Google is pausing an AI tool that creates images of people following inaccuracies in some historical depictions the model generated, the latest hiccup in the Alphabet-owned company’s efforts to catch up with rivals OpenAI and Microsoft.
Google started offering image generation through its Gemini AI models earlier this month, but over the past few days, some users on social media had flagged that the model returns historical images that are sometimes inaccurate, such as depicting the U.S. Founding Fathers as women or Nazi soldiers as people of colour.
“We’re aware that Gemini is offering inaccuracies in some historical image generation depictions,” Google said on Wednesday.
Since the launch of OpenAI’s ChatGPT in November 2022, Google has been racing to produce AI software rivalling what the Microsoft-backed company had introduced.
When Google released its generative AI chatbot Bard a year ago, the company had shared inaccurate information about pictures of a planet outside the Earth’s solar system in a promotional video, causing shares to slide as much as nine per cent.
Bard was rechristened as Gemini earlier this month and Google rolled out paid subscription plans, which give users access to a more powerful AI model.
“Historical contexts have more nuance to them and we will further tune to accommodate that,” said Jack Krawczyk, senior director of product for Gemini at Google, on Wednesday.