Google cuts out Gemini’s features to generate AI images of people after racial errors
Tech magnet, Google has temporarily stopped its Gemini AI model from generating images after a week-long backlash because of racial errors by the model.
In a statement, Google apologised for saying that the model was “missing the mark” after generating specific white figures (like the US founding fathers) or the Nazi-era German soldiers as people of colour. Pre-historic data suggest that these people were mostly white, but the model generated an overcorrection to long-standing racial bias problems by making them people of colour.
A lot of X users were complaining on the social media platform. It almost turned out to be a widespread mockery and outrage against Gemini-generated images that circulated on social media. One X user said, “It is embarrassingly hard to get Google’s Gemini to know that white people exist”
In a clarification post on X, Google said, “Gemini’s AI image generation does generate a wide range of people. And that’s generally a good thing because people worldwide use it. But it’s missing the mark here.”’
The tech magnate added, “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.”
One of the images that stood out for criticism was a depiction of the scene of black and Asian Nazi soldiers. Another one depicted four Swedish women to be people of color.
Over the years, we have seen AI models experience widespread backlash for failing to notice the disparity between people of color and white people, leading to stereotypes in their results.
With OpenAI posing fierce competition for Google, the tech giant has been racing to catch up with ChatGPT since 2023, and their race has suffered setbacks in the release of its AI models.
Just last year, Google apologised after Bard, its AI chatbot wrongly presented a demo that the James Webb Space Telescope took from outside the solar system. It is evident that Google needs to improve the performance of its AI models, hopefully, the new versions will be re-released into the market soon.