- Google’s CEO Sundar Pichai expresses strong condemnation of Gemini AI’s diversity errors, labeling them as “completely unacceptable.”
- Gemini AI generated historically inaccurate images, including racially diverse Nazi-era German soldiers and misrepresentations of US Founding Fathers, sparking widespread backlash.
- Pichai’s internal memo to employees acknowledges the ongoing efforts to address the problematic text and image responses generated by Gemini AI, highlighting Google’s commitment to continuous improvement despite the challenges of AI development.
Google’s CEO, Sundar Pichai, conveyed strong disapproval regarding the errors in diversity displayed by the Gemini AI, labeling them as “completely unacceptable.” In an internal memo obtained by The Verge, Pichai expressed that the inaccurate images and text produced by Gemini AI had not only offended users but also exhibited bias.
Last week, Google halted Gemini’s image generation feature after widespread discovery of racially inaccurate depictions, including Nazi-era German soldiers depicted as racially diverse, non-white US Founding Fathers, and misrepresentations of Google’s own co-founders. Although Google issued an apology for the oversight and pledged to reinstate image generation in the following weeks, Pichai’s memo marks the first widespread acknowledgment of the controversy by the CEO.
In the memo, initially reported by Semafor, Pichai noted that the company had been diligently working to rectify “problematic text and image responses” within the Gemini app, without explicitly stating that the issue had been resolved. He acknowledged the imperfection of AI, particularly at this nascent stage of the industry, and affirmed Google’s commitment to continuously improving despite the challenges.
