• 58008@lemmy.world
    link
    fedilink
    English
    arrow-up
    5
    ·
    3 months ago

    At least they have an AI-free option, as annoying as it is to have to opt into it.

    On a related note, it’s hilarious to me that the Ecosia search engine has AI built in. Like, I don’t think planting any number of trees is going to offset the damage AI has done and will do to the planet.

      • Balinares@pawb.social
        link
        fedilink
        English
        arrow-up
        1
        ·
        edit-2
        3 months ago

        I mean, the poll was like as not a publicity stunt, to draw attention to the fact DDG is not doing AI. All the same, the fact they are making “no AI” a selling point is noteworthy.

        EDIT: I stand corrected – apparently DDG does do AI presently. Hopefully they’re serious about reconsidering that, then.

    • Egonallanon@feddit.uk
      link
      fedilink
      English
      arrow-up
      1
      ·
      3 months ago

      You can turn all the AI features off on regular DDG search settings. Best I can tell that achuevescthe same as using the no AI filter.

  • setsubyou@lemmy.world
    link
    fedilink
    English
    arrow-up
    2
    ·
    3 months ago

    The article already notes that

    privacy-focused users who don’t want “AI” in their search are more likely to use DuckDuckGo

    But the opposite is also true. Maybe it’s not 90% to 10% elsewhere, but I’d expect the same general imbalance because some people who would answer yes to ai in a survey on a search web site don’t go to search web sites in the first place. They go to ChatGPT or whatever.

        • IronBird@lemmy.world
          link
          fedilink
          English
          arrow-up
          0
          ·
          edit-2
          3 months ago

          it just makes it evermore obvious to them how many people in their life are sheep that believe anything the read online, i assume? a false sense of confidence where one mught have just said 'i dont know"

          • evol@lemmy.today
            link
            fedilink
            English
            arrow-up
            0
            arrow-down
            1
            ·
            edit-2
            3 months ago

            So many people were already using tiktok or youtube as google search. I think AI is arguably better than those

            edit: New business, take your chatgpt question and turn it into a tiktok video. The Slop must go on

            • AmbitiousProcess (they/them)@piefed.social
              link
              fedilink
              English
              arrow-up
              0
              ·
              3 months ago

              The main problem is that LLMs are pulling from those sources too. An LLM often won’t distinguish between highly reputable sources and any random page that has enough relevant keywords, as it’s not actually capable of picking its own sources carefully and analyzing each one’s legitimacy, at least not without a ton of time and computing power that would make it unusable for most quick queries.

              • evol@lemmy.today
                link
                fedilink
                English
                arrow-up
                0
                arrow-down
                1
                ·
                3 months ago

                Genuinely, do you think the average person tiktok’ing their question is getting highly reputable sources? The average American has what, a 7th grade reading level? I think the LLM might have a better idea at this point

        • Ex Nummis@lemmy.world
          link
          fedilink
          English
          arrow-up
          0
          ·
          edit-2
          3 months ago

          First, its results are often simply wrong, so that’s no good. Second, the more people use the AI summaries, the easier it’ll be for the AI companies to subtly influence the results in their advantage. Think of advertising or propaganda.

          This is already happening btw, and the reason Musk created Grokipedia. Grok (and even other llm’s!) already use it as a “trusted source”, which it is anything but.

          • evol@lemmy.today
            link
            fedilink
            English
            arrow-up
            0
            arrow-down
            1
            ·
            3 months ago

            Okay but its a search engine, they can literally just pick websites that align with a certain viewpoint and hide ones that don’t, Its not really a new problem. If they just make grokpedia the first result then its not like not having the AI give you a summary changed anything.

  • Reygle@lemmy.world
    link
    fedilink
    English
    arrow-up
    1
    ·
    edit-2
    3 months ago

    Even that I would consider wildly unjust. User data would HAVE to be opt IN.

    • M137@lemmy.world
      link
      fedilink
      English
      arrow-up
      0
      ·
      3 months ago

      Because the poll just ended… it’s been opt out since before the poll and nothing has changed, yet (if anything does change). How is this not obvious?

        • Jako302@feddit.org
          link
          fedilink
          English
          arrow-up
          0
          arrow-down
          1
          ·
          3 months ago

          Asking an existing userbase for any kind of change will pretty much always result in a no.

          If the project requires minimal resources and doesn’t have a major downside, then implementing your own version before asking is fine.

          They didn’t serve a bunch of ex alcoholics a full bottle of whisky, all they did is make you scroll twice on your mouse wheel.

          • Ensign_Crab@lemmy.world
            link
            fedilink
            English
            arrow-up
            1
            ·
            3 months ago

            Asking an existing userbase for any kind of change will pretty much always result in a no.

            If you’re trying to position yourself as a search engine that hasn’t enshittified, don’t head down that road without asking. Know your userbase. They’re using duckduckgo for a reason.

  • Deestan@lemmy.world
    link
    fedilink
    English
    arrow-up
    1
    ·
    3 months ago

    Meanwhile, at HQ: “The userbase hallucinated that they don’t want AI. Maybe we prompted them wrong?”

    • Sockenklaus@sh.itjust.works
      link
      fedilink
      English
      arrow-up
      0
      ·
      3 months ago

      The prompt was bad: there was no option to vote for “a little bit of AI as a tool is not bad but don’t force feed it to me”.

      I think there were many people who voted for “no AI” who would’ve voted for “a little bit of ai” if they had the option.

      • eksb@programming.dev
        link
        fedilink
        English
        arrow-up
        1
        ·
        3 months ago

        There were probably also people who voted for “yes AI” who would have voted for “a little bit of ai when I explicitly ask for it” if they had the option.

  • Gorilladrums@lemmy.world
    link
    fedilink
    English
    arrow-up
    0
    ·
    3 months ago

    I think most people find something like chatgpt and copilot useful in their day to day lives. LLMs are a very helpful and powerful technology. However, most people are against these models collecting every piece of data imaginable from you. People are against the tech, they’re against the people running the tech.

    I don’t think most people would mind if a FOSS LLM, that’s designed with privacy and complete user control over their data, was integrated with an option to completely opt out. I think that’s the only way to get people to trust this tech again and be onboard.

    • Reygle@lemmy.world
      link
      fedilink
      English
      arrow-up
      1
      ·
      3 months ago

      I’m enjoying how ludicrous the idea of a “privacy friendly AI” is- trained on stolen data from inhaling everyone else’s data from the internet, but cares suddenly about “your” data.

    • GarboDog@lemmy.world
      link
      fedilink
      English
      arrow-up
      0
      ·
      3 months ago

      We can say maybe a personal LLM trained on data that you actually already own and having the infrastructure being self efficient sure but visual generation llms and data theft isn’t cool

  • Azal@pawb.social
    link
    fedilink
    English
    arrow-up
    0
    ·
    3 months ago

    I have found a couple times an LLM being good for searches. My example is scooters in the US, because of commute I’d like to find an electric scooter, the kind like a Vespa. But you do a search, you find all sorts about the standing scooters, look for electric motorcycles and they start getting to car level prices or aren’t street legal, it gets to be a mess. Chat GPT turned that search into a relatively quick one that I could find a local place that I’m going to check out when there’s not ice on the ground. But the important part was making it require links. As a tool, it can have its uses.

    So all of that said… Google and other searches doing AI did fuck all and nothing to help on this, and I agree 100% I do not want AI on the search for DDG

  • thegoodyinthehoody@sh.itjust.works
    link
    fedilink
    English
    arrow-up
    0
    ·
    3 months ago

    As much as I agree with this poll, duck duck go is a very self selecting audience. The number doesn’t actually mean much statistically.

    If the general public knew that “AI” is much closer to predictive text than intelligence they might be more wary of it

    • ikirin@feddit.org
      link
      fedilink
      English
      arrow-up
      1
      ·
      3 months ago

      I mean you Gotta Hand it to “Ai” - it is very sophisticated, and Ressource intensive predicitive Text.