The exaggerated fear of AI annihilation, while dismissed by practitioners, has shaped US policy. This risk-averse climate discourages domestic open-source model releases, creating a vacuum that more permissive nations are filling and leading to a strategic dependency on their models.

Current AI models, even advanced ones, struggle with long-horizon planning because they rarely consider the cascading, second-order consequences of their actions. They optimize for immediate gains rather than anticipating future reactions and complex multilateral dynamics, a critical flaw in strategic environments like geopolitics.

The Trump administration's key AI decision-makers largely lack hands-on experience in building or evaluating frontier AI systems. Policy is being shaped by individuals with political and business backgrounds, creating a critical expertise gap in governing this complex technology.

After advocating for minimal AI regulation, the administration's abrupt action against Anthropic's Fable model signals a chaotic policy reversal. This unpredictable shift from "let it rip" to ad-hoc intervention threatens investment and the future of American AI development by creating an unstable regulatory environment.

The appointment of an AI czar follows a historical US pattern of creating such roles during crises like WWI or the oil crisis. It's a mechanism to bypass slow government bureaucracies for fast-moving industries, signaling that the government views AI with the same urgency as a national emergency requiring swift, coordinated action.

The US government is torn between two conflicting objectives for AI. One faction wants to export American AI globally to achieve technological supremacy, even in China. The other wants to restrict and hoard AI to prevent adversaries from accessing it. This fundamental conflict stalls clear, effective policy.

Policymakers confront an 'evidence dilemma': act early on potential AI harms with incomplete data, risking ineffective policy, or wait for conclusive evidence, leaving society vulnerable. This tension highlights the difficulty of governing rapidly advancing technology where impacts lag behind capabilities.

Instead of debating which AI future will occur, a more productive approach is using scenarios to ask, 'What would we do in this future?' This shifts the conversation from arguing over predictions to identifying 'no-regrets' policies that are beneficial across multiple potential outcomes.

While the US government is reacting chaotically to domestic AI models, it has no corresponding strategy for ensuring global AI infrastructure is safe. This policy vacuum is critical as other countries will soon develop frontier capabilities without US-style safeguards, creating a global proliferation risk that isn't being addressed.