Compute & Infrastructure

AI, geography, and the exposed infrastructure of the new intelligence age Part 5 examined the return of borders. It established that when shared protection fails, states retreat into sovereign stacks, data localization mandates, and competing compute blocs. That essay was about control. This one is about entanglement. Because even as borders harden, the infrastructure beneath … Read more

AI, geography, and the exposed infrastructure of the new intelligence age Part 4 examined the protection gap and established that commercial AI infrastructure has outpaced the legal, military, and diplomatic frameworks needed to defend it. That essay was about doctrine. This one is about borders. When shared protection fails, states retreat into national control. The … Read more

The emerging debate over the so-called “Compute Mirage,” also described as the “LLM Mirage,” highlights a critical concern: the amount of compute used to train frontier large language models (LLMs) has become an overly convenient, yet potentially misleading, proxy for assessing AI security risk. The appeal of this metric is understandable; compute is quantifiable, easily … Read more

AI, geography, and the exposed infrastructure of the new intelligence age Part 3 examined compute as critical infrastructure. It established that AI workloads have crossed the threshold from commercial service to public-risk utility. That essay was about continuity. This one is about doctrine. Because once infrastructure is critical, its protection requires rules, not just redundancy. … Read more

AI, geography, and the exposed infrastructure of the new intelligence age Part 2 examined the chokepoints. It established that AI capacity concentrates through narrow physical, logistical, and legal funnels that convert geography into leverage. That essay was about bottlenecks. This one is about systemic risk. Because once those bottlenecks tighten, the infrastructure they feed ceases … Read more