Betul Abari | Anadolu | Getty Images
Google said Thursday it would suspend Gemini’s artificial intelligence-powered image generation feature, citing “inaccuracies” in historical photos.
Users on social media complained that the AI tool inaccurately generated images of historical figures such as the Founding Fathers of the United States as people of color.
Google said in a post on X on Wednesday that its AI capabilities “could create a wide range of people. People all over the world are using it, so that’s generally a good thing.” But the company said the software’s functionality “misses the point here,” adding that the tech giant is “working to immediately improve this kind of depiction.”
Google released its latest statement on Thursday, saying it was suspending Gemini’s human image generation feature and would soon re-release an “improved” version.
The image generation tool was released in early February through Gemini, formerly known as Bard. The company faces challenges now as Google tries to catch up with Microsoft-backed OpenAI.
Gemini refused to produce any images when inspected by a CNBC reporter Thursday morning.
While Google is facing issues with Gemini’s image creation, Open AI last week announced Sora, a new generative AI model that can generate videos from a user’s text prompts.
Google’s Gemini senior director of products, Jack Kracik, said Wednesday that the company’s image generation features reflect its “global user base” and that the company takes “representation and bias” seriously. Stated.
“We will continue to do this for open-ended prompts (images of people walking their dogs are universal!),” Krawczyk said in a post on X.
“There are more nuances in the historical context, so we’ll make further adjustments to accommodate that.”
Google launched its most advanced AI model, Gemini, at the end of 2023. This month, the company also rebranded ChatGPT competitor Bard to Gemini. Google also launched a subscription service to more powerful versions of its AI models.
