Google Pauses Gemini People Image Generation Over Historical Inaccuracies
Google temporarily stops its Gemini AI tool from generating images of people after the system produces historically inaccurate depictions, such as racially diverse Nazi soldiers. The company aims to fix the overcorrection of representation bias in its generative AI model.
Google halts the ability of its Gemini AI tool to generate images of people following severe backlash over historically inaccurate depictions. Critics point out that the system produces absurd images, such as portraying World War II-era Nazi soldiers as racially diverse, instead of reflecting historical reality.
The issue stems from Google's attempt to compensate for gender and racial representation biases commonly found in artificial intelligence systems. By aggressively programming the model to ensure diverse outputs, the AI inadvertently creates ahistorical images that ignore the actual demographics of specific historical periods and contexts.
Google acknowledges these problematic generation depictions and pauses the text-to-image feature for people while it works on a fix. The tool, formerly known as Bard, only recently introduced this image generation capability earlier this month before running into this significant public relations problem.