Brin’s “We definitely messed up.”, at an AI “hackathon” event on 2 March, followed a slew of social media posts showing Gemini’s image generation tool depicting a variety of historical figures – including popes, founding fathers of the US and, most excruciatingly, German second world war soldiers – as people of colour.
If you create an image generator that always returns clean cut white men whenever you ask it to produce a “doctor” or a “business man”, but only ever spits out black when when you ask for a picture of someone cleaning, your PR department is going to have a bad time.
And even worse, it actually reinforces that image within users.
Doing the opposite of that isn’t any better
And Gemini was perfectly happy to exclude blacks from prompts about eating fried chicken and watermelon.
Turns out you can’t fight every fire with more fire, more often than not it will burn everything down. You can’t solve something as complex as systemic racism with more systemic racism just against a different group.
Socially, it kind of is, though? When certain groups of people have been historically and chronically maligned, marginalized, persecuted, and othered, showing them in positive roles more frequently is actually a net benefit to those groups, and to society as a whole.
Like, yes, it’s very stupid that these systems are overwriting specific prompts, but also that’s the effect of a white supremacist society refusing to look at itself in the mirror and wrestle with its issues. If you want these big companies to let you use their resources to make specific things that could be used to highlight that white supremacy that people don’t want to acknowledge or address, you kind of have to… get them to acknowledge and address it.
Otherwise, build your own generative model.