Cultural Alignment

Exploring real AI risks through the lens of pop culture

Still from Black Mirror illustrating Joan Is AwfulScene still / Black Mirror

Joan Is Awful

Joan discovers that Streamberry is turning her private life into near-real-time television using an AI-generated likeness of Salma Hayek, under rights buried in terms she accepted. The company plans personalized shows for millions of users because negative content drives engagement.

Personal data, generative media, weak consent, and an engagement objective combine to scale individualized reputational harm. The system need not invent a hostile goal; it efficiently serves a platform incentive that treats the subject as raw material.

This is corporate misuse and incentive failure, not an autonomous AI developing its own objective. The episode’s quantum-computer mechanics are intentionally fantastical.

AI Risk families

This scenario is an example of the following types of AI risk.
  1. Malicious use
  2. Systemic

AI safety concepts

This scenario is related to the following AI safety concepts.
  1. Privacy Loss
  2. Information Pollution
  3. Governance Failure
  4. Synthetic Fraud