An Asian MIT student asked AI to turn an image of her into a professional headshot. It made her white with lighter skin and blue eyes.::Rona Wang, a 24-year-old MIT student, was experimenting with the AI image creator Playground AI to create a professional LinkedIn photo.

  • ExclamatoryProdundity@lemmy.world
    link
    fedilink
    English
    arrow-up
    86
    arrow-down
    11
    ·
    1 year ago

    Look, I hate racism and inherent bias toward white people but this is just ignorance of the tech. Willfully or otherwise it’s still misleading clickbait. Upload a picture of an anonymous white chick and ask the same thing. It’s going go to make a similar image of another white chick. To get it to reliably recreate your facial features it needs to be trained on your face. It works for celebrities for this reason not a random “Asian MIT student” This kind of shit sets us back and makes us look reactionary.

    • AbouBenAdhem@lemmy.world
      link
      fedilink
      English
      arrow-up
      53
      arrow-down
      6
      ·
      edit-2
      1 year ago

      It’s less a reflection on the tech, and more a reflection on the culture that generated the content that trained the tech.

      Wang told The Globe that she was worried about the consequences in a more serious situation, like if a company used AI to select the most “professional” candidate for the job and it picked white-looking people.

      This is a real potential issue, not just “clickbait”.

      • HumbertTetere@feddit.de
        link
        fedilink
        English
        arrow-up
        13
        ·
        1 year ago

        If companies go pick the most professional applicant by their photo that is a reason for concern, but it has little to do with the image training data of AI.

        • AbouBenAdhem@lemmy.world
          link
          fedilink
          English
          arrow-up
          8
          arrow-down
          1
          ·
          1 year ago

          Some people (especially in business) seem to think that adding AI to a workflow will make obviously bad ideas somehow magically work. Dispelling that notion is why articles like this are important.

          • Water1053@lemmy.world
            link
            fedilink
            English
            arrow-up
            3
            ·
            1 year ago

            Businesses will continue to use bandages rather than fix their root issue. This will always be the case.

            I work in factory automation and almost every camera/vision system we’ve installed has been a bandage of some sort because they think it will magically fix their production issues.

            We’ve had a sales rep ask if our cameras use AI, too. 😵‍💫

          • Square Singer@feddit.de
            link
            fedilink
            English
            arrow-up
            1
            ·
            1 year ago

            It’s a massive issue that many people (especially in business) have this “the AI has spoken”-bias.

            Similar to how they implement whatever the consultant says, no matter if it actually makes sense, they just blindly follow what the AI says .

      • JeffCraig@citizensgaming.com
        link
        fedilink
        English
        arrow-up
        6
        arrow-down
        1
        ·
        edit-2
        1 year ago

        Again, that’s not really the case.

        I have Asian friends that have used these tools and generated headshots that were fine. Just because this one Asian used a model that wasn’t trained for her demographic doesn’t make it a reflection of anything other than the fact that she doesn’t understand how MML models work.

        The worst thing that happened when my friends used it were results with too many fingers or multiple sets of teeth 🤣

      • drz@lemmy.ca
        link
        fedilink
        English
        arrow-up
        2
        arrow-down
        1
        ·
        1 year ago

        No company would use ML to classify who’s the most professional looking candidate.

        1. Anyone with any ML experience at all knows how ridiculous this concept is. Who’s going to go out there and create a dataset matching “proffesional looking scores” to headshots?
        2. The amount of bad press and ridicule this would attract isn’t worth it to any company.
        • kbotc@lemmy.world
          link
          fedilink
          English
          arrow-up
          5
          ·
          1 year ago

          Companies already use resume scanners that have been found to bias against black sounding names. They’re designed to feedback loop successful candidates, and guess what shit the ML learned real quick?

      • ExclamatoryProdundity@lemmy.world
        link
        fedilink
        English
        arrow-up
        6
        arrow-down
        10
        ·
        1 year ago

        Didn’t miss the point. It was trained on images of people. The majority of images it had access to were white faces because that’s what was available to scrape. Too many white people are represented in media. Isn’t that the underlying point? AI is merely reflecting that, as it was designed to do. That reflection is embarrassing. Like a toddler with a potty mouth. Not the kids fault.

        • kadu@lemmy.world
          link
          fedilink
          English
          arrow-up
          13
          arrow-down
          2
          ·
          edit-2
          1 year ago

          AI is merely reflecting that, as it was designed to do.

          Yes. And everybody knows that. Your comment doesn’t really add anything.

          The point is that biased datasets are a problem that should be fixed - AI models need to comprehend human diversity, and the limitations from the biased datasets we do have need to be properly communicated to users.

          • ExclamatoryProdundity@lemmy.world
            link
            fedilink
            English
            arrow-up
            1
            arrow-down
            11
            ·
            1 year ago

            No we’re on the same page. But since you came off hostile from the start, I’m going to assume that’s the kind of interaction you’re looking for. Look elsewhere.

              • ExclamatoryProdundity@lemmy.world
                link
                fedilink
                English
                arrow-up
                1
                arrow-down
                6
                ·
                1 year ago

                I’m allowed to do whatever I want and you’re allowed to be willfully obtuse. Doing great so far. Keep telling the world how it should work rather than try to understand it and offer solutions. It’s gonna work out for you I’m sure of it.

        • Blamemeta@lemmy.world
          link
          fedilink
          English
          arrow-up
          2
          ·
          1 year ago

          Too many white people, meanwhile 2/3rds of America is white. That’s the actual problem, it’s based on American datasets, and finding a balanced dataset is impossible.

    • hardypart@feddit.de
      link
      fedilink
      English
      arrow-up
      7
      arrow-down
      1
      ·
      1 year ago

      It still perfectly and visibly demonstrates the big point of criticism in AI: The tendencies the the training material inhibits.

    • Buddahriffic@lemmy.world
      link
      fedilink
      English
      arrow-up
      2
      ·
      1 year ago

      The AI might associate lighter skin with white person facial structure. That kind of correlation would need to be specifically accounted for I’d think, because even with some examples of lighter skinned Asians, the majority of photos of people with light skin will have white person facial structure.

      Plus it’s becoming more and more apparent that AIs just aren’t that good at what they do in general at this point. Yes, they can produce some pretty interesting things, but they seem to be the exception rather than the norm, and in hindsight, a lot of my being impressed with results I’ve seen so far is that it’s some kind of algorithm that is producing that in the first place when the algorithm itself isn’t directly related to the output but is a few steps back from that.

      I bet for the instances where it does produce good results, it’s still actually doing something simpler than what it looks like it’s doing.

    • notacat@lemmynsfw.com
      link
      fedilink
      English
      arrow-up
      6
      arrow-down
      6
      ·
      1 year ago

      You said yourself you hate inherent bias yet attempt to justify the result by saying if used again it’s just going to produce another white face.

      that’s the problem

      It’s a racial bias baked into these AIs based on their training models.

      • thepineapplejumped@lemm.ee
        link
        fedilink
        English
        arrow-up
        5
        ·
        1 year ago

        I doubt it is concious racial bias, it’s most likely that the training data is made up of mostly white people and labeled poorly.

        • notacat@lemmynsfw.com
          link
          fedilink
          English
          arrow-up
          4
          arrow-down
          2
          ·
          1 year ago

          I also wouldn’t say it was conscious bias either. I don’t think it’s intentionally developed in that way.

          The fact still remains though whether conscious or unconscious, it’s potentially harmful to people of other races. Sure it’s an issue with just graphic generation now. What about when it’s used to identify criminals? When it’s used to filter between potential job candidates?

          The possibilities are virtually endless, but if we don’t start pointing out and addressing any type of bias, it’s only going to get worse.

          • wmassingham@lemmy.world
            link
            fedilink
            English
            arrow-up
            3
            ·
            1 year ago

            What about when it’s used to identify criminals? When it’s used to filter between potential job candidates?

            Simple. It should not fucking be used for those things.

          • Altima NEO
            link
            fedilink
            English
            arrow-up
            2
            ·
            1 year ago

            I feel like you’re overestimating the capabilities of current ai image generation. And also presenting problems that don’t exist.

      • Blaidd@lemmy.world
        link
        fedilink
        English
        arrow-up
        5
        ·
        1 year ago

        They aren’t justifying anything, they literally said it was about the training data.