• sinematic@lemmy.zip
    link
    fedilink
    English
    arrow-up
    21
    ·
    6 days ago

    Absolute pricks. “Don’t do evil” they said.

    AI has biases. News are titled to be biased too. This is grounds for fake news.

    • Jiral@lemmy.org
      link
      fedilink
      English
      arrow-up
      3
      ·
      6 days ago

      Don’t do evil

      Hasn’t that been Google’s guiding principle for quite some time already?

      • ripcord@lemmy.world
        link
        fedilink
        English
        arrow-up
        1
        ·
        6 days ago

        “Don’t be evil” hasn’t been an official guiding principle for over a decade, no.

          • ripcord@lemmy.world
            link
            fedilink
            English
            arrow-up
            1
            ·
            6 days ago

            Ah, I was waking up and didn’t see the strike through

            Both here and on reddit you can just say “whoosh” although it wouldn’t really have totally made sense

    • sigmabot@lemmy.zip
      link
      fedilink
      English
      arrow-up
      1
      arrow-down
      4
      ·
      edit-2
      6 days ago

      I’d trust an LLM to summarize an article and give it an honest title over a piece of shit journalist that wrote it.

      • Lyrl@lemmy.dbzer0.com
        link
        fedilink
        English
        arrow-up
        1
        ·
        edit-2
        5 days ago

        In the short-lived news app Artifact, that was one of my favorite features. It was done on demand, and if a high portion of early viewers asked for a rewritten title, the rewrite would become the default for future viewserves.

        In the Artifact implementation, the LLM was specifically prompted by the app to summarize the article with an honest, non-clickbaity title. In Google’s case, they claim they are prompting the LLM to title the link to better tempt the searcher to click on it based on what they were searching for. Kind of the opposite. Yes, LLMs could do what you say, but that doesn’t seem to be how Google is setting it up.

  • imjustmsk@lemmy.world
    link
    fedilink
    English
    arrow-up
    56
    ·
    7 days ago

    What the fuck are we supposed to do when the services we use are monopolies which actively seeks to destroy its competition as well as enshittify itself to a point its unusable :|

    Time to start self hosting.

    • JensSpahnpasta@feddit.org
      link
      fedilink
      English
      arrow-up
      12
      ·
      7 days ago

      Time to start self hosting

      There is simply no usable self-hosted search engine. There are some projects like Yacy or Searxng, but they do not have their own index. And having your own self-hosted index is kind of impossible because you need to save the whole of the web on your own devices to be able to search it.

      • yourgodlucifer@lemmy.world
        link
        fedilink
        English
        arrow-up
        3
        ·
        6 days ago

        Yacy does have its own index it even allows you to crawl the webpages yourself from the settings page problem is that the results you get out of it fucking suck

        I really like the concept of yacy just wish the results were better but I think there is value in p2p search indexes yacy is probably the most reasonable search engine to self host since you don’t have to index everything yourself

        I have had some success with https://marginalia-search.com/

        Which does have its own index it is open source so you could theoretically self host but you would need to make your own index

        • JensSpahnpasta@feddit.org
          link
          fedilink
          English
          arrow-up
          2
          ·
          6 days ago

          Yeah, that’s a problem. So even if you want to search your local newspaper, you won’t be able to because of bot protection. Your crawler will be blocked. You also can’t put the whole of Reddit or even Lemmy in your search index.

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

            I do think this is one of the places where yacy has it’s strength since yacy is p2p you can retrieve results from other nodes that have indexed the page thus large websites can be collectively indexed by many nodes.

            I also don’t remember having problems with the crawler getting blocked but I also didn’t do a lot of crawls I can see this being a problem though.

        • athatet@lemmy.zip
          link
          fedilink
          English
          arrow-up
          9
          arrow-down
          1
          ·
          7 days ago

          I mean get a group of friends together and start pillaging Target stores.

    • nialv7@lemmy.world
      link
      fedilink
      English
      arrow-up
      3
      ·
      7 days ago

      ten year ago there were still things we could have done to stop google. now is too late…

      • badgermurphy@lemmy.world
        link
        fedilink
        English
        arrow-up
        3
        ·
        6 days ago

        Try not to give up hope! People said similar things about IBM, Yahoo!, AltaVista, AOL, Blockbuster Video, Standard Oil, The Dutch East India Company, and more! All of those are either in the dustbin of history or ghosts of their former selves.

        The reckoning will come to these companies that continue to seem successful in spite of providing objectively bad and worsening products; nothing has ever stopped the pendulum from swinging. When you see your chance to help, give it a push.

        • nialv7@lemmy.world
          link
          fedilink
          English
          arrow-up
          3
          ·
          6 days ago

          These entities may have gone but their ghosts never left. They only got up ended because of paradigm shifts, not because the system that allowed them to exist got fixed. The new stuff that replaced the old ones eventually goes on the same path. I am not hopeful anything ever will happen.

          • badgermurphy@lemmy.world
            link
            fedilink
            English
            arrow-up
            3
            ·
            edit-2
            6 days ago

            While that’s true for some of those, you never know when there will be a paradigm shift, and neither do they. Also, off the top of my head, I know that Yahoo! and IBM caused their own undoing through long periods of mismanagement. The world was in their hands and they couldn’t stay out of their own way. Standard Oil was broken up in direct response to the establishment and enforcement of federal anti-monopoly regulation.

            So, again, don’t give up hope! If the pendulum does not swing back the other way, it will the defy the sum of all human history. If you think about it, believing otherwise doesn’t even make sense, like believing if I keep throwing a ball on the air, eventually it will stay up there.

    • kazerniel@lemmy.world
      link
      fedilink
      English
      arrow-up
      3
      ·
      6 days ago

      I may switch to them as my primary provider. They seem to source results from Bing, and with the noai prefix hopefully won’t keep turning on AI summary like Ecosia does 😑

  • Bieren@lemmy.today
    link
    fedilink
    English
    arrow-up
    14
    ·
    6 days ago

    Before long they will use it to just create their own sites. Their own articles. Why replace just the headline. A whole internet of just google ai generated bs. Just search something on Google and everything from there on is ai generated. All results. All pictures. The maps. Everything. Nothing will be real.

  • LogicalDrivel@sopuli.xyz
    link
    fedilink
    English
    arrow-up
    30
    ·
    7 days ago

    I tried looking for an instruction PDF for some stage lights the other day. No amount of “google-fu” could find the right thing. Links labeled as one model showed stuff for other models when clicked, and the god awful summary that it shoves in your face was just full of contradictions of itself. Google is warm trash water at this point.

    • M0oP0o@mander.xyz
      link
      fedilink
      English
      arrow-up
      8
      ·
      6 days ago

      I find the same thing with vehicle years, I look for say a fuse box layout for a 1985 c20 truck and I will get nothing but genaric trash or stuff for 2021 model years. I used to be able to use “year” but even that has stopped working. Like how do you fuck model and year up?

  • 404found@lemmy.zip
    link
    fedilink
    English
    arrow-up
    10
    ·
    6 days ago

    What Google is accomplishing with this change is tighter control of their results.

    This is disappointing and only serves the interest of Google.

  • NewNewAugustEast@lemmy.zip
    link
    fedilink
    English
    arrow-up
    21
    arrow-down
    1
    ·
    6 days ago

    Wow. Thats horrible. Google has been a shitty search engine for a while now, but this is even worse.

    I quit using them as my main search probably 5 years ago. But I get that most people still use it and most people will be misled.

    Awful.

  • Fit_Series_573@lemmy.world
    link
    fedilink
    English
    arrow-up
    20
    ·
    6 days ago

    Apparently this has been going on for a while but The Verge started sounding the alarm 3 months ago. They know most people won’t read more than a couple sentences past a headline if they even open the article at all. Confirmation bias is about to go wild

    • BenchpressMuyDebil@szmer.info
      link
      fedilink
      English
      arrow-up
      1
      ·
      6 days ago

      You can hide the AI preview using udm14.com or adding &udm=14 query param to your google queries (can be done automatically by adding browser search engine with this param)

    • cley_faye@lemmy.world
      link
      fedilink
      English
      arrow-up
      1
      ·
      6 days ago

      You can use the web page, aka udm14. You can also set this in your browser so that typing in the search/address bar directly opens this.

      I have no idea if this particular development affects these results, but so far it’s been nice. No AI summary, no “similar” advertisement, no “questions about…”. Just plain results. Like, I don’t know, that old google website.

  • I_Has_A_Hat@lemmy.world
    link
    fedilink
    English
    arrow-up
    6
    arrow-down
    1
    ·
    7 days ago

    Ok wait… Can I use AI to change clickbait headlines into actual summaries of articles, without hyperbole, or people getting “slammed”? I am sure googles goal is only to create more clickbait, but if there was a way to do the opposite, that would be the first good use of AI I’ve seen.

    • Deestan@lemmy.world
      link
      fedilink
      English
      arrow-up
      6
      ·
      7 days ago

      Yeah, friend of mine did that as a hobby project 8ish years ago. You can do that with just a small self-trained language model.

      So, with AI as in the current RAM-sucking, earth-boiling monstrosities that are being pushed? No, it’s not a valid use case for that AI because it was already solved.

      • Ebber@lemmings.world
        link
        fedilink
        English
        arrow-up
        3
        ·
        7 days ago

        There was a story a few years back where a Danish guy made an extension to “Unbait” headlines on the particularly egregious news sites here. Of course they complained, i believe on grounds of copyright infringement, and I think they sued the guy.