more biased and skewed the results get. So I don’t see that happen in the very near future. But we definitely should keep up doing the research and pushing that.

        Tagging and organizing is something I’d like an AI for.

        • pizzaboi
          link
          fedilink
          English
          29 months ago

          Imagine spending hours writing and editing something with care only for an LLM to “summarize“ it, completely missing any nuance or sarcasm, removing any creative bits or humor, while also making the wrong point altogether. To top it off anyone unwilling to read your story, their time is valuable after all (but not yours, apparently), will now repeat the LLM’s interpretation to anyone they’d like, whether it’s accurate or not.

          It’s an abysmal direction to go for misinformation and even more abysmal for writers. Good content becomes irrelevant and people become less and less willing to pay for a writer’s time and expertise. Why not write with an LLM if a large percentage of your readers summarize the piece with an LLM anyways? Just need more eyeballs to justify our Google Ads spending.

          Built into a “private” browser or not, it’s just another nail in the coffin of a web built by and for humans.

          • @rufus@discuss.tchncs.de
            link
            fedilink
            39 months ago

            I think you’re completely right with that assessment. Journalist used to be a reputable profession. And explaining things and processing raw information into something that can be consumed by the reader, deemed important. Especially getting it right. There is a whole process to it if you do it professionally. And curating content and deciding what is significant and gets an audience is equally as important.

            Doing away with all of that is like replacing your New York Times with your 5-year-old and whatever she took from watching the news.