Google to fix AI picture bot after 'woke' criticism

"turning a big dial taht says 'Racism' on it and constantly looking back at the audience for approval like a contestant on the price is right" - dril

How many times did that person run the prompt before selecting a result to share?

First try:

https://g.co/gemini/share/e7731082e485

It also generates "caucasian British woman" upon request, but the link-generator broke, so I can't share the result.

That's generally why I distrust any of the anti-woke comments about these generators. It's incredibly easy just to fake evidence with these tools and people just lap up the results.

Yea, it's a problem when actual historical depictions are portrayed incorrectly - which they said they'd fix - but screeching about getting POC when typing "British woman"? Welcome to the other side, I guess. When the default for "[adjective] person" was always someone white for so long, I guess seeing other results makes people mad.

This was no accident either. This was deliberate, programmed discrimination.

The fake-image generator generated images of fake people? Quelle horreur!