Meta admits its first ‘superintelligence’ was too stupid to survive for three days
Key Points:
- Meta launched "Muse Image," an AI image generation product from its Superintelligence Labs, on July 8, allowing Instagram users to apply AI-driven filters and create personalized images featuring friends by tagging their public accounts.
- The product faced swift and widespread backlash, notably from the actors’ union SAG-AFTRA, which condemned the lack of a clear opt-in for using users' images and highlighted potential dangers and harms.
- Within 72 hours of launch, Meta withdrew the Muse Image product, acknowledging that the feature "missed the mark" despite its intent to offer a creative tool with user control over public content references.
- Some filters used in Muse Image were designed by Instagram creators leveraging Meta AI, but involving creators in the development process did not prevent the backlash.
- The incident underscores Meta’s ongoing challenges with privacy, community trust, and managing the social implications of AI-powered features on its platforms.