Google’s image generation tool, Gemini, faced a setback this week as the tech giant decided to pause its feature that creates historical images of people. The decision came in response to complaints from users who pointed out that the generated images inaccurately depicted people of different races.
Google Senior Director of Product, Jack Krawczyk, not only acknowledged the issue but also took to social media to explain the team’s efforts in adjusting the AI model for more historically accurate results. However, the fix turned out to be more complex than anticipated, ultimately leading Google to disable the tool’s ability to generate images of people altogether.
Speaking on the issue, Google’s SVP Prabhakar Raghavan explained that the model behind Gemini had inadvertently overcompensated and was too conservative in depicting a range of people. As a result, Google plans to make significant changes and conduct extensive testing before reinstating the feature.
The pause on Gemini’s image generation feature serves as a reminder of the challenges that come with developing AI technologies, particularly in ensuring accurate and respectful representation of diverse individuals. Google’s commitment to addressing the issue and taking the necessary steps to rectify the situation demonstrates a responsible approach to AI development in the tech industry. Stay tuned for updates on when the tool will be back up and running with improved accuracy and inclusivity.
“Introvert. Avid gamer. Wannabe beer advocate. Subtly charming zombie junkie. Social media trailblazer. Web scholar.”