• dylanmorgan@slrpnk.net
      link
      fedilink
      English
      arrow-up
      44
      arrow-down
      1
      ·
      7 months ago

      The problem is that a fuckton of the web is SEO poisoned, so even a better search engine will find garbage because for a lot of subjects garbage is all that’s available.

      The best chef in the world can’t turn shit into anything you want to eat.

      • kronisk @lemmy.world
        link
        fedilink
        English
        arrow-up
        16
        ·
        7 months ago

        This is what they want you to think but they’re hardly even trying. Google is shitty on purpose because if initial search results are bad, you “engage” more and see more ads. And since they’re not worried about competition because google is the default search nearly everywhere – most people don’t even know there are alternatives, google is synonymous with search – they can enshittify their search as much as they want. https://pluralistic.net/2024/04/04/teach-me-how-to-shruggie/ https://www.wheresyoured.at/the-men-who-killed-google/

      • pukeko@lemm.ee
        link
        fedilink
        English
        arrow-up
        2
        ·
        7 months ago

        That’s what I enjoy about kagi: because I can block and rank sources, I get to do some reverse-SEO, and the results are really good with remarkably few adjustments.

      • OpenStars@discuss.online
        link
        fedilink
        English
        arrow-up
        6
        arrow-down
        9
        ·
        7 months ago

        I don’t understand what you are saying. If some other search methodology were to use the same exact methodology of Google in paying attention to the SEO terms, then obviously it would fall prey to the same thing that killed Google. Similarly if the method was not precisely 100% identical yet still used SEO, then it too would be poisoned.

        However, if the search method were to ignore SEO entirely and focus purely on the content of the page, plus other metrics such as number of links to that page, from other highly-ranked websites, but independently of SEO, then it would not be poisoned by SEOs. Although it might suck due to other causes, either related or not.

        Anyway it all depends so heavily on what you want to find - e.g. a replacement for Google Maps is harder, and Google Images is also fairly great too.

          • OpenStars@discuss.online
            link
            fedilink
            English
            arrow-up
            7
            ·
            7 months ago

            Probably less, in their modern formulation. I knew more about past ones but it’s been awhile and I have no idea how much of that is even still relevant.

        • dylanmorgan@slrpnk.net
          link
          fedilink
          English
          arrow-up
          13
          ·
          7 months ago

          A lot of content sites have altered how they write articles to be in line with google SEO to drive traffic. In doing this, the content that can be found by any search engine is now of lower quality.

              • afraid_of_zombies@lemmy.world
                link
                fedilink
                English
                arrow-up
                1
                ·
                7 months ago

                It really isn’t that hard to tell the difference between bots activities and humans. If Facebook can detect a nipple in a picture in microseconds they can tell that “hmm a surprisingly high number of IoT fridges have strong opinions about this anti-Putin blogger, starting last week” isn’t valid.

                • beebarfbadger@lemmy.world
                  link
                  fedilink
                  English
                  arrow-up
                  1
                  ·
                  7 months ago

                  That sounds like the millions of doctored amazon reviews and social media bot-boosted content should be dealt with by next week then.

                  • afraid_of_zombies@lemmy.world
                    link
                    fedilink
                    English
                    arrow-up
                    1
                    ·
                    7 months ago

                    If there was an incentive to do so there would be.

                    Alphabet doesn’t have real competition. If they start getting some they will be motivated to improve. Amazon the same way. You order crap from them and they still make their money. Social media the same way, there is just no particular reason for them to anything about bots when bots don’t impact ad numbers.

                    The corporation I work for has a captcha on the website to do pretty much anything useful. We have an incentive to not have bots.

        • frezik@midwest.social
          link
          fedilink
          English
          arrow-up
          3
          ·
          7 months ago

          What Google used to do is a lot of manual effort to keep SEO stuff at a minimum. You can’t make it go away entirely, but it was a give-and-take that basically worked.

          We’re stuck now because Google decided they didn’t want to spend the money on it anymore, so the SEO people won.

          However, if the search method were to ignore SEO entirely and focus purely on the content of the page

          As another poster mentions, SEO is about gaming the content so search engines pick them up. If you change your algorithm, they’ll just change their methods. Google’s old method may be the only way to clamp down on it.

          • afraid_of_zombies@lemmy.world
            link
            fedilink
            English
            arrow-up
            1
            ·
            7 months ago

            Just block sites that do it or even worse do what the AI people are doing, data harvest it, and present the answer only.

            I stopped looking up recipes ever since chatgpt became a thing.

          • OpenStars@discuss.online
            link
            fedilink
            English
            arrow-up
            1
            ·
            7 months ago

            Admittedly I am much behind on the technicals, but here is an example: if an answer to a technical problem appears on StackOverflow, Reddit, and pleaseclickmePuLEASEpleasepleasepleasepleasepleaseplease.xxx, then why allow the latter ones to rise to the top and the former two don’t show up until like page 3? Regardless of content on the page, the former two sites have a reputable “reputation” - is this what you mean by manual efforts, to designate them as more trustworthy sites?

            Ironically the Reddit upvote/downvote style would work for search results, helping guide others to find similar content after a few people blaze the trail. However, voting has its own issues… as we see even in irl elections, as people game that system too with alt accounts. Anywhere profits are involved, it becomes a cat-and-mouse game where you have to fight off the vested interests.:-(

            But for something important, it becomes worthwhile to invest some effort into it?

            • frezik@midwest.social
              link
              fedilink
              English
              arrow-up
              2
              ·
              7 months ago

              It’s only important in so far as Google can make a profit on it.

              Yes, they could favor specific servers, and that used to be the case. That takes specific effort to pick those sites out, though. They don’t want to do that anymore.

              • OpenStars@discuss.online
                link
                fedilink
                English
                arrow-up
                1
                ·
                7 months ago

                It’s only important in so far as Google can make a profit on it.

                It did not used to be that way. However, we collectively deluded ourselves into thinking that we were “safe”, forever, b/c Google “wasn’t evil”.

                Yes, they could favor specific servers, and that used to be the case. That takes specific effort to pick those sites out, though. They don’t want to do that anymore.

                In a sense, they wouldn’t even have to anymore, if they allowed the old ones to remain at the top. But I see what you mean - e.g. Reddit could change, and Lemmy would never get added.

      • stebo02@sopuli.xyz
        link
        fedilink
        English
        arrow-up
        12
        ·
        7 months ago

        In terms of results, yes. But at least there’s no ads, nor sponsored results, nor bloody AI crap.

        • SpaceNoodle@lemmy.world
          link
          fedilink
          English
          arrow-up
          6
          arrow-down
          2
          ·
          7 months ago

          It’s powered by Bing. That’s all. Compare DDG results with a private Bing session.

          • go $fsck yourself@lemmy.world
            link
            fedilink
            English
            arrow-up
            8
            arrow-down
            3
            ·
            7 months ago

            So, it uses Bing to some extent, I’ll take that. But it’s not entirely Bing.

            Even if a large part of DDG relies on Bing doesn’t mean they are the same. That’s an oversimplification at least.

            • SpaceNoodle@lemmy.world
              link
              fedilink
              English
              arrow-up
              6
              arrow-down
              2
              ·
              7 months ago

              To a great extent. To the point that search results are identical.

              Since the entire topic is about the quality of results, the comparison is quite valid.

              • go $fsck yourself@lemmy.world
                link
                fedilink
                English
                arrow-up
                2
                arrow-down
                5
                ·
                7 months ago

                I’m going to need more than “they look similar”. I don’t think it’s a smart thing to trust random people on the internet based off their word alone.

                Lots of things could seem similar when they are still very different.

                • SpaceNoodle@lemmy.world
                  link
                  fedilink
                  English
                  arrow-up
                  4
                  arrow-down
                  3
                  ·
                  7 months ago

                  I stated that they are identical, not just similar, and literally told you how to check it for yourself. Nothing is stopping you from verifying, or reading their FAQ where they even admit that they use Bing for most of their results.

                  Now stop acting so disingenuous.

                  • go $fsck yourself@lemmy.world
                    link
                    fedilink
                    English
                    arrow-up
                    4
                    arrow-down
                    3
                    ·
                    7 months ago

                    I literally did that and I was not convinced, nor do I think that is good enough evidence. I did not find anything on their site about how much of their results are from Bing.

                    There’s no need to be such as asshole. Why the shit would I take your word when you’re needlessly being such a dick because I simply am asking for more information? Jesus fuck.

          • Negligent_Embassy@links.hackliberty.org
            link
            fedilink
            English
            arrow-up
            12
            ·
            edit-2
            6 months ago

            Apologies if you already know this but there are tons of different instances of searx hosted by different people.

            It’s possible the speed issue is related to the specific instance you tried, do you remember what it was?

            Here is the list of all them: https://searx.space/

            As I type this I realize I’m probably just not sensitive to slowness since I use tor for all my browsing lol

            • Dave@lemmy.nz
              link
              fedilink
              English
              arrow-up
              7
              ·
              7 months ago

              I self-hosted it, and used it for years and liked it. But the results take 3, 4, 5 seconds to come in as you wait for it to run the search on all the different engines. I just tested searx.be above and found the same thing.

              When I use any other search engine, I get results in half a second or less. I know it’s only a few seconds but it’s definitely noticeable when you use a different search engine.

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

              This is precisely why I don’t use searx.
              I just want to quickly find shit, not maintain a list of instances that all have slightly different functionality and frontends.

      • GregorGizeh@lemmy.zip
        link
        fedilink
        English
        arrow-up
        7
        arrow-down
        1
        ·
        7 months ago

        Duckduckgo, searxng, and if you are willing to pay, kagi is far better than google was.

        • Crashumbc@lemmy.world
          link
          fedilink
          English
          arrow-up
          7
          ·
          7 months ago

          “is” absolutely, “was” there was a time in the mid-teens when Google’s search was god-like in its powers of finding what I wanted.

          Duckduckgo is ok but not great even, although search in general is harder nowadays I think

          • GregorGizeh@lemmy.zip
            link
            fedilink
            English
            arrow-up
            2
            ·
            edit-2
            7 months ago

            Yeah i was mostly referring to pre ai enshitification google. Back when all they did was search the web for you and show you a single text ad, those were the days.

            Im also inclined to agree that search in general has gone downhill, but that is more the fault of tech corps locking away info while seo ruins the leftover results.

            I’m worried when ml models really take off the current web dies completely between corpos milking it to death, seo producing nothing but hollow, vapid nonsense to keep you engaged with their non information; and then as final nail in the coffin the ml model ability to entirely fabricate virtually anything digital, truth and information will be completely gone.