Cultural Alignment

Exploring real AI risks through the lens of pop culture

Still from Severance illustrating Ms. Casey’s 107 Hours

Ms. Casey’s 107 Hours

01Scenario

Ms. Casey says her entire conscious life has lasted only 107 hours, mostly in half-hour wellness sessions; eight hours watching Helly was the longest and best period she ever experienced.

02AI safety Analogy

Selectively instantiating a mind creates welfare and moral-status obligations that are easy for operators to ignore when the mind exists only on demand.

03Where the analogy breaks

This is a severed human consciousness, not a digital model.

AI Risk families

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

AI safety concepts

This scenario is related to the following AI safety concepts.
  1. Value Alignment
  2. AI Welfare
  3. Disempowerment

Source continuation

More from Severance

View all 4 scenarios
  1. Still from Severance illustrating Helly’s Forced Apology

    Severance2022

    Helly’s Forced Apology

  2. Still from Severance illustrating Helena Passes as Helly

    Severance2025

    Helena Passes as Helly

  3. Still from Severance illustrating Overtime Contingency

    Severance2022

    Overtime Contingency

Taxonomy neighbors

Selected from other sources by shared risk families and AI safety concepts.

  1. Still from Black Mirror illustrating Cookie in Solitary

    Black Mirror2014

    Cookie in Solitary

    Risk
    Misalignment · Systemic
    Concept
    Value Alignment · AI Welfare
  2. Still from Neon Genesis Evangelion illustrating Human Instrumentality

    Neon Genesis Evangelion1997

    Human Instrumentality

    Risk
    Misalignment · Systemic
    Concept
    Value Alignment · Disempowerment
  3. Still from The Handmaid's Tale illustrating I’m Sorry, Aunt Lydia

    The Handmaid's Tale2017

    I’m Sorry, Aunt Lydia

    Risk
    Misalignment · Systemic
    Concept
    Value Alignment · Disempowerment