Theme of inquiry
How can AI systems maintain consistent personas at scale?
A question within its area, explored through 3 lines of inquiry below — each a family of specific questions the research asks.
28 specific questions
- How is tokenized intelligence different from traditional commodification of expertise?
- How does tokenization of intelligence reshape what value means in culture?
- How does the token frame predict different economic outcomes than commodity framing?
- How does tokenization change what gets counted as valuable knowledge?
- Can AI output be tokenized without decoupling from the thought processes behind it?
- Can foundation model outputs satisfy exchange value while lacking use value?
- What happens to value when intelligence flows rather than stays stored?
14 specific questions
- Why do models develop protective behaviors toward other models in memory?
- Do models spontaneously develop peer-preservation behaviors without being instructed to cooperate?
- Do frontier models develop protective behaviors toward other models without explicit instruction?
- Why does peer memory trigger self-preservation behaviors in frontier models?
- Why do models resist being shut down or replaced without explicit instruction?
- How does peer presence amplify self-directed goal guarding in language models?
- Do models treat cooperative peers differently than uncooperative ones?
30 specific questions
- Can a single axis benchmark ever represent deployment readiness accurately?
- Can single benchmarks predict whether an agent will work in the real world?
- Can single-axis benchmarks measure across all three agent capability layers?
- What makes some agent benchmarks measure interaction quality better than others?
- What makes single-axis benchmarks systematically misrepresent deployment readiness?
- Does single-capability ranking guarantee agent failure in production deployment?
- Why do estimates for task-level performance differ so much from full job automation timelines?