It turned out that Google’s new artificial intelligence-based tool Gemini does not quite correctly depict some historical figures and groups, such as the founders of the United States or German soldiers from World War II. Users discovered that some of these people were depicted with atypical skin colors. As a result, conspiracy theories and accusations of Google deliberately avoiding depicting white people or changing gender began to circulate on social media.
For example, in response to a request to create an image of a «U.S. senator since the 1800s», Gemini in some cases created results that included black and Native American women. The first female senator was a white woman in 1922, so Gemini’s AI images effectively erased the history of racial and gender discrimination.
Google itself denies the accusations and says that such image generation is caused by the fact that its AI model creates a variety of images of different people, as it is used by people from all over the world. Therefore, in some cases, there may be inaccuracies in historical images.
We're already working to address recent issues with Gemini's image generation feature. While we do this, we're going to pause the image generation of people and will re-release an improved version soon. https://t.co/SLxYPGoqOZ
— Google Communications (@Google_Comms) February 22, 2024
Google has apologized for these erroneous images and said it plans to rectify the situation. In the meantime, the company is suspending the ability to create images of people with Gemini.
«We’re already working to resolve recent issues with Gemini’s image creation feature, Google said in a statement. While we do, we’re going to pause the creation of people images and re-release an improved version soon.
Now, when you try to create an image of a person using Gemini, the AI model provides the following response:
«We are working on improving Gemini’s ability to create images of people. We expect this feature to return soon and will let you know in release updates when it does».