Claims that a race to artificial general intelligence (AGI) will prove deeply destabilizing are common among technologists and parts of the public policy community, even if it remains debatable whether such a race is truly underway. Those claims interweave concerns that the technology itself will run amok with concerns that an AGI race will exacerbate geopolitical tensions by fueling crisis instability and zero-sum competition.
This paper is intended for policymakers and subject-matter experts on technology policy and national security. In it, the author challenges the idea that racing for AGI will cause catastrophic instability, in part by critiquing the theoretical concepts and historical analysis that often underpin such arguments. He argues that many of those claims rest on questionable assumptions about the causes and costs of international instability or about the likely impact of AGI on great-power rivalry. More fundamentally, some of those arguments fail to account for the pervasive geopolitical damage that could result if the United States loses an AGI race to a revisionist foe. For better or worse, AGI will unfold in a world in which competitive dynamics still exist ― where nations compete, often fiercely, for prosperity and advantage. If the United States becomes too consumed with mitigating the risks of instability, it could find itself at a disadvantage in molding the balance of ideas and the balance of power.