Google suspends AI image generation of people with Gemini

Google has announced that it is pausing the ability of Gemini, its artificial intelligence system, to generate images of people after the Black Nazis case. The tool, which has attracted interest for its ability to generate different images of the Founding Fathers of the United States and Nazi-era German soldiers, has been shown to have inaccuracies that have led to new problems.

The problems with the images generated by Gemini

Gemini, Google’s artificial intelligence system, has recently been found to generate inaccurate historical images. Specifically, instances have been reported where Gemini generated images of people that did not match the historical characteristics of the requested groups or figures. For example, users requesting images of the Founding Fathers of the United States got results with non-white people, while the first female senator was not elected until 1922.

These inaccuracies have raised online conspiracy theories that Google intentionally avoids depicting white people. In the face of this criticism, Google has decided to pause the generation of images of people via Gemini and work to improve this feature.

Google’s response to diversity issues

In response to the diversity issues that have arisen with Gemini-generated images, Google issued a statement saying that it is already working to address these issues. During this time, the generation of images of people will be paused and an improved version will be released in the future.

Google also stressed that Gemini ‘s goal is to generate diverse images to counter the gender and race stereotypes present in generative artificial intelligence. However, recent errors show that there are still issues to be resolved to ensure historical correctness and accurate representation of people.

Competition with OpenAI and Microsoft Copilot

Google’s introduction of Gemini is part of a fierce competition in the field ofgenerative artificial intelligence. OpenAI and Microsoft Copilot are just some of Google’s competitors that offer similar tools for generating images based on textual input.

These tools leverage advanced algorithms to create collections of images from textual descriptions. However, as Gemini’s recent failures demonstrate, generating realistic and correct images is still a challenge for artificial intelligence.

Next steps for Gemini

After pausing the generation of images of people via Gemini, Google is focusing on improving this feature. The company has assured that it plans to release an improved version soon and notify users about future updates.

Temporarily pausing the generation of people images is a sign that Google is listening to users’ concerns about the accurate representation of historical people. It demonstrates the company’s commitment to ensuring that generative artificial intelligence is used responsibly and respectfully.