Google admits its Gemini AI was problematic with racially diverse images, vows to fix it

AI SaaS

Google LLC today admitted it “missed the mark” after its the image generator in its generative artificial intelligence Gemini was accused of going too far in trying to promote diversity by consistently making white historical characters nonwhite.

As a result, the company has paused generation of images of people until it can come up with a fix and will release an “improved” version “soon.”

Gemini’s efforts to diversify sparked outrage among many, who called Gemini’s images historically inaccurate. Some examples include the United States’ founding fathers not being white and the 263 Popes in history – all white Europeans – being the wrong shade. It was reported that Vikings, Nazis and even famous Canadian hockey players — men to women — were more often than not, or all the time, misrepresented in terms of race or sex.

Matters were made worse when some users said they asked Gemini to create a white person in history, and the AI was unable to do it, but it could produce an image of a black person seemingly without issue. The controversy ignited even more when a Google employee admitted on X that it is “embarrassingly hard to get Google Gemini to acknowledge that white people exist.”

It seems obvious that Google was trying to make its AI more diverse, but the results perhaps went too far in the other direction. Some have accused the company of trying to wipe white folks from history, although it’s much more likely Google was trying its best to avoid not being accused of being racist, anathema to any high-profile company’s bottom line. Still, Google accepted it might have gone too far, and the company may have been concerned after the “go woke, go broke” adage was invoked by a large number of people online.

“We’re working to improve these kinds of depictions immediately,” Jack Krawczyk, Google’s senior director of product management for Gemini Experiences, said in a statement. “Gemini’s AI image generation does generate a wide range of people. And that’s generally a good thing because people around the world use it. But it’s missing the mark here.”

A fix is in place at the moment. According to The Verge, Gemini is now refusing to create images which might lead to more outrage on the internet. The article said Gemini wouldn’t generate Nazis, Vikings or “an American president from the 1800s.”

AI has had its problems with racial bias in the past, and companies will surely do everything to avoid being thrust into the spotlight in this regard, but in this case, Google’s attempt to remain on the good side of people backfired.

Photo: Pawel Czerwinski/Unplash

Your vote of support is important to us and it helps us keep the content FREE.

One click below supports our mission to provide free, deep, and relevant content.  

Join our community on YouTube

Join the community that includes more than 15,000 #CubeAlumni experts, including Amazon.com CEO Andy Jassy, Dell Technologies founder and CEO Michael Dell, Intel CEO Pat Gelsinger, and many more luminaries and experts.

“TheCUBE is an important partner to the industry. You guys really are a part of our events and we really appreciate you coming and I know people appreciate the content you create as well” – Andy Jassy

THANK YOU

AI SaaS

Leave a Reply

Your email address will not be published. Required fields are marked *