Google pauses Gemini AI image generation of people
Several users complained that Gemini AI was generating inaccurate images of historical figures, prompting Google to pause use of the service.
Google’s new Gemini AI image generation service has earned the company some heated critique from the public, pushing them to pause use of the service for upgrades and improvements. Gemini AI was reportedly receiving complaints for generating blatantly inaccurate images of historical figures and events, which prompted the suspension.
Google announced its decision to pause use of Gemini AI on the Google Communications Twitter this week. The service launched near the end of December 2023, and allows users to put in a prompt, which the AI program would then use to generate an image as close to the prompt as possible. Unfortunately, since then, users have complained of inaccurate images of historical figures and events. One such example was the U.S. Founding Fathers depicted as various ethnicities other than what they were.
Google said in its response to the matter that Gemini is built to offer diverse depictions of people, but it shouldn’t be used in such a controversial manner.
Gemini AI senior director Jack Krawczyk added that making sure historical pictures aren’t fudged is a little harder to train into an AI model and will take time as Google prepares updates for the program. He clarified that general use of Gemini should still prove fruitful.
“We will continue to do this for open ended prompts (images of a person walking a dog are universal!),” Krawczyk wrote on his now-protected Twitter. “Historical contexts have more nuance to them and we will further tune to accommodate that.”
With Gemini AI down until further updates are implemented, stay tuned for new details as Shacknews follows Google and this story.
-
TJ Denzer posted a new article, Google pauses Gemini AI image generation of people