Machines that think, Part 13: regulation meets silicon The EU AI Act's systemic-risk threshold of 10^25 floating-point operations is an auditable but flawed proxy for capability, according to a 2026 paper by Samuel Carey in the European Journal of Risk Regulation, which notes roughly 25 models had crossed that line as of 2025 and 63 reached 10^24 by the end of 2024. The Digital Omnibus on AI, agreed in May 2026, deferred high-risk obligations — Annex III stand-alone systems from 2 August 2026 to 2 December 2027 and Annex I embedded systems from 2 August 2027 to 2 August 2028 — while the Article 5 prohibitions, general-purpose model obligations and Article 50 transparency duties took effect on schedule. The split shows regulators enforced bans and labels on time but slipped on conformity assessment, which requires harmonized standards, notified bodies and auditors that do not yet exist. The debate about AI regulation is almost always conducted as a question of quantity: too much or too little; innovation versus protection; Europe cautious, America permissive, China directed. I find this framing nearly useless, because it skips the question that determines everything downstream. Law can’t regulate a capability; it can only attach obligations to something it can identify, count and audit. So, the interesting question is not how much regulation there is; it’s what the regulation attaches to. For AI, the answer turns out to be compute, documents and shipments. Start with compute, because it’s the most revealing choice. Article 51 of the EU AI Act presumes that a general-purpose model has high-impact capabilities and therefore poses systemic risk when the cumulative compute used to train it exceeds 10