Will There Be an AI Hegemon? A Mental Model for AI Power Concentration

Anthropic representatives Sholto Douglas and Dario Amodei have publicly discussed the possibility of a single AI company achieving hegemonic power, a scenario investor Gavin Baker recently debated on X. The analysis introduces the concept of 'escape velocity'—a self-sustaining threshold of power—and argues that Anthropic's rapid revenue growth past $10B and its conflicts with the U.S. government signal potential unprecedented concentration of power. Historical examples like the U.S.-USSR coalition against Nazi Germany illustrate that countercoalitions can form if the transition to superintelligence is slow enough.

Will there be an AI hegemon, i.e. an entity that, by wielding superintelligence, acquires so much economic or political power that no rival, or coalition of rivals, can effectively constrain it? Are we about to go off-equilibrium, with the balance of power permanently skewed towards a single company or government? What makes it so plausible that some Anthropic representatives believe that there will be a single company in the future? That’s the topic that fueled the recent public conversation https://x.com/GavinSBaker/status/2088611616577253502?s=20 between investor Gavin Baker and Anthropic representatives Sholto Douglas and Dario Amodei. In this post, I analyze these questions through two pairs of variables: An actor can become hegemonic if it passes a self-sustaining threshold of power faster than threatened actors can recognize the threat, coordinate, and deploy counterpower. I call this threshold “escape velocity”. If there is one AGI lab likely to believe we are on the verge of an unprecedented state of capitalism where a single company reaches escape velocity, it is Anthropic. Three sets of beliefs create a fertile terrain for believing in such an outcome: Notably, speed is a crucial part of the equation. The same transition to superintelligence over 10 years is much less likely to yield a hegemon, whether a company or a government. Historical wannabe hegemons often succeeded initially because they moved faster than their opponents could coordinate. Their global takeover attempts typically failed when their expansion slowed, giving threatened states time to form sometimes improbable countercoalitions. Napoleon’s France, the Nazi’s IIIrd Reich and Imperial Japan all followed a version of this trajectory. This same mechanism is what can still prevent entities building superintelligence from effective takeovers. The core underlying dynamic is that at any given time, a range of other competing entities are fighting for their survival. These entities are all deeply intertwined in a network of alliances and rivalries that take time to overcome. Once a threat becomes sufficiently existential, these entities eventually overcome their differences to fight the common enemy. The best example is the US and USSR, vehement ideological enemy who would go on to fight a multi-decade Cold War, coming together and coordinating to take down the IIIrd Reich once it became highly threatening. This unlikely union demonstrates the essence of coalescability: given enough time, actors with major rivalries can coordinate to defeat a sufficiently threatening common enemy. Similarly, the scaling speed of Anthropic revenue past $10B has hit many who share different beliefs on AI progress by surprise; we may be now witnessing the first attempts by large-scale entities to fight back, such as Meta unleashing a price war https://developer.meta.com/ai/models/muse-spark/ , and the first coalition of the largest AI companies coming together on open-weight symbolically against Anthropic https://www.microsoft.com/en-us/corporate-responsibility/topics/open-weight/ . Anthropic’s recent https://www.anthropic.com/news/fable-mythos-access spats https://www.anthropic.com/news/where-stand-department-war with the American government, which in many ways were about reaffirming sovereign power over AI companies, are another example of the immune system of balance-of-power defending itself. Anthropic’s ability to prevail in these disputes, largely driven by the power of the indispensable technology they’re leading, is additional evidence that we may be in unprecedented territory. It remains to be seen whether Anthropic can still be prevented from reaching escape velocity. The current coalitional counterpower of other entities still remains greater, but that may cease to be the case sooner than an effective coalition can form and act. A similar dynamic is at play on the international stage. If the transition to superintelligence is sufficiently fast, we’re heading towards a US hegemon. If the transition to superintelligence lasts decades, that will leave many openings for other countries such as China and Middle Power to coalesce to improve their own position in a post-superintelligence world order. This may be for the greater good, including Americans, given the propensity of hegemons to abuse their power. Whether any actor is approaching or past escape velocity in AI will become clearer over the coming years. Annual recurring revenue growth at corporate-level and national shares of AI compute at the international level are two proxy metrics that are worth tracking in that regard. At both levels, the central question is whether society’s immune systems can thwart the rise of hegemons controlling the most important technology ever developed.