OpenAI wants everyone fat and black
(twitter.com)
You're viewing a single comment thread. View all comments, or full comment thread.
Comments (40)
sorted by:
The cope is so funny across all of these. So many defenders saying 'the warm tone from the filter makes everything darker each cycle'
Ok... Why does it change the facial features and weight? It's obvious, in-your-face DEI bias, and they still wont accept it.
If it's a couple, it will race-mix them. If it's a solo, it will black-ify them. That's the pattern, not some complicated excuses.
If that were true, then it should be possible to repeat this experiment with a different image and turn a black person into a white person.
The weight seems more understandable to me than the blackness.
Most people nowadays are fat or morbidly obese. I assume most pictures the AI has seen are fat or morbidly obese.
But why would it make people black? And why would this be true of every single AI, even after they've gotten egg on their faces? It's just bizarre.
The 'warm tone' would make sense if it were just making people darker. But it's not. It's not a dark-skinned white woman. You also get more black facial features.
Because as you guessed it it's not really an AI... or perhaps this is all AI is. What happens is the chatbots are given a set of basic instructions in the form of "here's what you're able to do for the user... you can search the web for related pictures, you can generate images with GPT-image-1, you can do the following edits to user uploaded pictures...", then they have something like an internal API that links to the set of available tools and has instructions on how to call each one, most of them being commands to the latest image generation model. Unless the toolset includes a "literally just copy/paste the picture" tool, it's not going to be able to generate that response. In this case GPT-image-1 interpreted the prompt as the user requesting some kind of inpainting so that's what it did, I guess with hallucination at whatever the default setting is so it pulled style guidance from its biased model.