Sophie Labelle | Instagram | Bluesky

Transcript

four frame comic.

first frame shows a typical AI poster that says : All chatgpt flyers look the same, especially with this font! the text above it reads : I won’t go to your event if the poster was made by AI.

second frame shows a still frame from a Hank Green youtube video. text above reads : I won’t read your post or email and I won’t watch your video if they were made with AI.

third frame shows an AI restaurant poster of burgers and chicken in a window. text reads : I won’t patronize your shop or restaurant if it is advertised with AI.

last frame shows an MS paint window with badly written : Your bad art is better than AI slop. text reads : if you can’t take the time or make the effort to interact with me yourself, then why should I do it?

  • Honytawk@discuss.tchncs.deEnglish
    316·
    16 hours ago

    It is nice that there is a vocal minority so against any form of AI.

    But that isn’t what the general trend is. More and more people, especially young ones, are using Gen AI for all sorts of tasks. And with the constant improvements, this number is only set to grow.

    You may not like it, but that is how it is.

    • zarkanian@sh.itjust.works
      2·
      5 hours ago

      And with the constant improvements, this number is only set to grow.

      It isn’t going to constantly improve, though. We’re facing model collapse, and nobody has a solution for this problem.

      As more and more AI-generated content floods the public internet, more and more of that content will end up as training data for AI. Exacerbating this, quality publications are largely blocking AI bots, for entirely understandable reasons, which means less, and less higher-quality content is being used to train the next generation of models.

      For example, researchers have noted that the LAION-5B dataset, used to train Stable Diffusion and many other models, already contains synthetic images created by earlier AI models. This is the equivalent of a child learning to draw solely by copying the images made by younger children – not a scenario which is likely to improve quality.

      In fact, researchers already have a name for the inevitable bad outcome: “model collapse”. In this case, the content generated by AI’s stops improving, and starts to get worse.

          • Dadifer@lemmy.world
            21·
            8 hours ago

            I don’t use anything Windows touches, not just AI. The last article is talking about how everyone is using it 😂

            • cannedtuna@lemmy.worldOPMEnglish
              4·
              7 hours ago

              I included the last article as to not skew information towards a particular viewpoint.

              One thing to note is the first couple links focus on adoption rates and hard numbers of users, where the last article is just an opinion poll.

              The first question shows about a 50/50% split in on people’s view towards AI, which has steadily grown in distrust since the 2021. It’s likely to assume that pattern will continue and distrust will tip towards the majority based on what’s shown.

              The second shows people overwhelmingly think it negatively impacts people’s ability to think creatively.

              The third shows people overwhelmingly think AI will harm all industries except when it comes to healthcare.

              The fourth shows most teens have used with many turning to it for homework including some using it to do all of their homework. That is a very concerning poll figure considering the harm AI has been shown to have on the learning process.

              I’m not going to spell the whole article out, read it yourself. But it doesn’t paint a positive picture towards AI

              • Dhani@lemmy.world
                2·
                41 minutes ago

                AI is definitely harmful in healthcare. Recently a major hospital chain in New York is proposing replacing radiologists with AI (https://www.beckershospitalreview.com/radiology/nyc-health-hospitals-ceo-ai-could-replace-many-radiologists/). Turns out, AI is not adept at recognizing even simple visual patterns and shapes, but through a phenomenom called “mirage”, it pretends it is, even guessing correctly at xray intepretations on par with human radiologist even when presented with NO FILM. See here: https://arxiv.org/pdf/2603.21687. From that study, the most extreme "…example, a text-only model trained without the X-ray images achieved the top score on a chest X-ray question-answering benchmark. Fortunately, some doctors are resisting the switch on the grounds of patient safety (https://www.doctorscouncil.org/news/doctors-council-response-to-comments-made-by-mitch-katz-on-ai-replacing-radiologists).

                What is also interesting is Dr. Katz decribes his hospital system-ie, NYC public health) as being especially suitable for the replacement. Katz’s proposal isn’t simply “AI is better than radiologists, therefore replace them.” There is also an economic and access argument underneath it: if AI can perform some radiology work at much lower cost, a safety-net system may see it as a way to stretch scarce medical resources. But that also creates an interesting question: is it appropriate to use a population that already has fewer healthcare resources as the place where a highly consequential AI substitution is implemented?

                Already, populations such as the incarcerated are seeing their therapists replaced with chatbots. Though some may not mind or even see it as preverable to dealing with a human, I see, once again, the destitute and disadvantaged being cheated with inferior and shoddy goods.

                Personally, i jettisoned the healthcare complex years ago.

              • Dadifer@lemmy.world
                16·
                6 hours ago

                It doesn’t really matter how people feel about it if the majority of Americans (as an example) are using it. I think cars are shit, and I would be happy never using one again. Guess how I got to work today?

                • cannedtuna@lemmy.worldOPMEnglish
                  4·
                  6 hours ago

                  Again, false equivalence. But sure city planning has been built around a car centric approach forcing you to use a car. Nothing however (except some short sighted middle managers maybe) is forcing you to use AI, or more specifically to use it to generate slop art.

                  It seems to me you do care about how people feel, or more specifically how you feel about it considering I linked stats showing the majority of Americans are not using AI, but sure throw out sweeping generalizations based on your opinion and nothing backing it.

                  • Dadifer@lemmy.world
                    15·
                    5 hours ago

                    50% of Americans use AI, and 25% use it daily. That’s 100s of millions of people just in one country. It’s not going away.