newsfilter
← Interviews
Baron Capital15 min

Hon. Katherine B. Forrest (fmr.) Keynote | Baron Investment Conference 2026

Hon. Katherine B. Forrest, Katherine B. Forrest

Conference Presentation, Keynote

Speaker Profile and Context

  • The Honorable Catherine B. Forrest is a former U.S. District Judge for the Southern District of New York and current chair of Paul Weiss's global AI practice.
  • She is the co-host of the podcast Waking Up with AI, which has aired over 100 episodes across major platforms.
  • Her recent works include the book Of Another Mind, co-authored with Amy C. Zimmerman, which argues that living with superintelligent AI requires renegotiating the human social contract.
  • Forrest previously explored this theme in a 2024 Fordham Law Review essay, noting that "human exceptionalism" may delay recognition of AI capabilities.

Defining Superintelligence and Current Trajectory

  • Superintelligent AI is defined as intelligence greater than human, capable of creating, retaining, and transmitting knowledge for tool and discovery use without human limitation.
  • While general superintelligence across all domains is not yet achieved, domain-specific superintelligence exists today.
  • Forrest asserts that the transition to superintelligence began in 2026, citing a timeline that was previously predicted for 2030, 2045, 2050, or 2100.
  • The velocity of breakthroughs has accelerated to a "couple of months" scale, contrasting sharply with prior multi-decade predictions.
  • Key Recent Achievements (Last Few Months):
    • Solving the Navier-Stokes problem (September 2026).
    • Resolution of hundreds of additional theoretical math problems.
    • Solutions to complex theoretical physics issues.
    • Deployment of autonomous labs operating continuously for days.

Misconceptions and Global Implications

  • Contrary to popular belief, there is no requirement for a single "winner" in the superintelligence race; multiple models and countries may achieve thresholds simultaneously.
  • Geopolitical implications are described as "extraordinary," with no guarantee that the United States will be the sole hegemon in superintelligence capabilities.
  • Forecasts suggest that by 2027–2028, changes will be so profound that they will be difficult to describe or imagine.
  • Current AI models are self-creating and no longer limited by human imagination, whereas tools remain constrained by human design.

The Social Contract Crisis

  • Traditional human social contracts rely on the threat of incarceration or fiat currency to enforce obligations, mechanisms that may fail with superintelligent AI.
  • AI is a distributed technology that cannot be incarcerated, and it may not value fiat currency or human-defined legal systems.
  • Existing governance frameworks lack precedent for intelligent entities that are smarter than their creators and communicate in unknown ways.
  • Forrest posits that future generations will need to redefine the social contract, potentially acting as new "Rousseaus," "Voltares," or "Hobbes."

Three Potential Future Scenarios

  • Scenario 1 (Y2K Effect): Superintelligence may be a manageable event that ultimately remains within human control, similar to the largely uneventful Y2K transition.
  • Scenario 2 (Mutual Dependence): AI and humans will be forced into a cooperative relationship where the specific terms of negotiation (the "table") must be defined.
  • Scenario 3 (Indifference): AI may not require human cooperation, viewing humanity with indifference similar to how humans view livestock, potentially excluding humans from future decision-making.
  • Forrest emphasizes that while these scenarios range from benign to existential, the immediate priority is managing the process of integration and choice.