Trump AI adviser rules out formal licensing; ad hoc regime continues

The article presents a sharp critique of the Trump administration's approach to AI policy. Rather than adopting formal, transparent licensing rules, the administration has chosen to regulate AI through ad hoc decisions—a framework that the author argues amounts to regulatory arbitrariness masked as deregulation. Krishnan's statement that he thinks only "about this government and this moment in time" rather than how future administrations might exploit these precedents suggests a short-term orientation that could invite political abuse of emergency powers.

Underlying this policy choice, the author identifies a deeper problem: influential figures shaping White House policy flatly refuse to acknowledge the capabilities of existing closed frontier models and have done so for years. This disconnect between reality and policy understanding is presented as a root cause of poor regulatory decisions. The author outlines three conceptual "pills" that policymakers must swallow: first, understanding what current AI can already do (the "AI pill"); second, grasping what it will likely do in the near future (the "AGI pill"); and third, reckoning with superintelligence (the "ASI pill"). Without at least the first two, the author contends, AI policy cannot be coherent. The framing suggests that the administration's anti-licensing stance may partly stem from denial about existing model capabilities rather than genuine confidence in market-driven alternatives.