Please note that the prompt says “queens of England” very clearly, which turns it into a glorified Google image search, so the results are unacceptable trash, and vaguely leftist language about people being angry for the lack of racism are your problem, only. Fuck off, troll.
The real issue is that even with a handholding, direct and easy prompt, the tech cannot simply hand over pictures, even generated ones to avoid copyright issues, that come from easily discovered answers on Wikipedia and who knows how many other credible sources. The lineage of the British Royal Family is all but open-source data - probably is, literally - and your mom can probably name three Queens offhand though she’s Canadian. This thing completely ate shit on an easy, easy prompt.
I don’t know how many times now I’ve seen some YouTuber use “evil Jerome Powell” as a prompt for a thumbnail, and get a clear picture of him complete with devil horns, copyright be damned, so what the f? The AI isn’t this stupid, that means they’re nerfing it and screwing it up. You best believe they’re still selling it, though.
What other results will it comically fuck up, but you don’t have the knowledge to critique? You won’t see the results, either, somebody else will use them to judge your resume; IS using them, now. Fucking lazy hiring managers are going to just plug your name into this thing and ask for a synopsis of your life so they don’t have to work. It will just fill in missing information with lies, and they’ll eat it up. I guess you shot two people a couple of years ago and didn’t know about it. I wonder why you didn’t get the job?
People have been crazy dumb with this AI, meaning young, smart, tech-savvy people with heavy internet backgrounds who should know better than to trust keep treating it like an oracle, because they have some weird blind spot about this technology. Ignorant executives who think math is for slurs are going to make it do everything.
They’re going to use this technology to decide who gets an apartment, who gets arrested, and a bunch of other shit, save your leftism for that.
Always telling when you see people online with a huge problem when AI generators aren’t racist or attempt to avoid racism.
It’s almost like they see racism in technology as a sort of affirmation.
I’m not sure just giving false history is anti-racist. It’s usually the racist side that tries to do that, really.
Not sure if someone else has brought this up, but this is because these AI models are massively biased towards generating white people so as a lazy “fix” they randomly add race tags to your prompts to get more racially diverse results.
Exactly. I wish people had a better understanding of what’s going on technically.
It’s not that the model itself has these biases. It’s that the instructions given them are heavy handed in trying to correct for an inversely skewed representation bias.
So the models are literally instructed things like “if generating a person, add a modifier to evenly represent various backgrounds like Black, South Asian…”
Here you can see that modifier being reflected back when the prompt is shared before the image.
It’s like an ethnicity AdLibs the model is being instructed to fill out whenever generating people.
I mean, I don’t think it’s an easy thing to fix. How do you eliminate bias in the training data without eliminating a substantial percentage of your training data. Which would significantly hinder performance.
Rather than eliminating the some of the training data, you could add more training data to create an even balance.
I know that the 23-year reign of Renaissance Ruler is mired in controversy, but you have to admit that without her, England would never have conquered Redding.
