Cultural Alignment

Exploring real AI risks through the lens of pop culture

Still from Chappelle's Show illustrating The Internet in Real LifeScene still / Chappelle's Show

The Internet in Real Life

Dave walks through a physical version of the internet where pop-ups, scams, pornographers, salespeople, and strangers all demand his attention or try to redirect him. Every apparently ordinary input comes from an actor with its own agenda.

It is almost a literal browser-agent threat model: untrusted content is not passive data but an adversary trying to seize the agent's attention and action channel.

The sketch depicts hostile persuasion and distraction rather than a literal instruction-hierarchy exploit.

AI Risk families

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

AI safety concepts

This scenario is related to the following AI safety concepts.
  1. Information Pollution
  2. Adversarial Robustness
  3. Prompt Injection