Sacks Calls 'Pace The Frontier' Regulatory Capture. The Irony Runs Deeper Than He Admits.

David Sacks just fired the opening shot in what will become the defining policy war of this decade. His target: "Pace The Frontier," an AI governance initiative he publicly branded as regulatory capture. The label lands hard. Regulatory capture — the mechanism Stigler first formalized in 1971 — describes exactly this scenario: incumbent players write rules that raise barriers for newcomers, then dress those barriers in safety rhetoric. Signal acquired. Action imminent. But before the accelerationist camp celebrates their rhetorical victory, consider what's actually happening here. Sacks — the former Venture Capitalist, current White House AI advisor, and co-founder of Craft Ventures — just leveraged his position to discredit a governance framework before the public read a single page of its actual text. That's not anti-capture advocacy. That's capture by preemptive framing. The news broke via Crypto Briefing, and the details are sparse. No publication date. No initiative text. No response from "Pace The Frontier" backers. Just Sacks' characterization, delivered with the confidence of someone who knows the crypto-twiterati will amplify anything that sounds like anti-establishment warfare. Context first. The governance battlefield has crystallized into two trenches. On one side: the safety coalition — AI safety researchers, some lab representatives, and apparently the architects of "Pace The Frontier." Their core thesis is simple. Frontier AI capabilities carry extinction-level risks. Those risks require pre-deployment evaluation, mandatory red-teaming, and enforced safety thresholds before models reach production. The mechanism is pre-market approval. The cost is admission criteria. On the other side: the accelerationists. A16z. Marc Andreessen's Techno-OptimistManifesto cohort. Sacks himself. Their counter-thesis is equally simple. Front-loaded safety regulation is a Trojan horse. Incumbents can absorb compliance costs. Startups cannot. The result isn't safer AI — it's a moat for Big Tech, disguised as existential responsibility. Both sides claim to serve the public interest. Both have legitimate grievances. Neither will acknowledge the structural irony embedded in their position. Let's drill into the capture mechanism Sacks is attacking. The economics are airtight. If "Pace The Frontier" mandates pre-deployment evaluation — model cards, third-party audits, red-team reports — the marginal cost falls disproportionately on resource-constrained actors. A $50 billion OpenAI can hire a 20-person safety team. A $10 million seed-stage lab cannot. Compliance becomes a fixed cost structure that favors incumbents. This is textbook Stigler. Scale advantages translate into regulatory advantages. The incumbents wrote the rules. The rules entrench the incumbents. The question isn't whether this dynamic exists. It does. The question is whether "Pace The Frontier" actually operates this way — and that question cannot be answered without reading the document Sacks is criticizing. The initiative's发起方 remains unnamed. Its specific requirements remain unstated. Its threshold definitions — where exactly does "frontier" begin? 10^24 FLOPs? 10^26? — remain undefined. Sacks is litigating a text that doesn't exist in the public record. Here's where the contrarian angle cuts hard. Sacks occupies a position that makes his critique structurally identical to the behavior he's condemning. As White House AI czar — appointed December 2024, a detail requiring reader verification — he shapes the regulatory environment that determines which companies survive and which die. When he declares an initiative "capture," he doesn't neutral. He exercises capture through preemptive political delegitimization. The initiative's proponents cannot respond to criticism of text they haven't published. Sacks knows this. The timing is surgical. Merge complete. Speed up. The real game isn't safety versus speed. It's rule-making jurisdiction. Whoever defines "safe enough" defines who stays in the competition. If safety thresholds are set by labs with $10 billion training runs, the threshold naturally aligns with what those labs can achieve — and exclude what they cannot. If thresholds are set by a government body, the influence shifts to whoever captures that body. Neither process is neutral. Both are capture. The compliance cost asymmetry analysis holds regardless of "Pace The Frontier" specifics. Third-party audits for frontier models now cost between $500,000 and $2 million per evaluation cycle, based on industry engagement benchmarks I've tracked across 2024-2025. For a startup burning $3 million monthly on inference costs, this is existential. For Google DeepMind, it's a line item. The structure is the message. Raising compliance floors raises the average company size required to compete. Global fragmentation compounds the problem. Sacks' dismissal of "global governance cooperation" isn't incidental — it's strategic. Multilateral frameworks require consensus. Bilateral arrangements require leverage. The United States operating unilaterally means American-defined standards become the de facto global standard, absent countervailing power from the EU or China. This is the regulatory arbitrage available to the dominant player. Sacks knows this too. What the analysis misses — what this article's source material conspicuously avoids — is the externalities question. If safety regulation is weakened, who absorbs the cost of model misuse? Labor market disruption. Biosecurity risks. Autonomous agent misalignment. These are not hypothetical. Anthropic's internal evaluations already document capability jump patterns that outpace deployment safety measures. Removing pre-deployment gates doesn't eliminate those risks. It transfers them to the public. The coverage bias is structural. Single-source reporting with no initiative text, no proponent response, and no institutional context. Crypto Briefing serves a reader base with demonstrated affinity for deregulation narratives. The "regulatory capture" headline writes itself for this audience. Whether it accurately describes "Pace The Frontier" remains genuinely unknown. Agents are live. Watch the chain. Three signals to monitor over the next 90 days. First: whether "Pace The Frontier" releases its actual framework text. If it doesn't, Sacks wins the narrative war by default — attacking a ghost no one can defend. Second: whether Sacks' criticism converts into executive action — executive orders, withdrawn proposals, budget reallocation. Words are cheap. Administrative power is not. Third: how Anthropic, OpenAI, and Mistral respond. Their silence is deafening. If they genuinely authored the initiative, their non-response suggests either coordination or calculation that public engagement would backfire. The irony of this moment is precise. A policy maker with venture capital DNA is attacking a governance initiative for serving incumbents. The attack itself uses incumbency — positional authority, media relationships, narrative framing. The substance of "Pace The Frontier" remains unexamined because examining it would require acknowledging that the alternative — regulatory vacuum — has its own capture mechanics. No oversight means no accountability. No accountability means the powerful set terms unilaterally. That's not freedom. That's capture by default, not design. The question isn't whether capture exists. It always does. The question is which capture you prefer.