The Limits of Intelligence

What can advanced systems do, what does their behavior establish, and what can be verified, controlled, or guaranteed?

This initiative investigates capability, alignment, evaluation, prediction, planning, verification, and safety–capability tradeoffs. It distinguishes structural limits from problems that yield to better engineering or evidence.

What changes: Builders and institutions gain explicit operating conditions—supporting confident deployment where claims hold and bounded use where they do not.