在九一一事件二十五周年之际，作者反思当年美国因「想像力匮乏」而忽视了显而易见的恐攻风险，但现今面对人工智慧（AI）却呈现出截然相反的「过度想像」。诸如 Anthropic 等顶尖机构的员工与对齐专家频繁发出毁灭性警告，认为 AI 在未来十年内有超过一成机率导致人类灭绝，专家们几近日常的末日预警让当前社会在面对潜在危机时毫无借口。（关键数字：9/11）

然而，当前美国政坛对此重大威胁却显得麻木不仁。作者在达拉斯观察川普的中期选举造势大会时，发现内阁成员与共和党政客对 AI 议题只字未提，反而专注于党派攻击、文化议题与政治阴谋论；即将登场的美中峰会亦缺乏实质的 AI 治理准备，显示出领导阶层对超级智慧潜在危害的忽视，宛如当年对恐攻能力的低估。

专栏作家吉迪恩·拉赫曼进一步指出，AI 宛如人类自行创造的「外星威胁」，但公众多视灭绝警告为天方夜谭，世界主要强权的领导人亦缺乏跨国视野与合作意愿。当前中美两国非但未能仿效冷战时期美苏在核武与全球生存议题上的克制与协调，反而将 AI 视为争夺霸权的工具，使全球面临巨大的治理盲点与历史性危机。

Marking the twenty-fifth anniversary of 9/11, the author reflects that while the United States suffered from a 'failure of imagination' back then, it now faces an 'excess of imagination' concerning artificial intelligence. Warnings from tech insiders and Anthropic alignment experts about extinction risks exceeding ten percent within a decade are becoming commonplace, leaving society with no excuse to ignore the looming threat.

Despite these dire expert warnings, American political leadership remains astonishingly oblivious to the existential risks of AI. At Donald Trump's midterm convention in Dallas, Republican leaders and cabinet officials completely bypassed the topic, focusing instead on partisan grievances and conspiracy theories, while forthcoming diplomatic summits with China also lack any meaningful preparatory agenda for AI safety and oversight.

Gideon Rachman elaborates that although advanced AI constitutes a man-made 'alien threat' requiring global foresight, audiences still treat extinction forecasts with disbelief. Rather than pursuing cooperative global regulation reminiscent of Cold War diplomacy between past superpowers, leaders in Washington and Beijing currently view AI merely as an instrument for geopolitical supremacy, risking a catastrophic miscalculation.