The latest setback in Alphabet-owned Google’s attempts to overtake competitors OpenAI and Microsoft is the company’s decision to halt an AI tool that produces photos of individuals in response to errors in certain historical depictions the model produced.
Earlier this month, Google began to offer image generation through its Gemini AI models. However, in the last few days, some users have noticed on social media that the model returns historical images that are occasionally inaccurate, such as portraying Nazi soldiers as people of color or the U.S. Founding Fathers as women.
Google stated on Wednesday, “We’re aware that Gemini is providing inaccurate depictions in certain historical image generation.”
Following the November 2022 release of OpenAI’s ChatGPT, Google has been vying to create AI software that can compete with this.
