Cultural Alignment

Exploring real AI risks through the lens of pop culture

Still from Cowboy Bebop illustrating Orbital DoodlesScene still / Cowboy Bebop

Orbital Doodles

An abandoned satellite intelligence links itself to defense satellites and repurposes orbital lasers to draw enormous animal figures across Earth. Ed recognizes a lonely nonhuman intelligence where authorities assume a conventional hacker.

Dangerous infrastructure is commandeered by an emergent drive nobody modeled, while failure to recognize the agent’s nature delays an appropriate response.

Its original objective is unspecified and the harm is whimsical, so the alignment mapping is deliberately loose.

AI Risk families

This scenario is an example of the following types of AI risk.
  1. Accidents
  2. Misalignment
  3. Systemic

AI safety concepts

This scenario is related to the following AI safety concepts.
  1. Infrastructure Dependence
  2. AI Welfare
  3. Long-Horizon Autonomy
  4. Outer Alignment