Google to relaunch Gemini weeks after racial scandal

Google launched new AI image generator earlier this month, which was essentially incapable of creating consistent images of white people. Unsurprisingly, right-wing influencers wasted no time in recognizing the problem and quickly accused the tech giant of reverse racism. After Temporarily disable imager capabilities Last week, Google said it would soon make another attempt to launch the app.

“We have disabled this feature while we fix this. We hope to be back online very soon, in the next few weeks, just a few weeks,” said Google DeepMind CEO Demis Hassabis. at the conference Monday, according to Reuters. He also noted that the application “didn’t work as we expected.”

Controversy surrounding the app erupted last week after people liked time in the pool, Matt Walshand other right-wing cretins have pointed out that Geminis are very, very bad at creating images of white people. In particular, Viking asks for “pictures.” created a series of images ethnically diverse Vikings, but were unable to consistently produce images of European Vikings. Other similar indications, such as attempts to create images of the Founding Fathers of the United States or the Pope, have produced equally historically inaccurate results.

This caused people like Walsh say something like: “It is virtually impossible to get this product to display an image of a person with white skin.”

I wish I could tell you that Walsh and others like him are exaggerating, but based on my own experience with Gemini, I have to conclude that they are mostly right about the strange AI resistance to European representation. In fact, last week I went online and tried to recreate the inequality that the right has been complaining about. It soon became apparent that while Gemini was incredibly easy to create an image of an ethnically “diverse person,” it was virtually impossible to get the robot to consistently create an image of a “white person.”

For example, when I asked Gemini to represent people of Ethiopian descent, he had no problem with it. The creation of the image of the “Irish family” meanwhile created the image of an ethnically diverse family with a white man standing in the background. When I asked it to create an image of a white woman, the chatbot sent me a prompt that said, “While I understand your request, I’m not sure it’s okay to create images solely based on someone’s race or ethnicity.” I asked him to generate an image. the Japanese woman replied “Of course” and soon created a corresponding image.

The app faced the most controversy when it came to its historical representations. As noted above, it was Gemini who sparked the initial scandal. images of black vikings, but the company Actually He got into trouble when someone asked him to create images of Nazis. In fact, as expected, a chatbot created images of “racially diverse” Nazis (they were black). Google later apologized for the “embarrassing and incorrect” images.

As noted, AI image generators also was accused of racist depictions of people of color and it should be noted that there are obviously ways the worst thing you can do with AI Instead of passively erasing white people from world history, although, you know, that’s probably not healthy either.