In February, Google’s Gemini AI model received major backlash from users as they realized the chatbot was unable to create accurate racial depictions and images of white people.
The users noticed that it would not create images of white people. When asked to create images of U.S. founding fathers and Catholic popes, it would depict women and men of various races.
Following this incident, the human image generation ability was paused for users.
However, in a blog post released on Wednesday, Google announced that it will bring the feature back for users paying for the English version of the Gemini Advanced. The company also revealed it is bringing the latest image generation model, Imagen 3, into the chatbot.
These updates were implemented to revoke the company’s image, which received a massive hit after giving historically inaccurate results to the users.
It said, “We’ve worked to make technical improvements to the product, as well as improved evaluation sets, red-teaming exercises, and clear product principles.”
Dave Citron, a Google senior director, wrote, “With Imagen 3, we’ve made significant progress in providing a better user experience when generating images of people.”
He said the services won’t support the generation of photorealistic, identifiable individuals, depictions of minors, or excessively gory, violent, or sexual scenes.
The company reminded users that even with significant improvements in the AI model, the technology would have errors.
“Of course, as with any generative AI tool, not every image Gemini creates will be perfect, but we’ll continue to listen to feedback from early users as we keep improving.”
