For the Republic
← Command Center / 📄 Article / 2026-03-15

Buying Both Doors

Draft Complete — Pending Author Review

Outline

4/10

Article Outline

Working Title

Buying Both Doors

Target Length

~1,500 words

Structural Choices

  • Shape: Cold Contrast. The thesis is built on a gap -- what voters see (ads about ICE, healthcare, Trump) versus what the money wants (no AI regulation). That gap is the engine. Opening with the juxtaposition does the argumentative work immediately and lets the framework name itself out of the tension rather than being imposed from above.
  • Counterargument approach: Woven throughout. The strongest objections (money doesn't guarantee outcomes, the patchwork problem is real, the industry isn't monolithic) arise naturally at specific points in the argument. Addressing them in-line keeps momentum and avoids the dedicated steelman section that the last three articles all used.
  • Closing approach: The lingering question. The last three articles all closed with callbacks to their openings. This one should end on an unanswered question the reader carries with them -- one that reframes "bilateral capture" from an AI story into a permanent structural feature of post-Citizens United democracy.
  • Recent articles used: "Constitutional Arbitrage" (Framework-First, dedicated steelman, callback close), "The Racket Never Dies" (Framework-First/Narrative hybrid, woven steelman, callback close), "The Arsonist's Invoice" (Framework-First/Cold Contrast hybrid, dedicated steelman section, callback close). All three open by establishing a conceptual framework early and close by returning to their opening image. This outline breaks both patterns.

Structural Overview

The article opens in the gap between what a voter sees and what the money is buying -- two juxtaposed realities that share a ZIP code but not a purpose. The tension between them IS the thesis, and it gets named ("bilateral capture") only after the reader has already felt it. From there, the article traces the mechanism's architecture through FEC filings, shows that it already worked once (Fairshake/crypto), puts a human face on it (the Bores case), and lands on a question that makes the reader realize this isn't about AI at all -- it's about what primaries are for in a system where the bracket can be bought. The reader's journey: confusion (why are AI companies running ads about immigration?), recognition (oh -- the ads aren't the product, the candidates are), unease (this already worked), and finally a question that won't go away.

The Gap (~300 words)

Opens with: Two realities, same congressional district, same week. Reality one: a voter in Chicago's northwest suburbs sees attack ads. One is about ICE. Another about healthcare costs. A third about Trump. They feel like standard midterm fare. Reality two: the FEC filings behind those ads. $62 million flowing into four Illinois Democratic primaries from AI, crypto, and AIPAC PACs. Leading the Future split $5 million to Think Big (D) and $5 million to American Mission (R) -- a perfectly balanced bipartisan bet. Meta's PACs mirror the structure: Forge the Future (R) and Making Our Tomorrow (D). None of the ads mention AI. Not one. Purpose: The juxtaposition is the argument in miniature. The reader should feel the dissonance before they can name it. This is not a framework being explained -- it is being experienced. The gap between what the voter sees and what the money wants is the entire article compressed into a single scene. Key evidence/examples: NBC News documentation of AI PAC ad content (immigration, healthcare, Trump -- never AI). FEC filings showing Leading the Future's dual-PAC architecture (NOTUS). The $62 million figure from WBEZ Chicago reporting. Meta's parallel PAC structure. Leads into: A single beat -- something like: "There's a word for what's happening here, and it isn't lobbying." -- that bridges into naming the mechanism.

The Architecture of "Both Doors" (~350 words)

Argument beat: Name bilateral capture, but let it emerge from the evidence just presented rather than arriving as a definition. The concept has three moving parts: (1) the dual-PAC architecture ensures bipartisan coverage by design, not accident; (2) the ad misdirection confirms the industry knows its actual agenda is unpopular -- 69-80% of Americans want more AI regulation, so the real transaction must be laundered through other issues; (3) the harvest, where the policy return matures. The harvest is already visible: a dormant commerce clause argument published by Andreessen Horowitz became a presidential executive order preempting state AI laws three months later. Governor Hochul crossed out the text of New York's strongest AI safety bill and replaced it with industry-preferred language after a Leading the Future donor held fundraisers for her. Key evidence/examples: The a16z-to-executive-order pipeline. The Hochul RAISE Act gutting (Popular Info, TechPolicy.Press). The 69-80% polling on AI regulation. Josh Vlasto as the human bridge -- the Democratic operative who ran Fairshake's 2024 crypto strategy now runs Leading the Future. Same person, same playbook, new industry. Counterargument woven in: Acknowledge directly that the state patchwork problem is real -- 1,200+ bills, contradictory requirements, PerceptIn's $344,000 compliance nightmare. The industry's policy preference (federal coherence over state fragmentation) is substantively defensible. The problem is the method: the appropriate response to fragmentation is democratic deliberation over a federal framework, not industry pre-selection of the legislators who will write it. Separate the policy from the politics in one clean move. Relationship to thesis: This is where the reader goes from "that's weird" to "that's engineered." The dual-PAC structure isn't inference -- it's documented financial architecture. And the harvest evidence shows it's already delivering returns before the 2026 elections even happen. Transition: "If this sounds theoretical, it isn't. Someone already ran this experiment."

The Proof of Concept (~300 words)

Argument beat: Fairshake's 2024 crypto playbook is the proof that bilateral capture works -- not as theory but as documented cause-and-effect. $290 million in bilateral spending. Sherrod Brown defeated. The GENIUS Act passed. Industry-friendly regulators installed. And now the same operative (Vlasto) has carried the playbook to AI. This isn't a new strategy being tested. It's a proven one being scaled. Then make it human: Alex Bores, a former Palantir data scientist who left over the company's ICE contract and became the country's leading state-level AI regulator, is now being attacked by a PAC funded by Palantir's co-founder Joe Lonsdale. $1.8 million in negative ads -- about his work at Palantir. The industry he tried to regulate is using his own resume as a weapon against him. The Bores case puts a face on the abstraction and makes the mechanism visceral. Key evidence/examples: Fairshake spending data and outcomes. The Vlasto bridge from crypto to AI. The Bores case study (NOTUS, NBC News). The $125M Leading the Future + $65M Meta vs. $20M Anthropic Public First spending ratio (~6:1). Counterargument woven in: Engage the political science evidence honestly. Money doesn't mechanically determine outcomes -- AIPAC spent $14.5M to defeat Summer Lee and failed. The relationship between spending and winning is probabilistic, not deterministic. But bilateral capture doesn't need a 100% success rate. It needs a high enough success rate to change the expected value calculation for any candidate considering a pro-regulation position. In low-turnout primaries, where a few million dollars can constitute a majority of all spending, the marginal impact per dollar is enormous. And Anthropic's counter-spending -- real, significant, $20M -- proves this is a choice by specific actors, not an inevitable feature of industry participation. But the 6:1 ratio tells the structural story. Relationship to thesis: This section answers the "does it actually work?" question with evidence, not assertion. It also humanizes what could otherwise remain an abstraction about PAC architecture.

The Real Election (~250 words)

Argument beat: Zoom out. What makes bilateral capture different from ordinary political spending is not the amount -- it's the target. Traditional lobbying persuades legislators after they're elected. Bilateral capture selects them before they're nominated. The decisive moment has moved upstream to the primary, where turnout is low, voter awareness is minimal, and the general election becomes a formality -- a choice between two pre-approved options. This is the difference between playing the game and rigging the bracket. And it's not just AI. AIPAC perfected this model ($100M+ in bilateral primary spending). Crypto proved it could be replicated. AI is scaling it. Whatever industry needs to avoid regulation next will adopt it. Citizens United is the infrastructure -- billionaire election spending has increased 163x since 2010, and over 80% of 2024 billionaire spending used channels prohibited before that decision. Bilateral capture is the most sophisticated expression of a fifteen-year trend. Name it now, while the mechanism is still visible. Once the primaries are over and the captured candidates are seated, it disappears. Key evidence/examples: Roosevelt Institute data on post-Citizens United spending. The AIPAC structural parallel (used precisely -- AIPAC perfected the mechanism, AI adopted it -- without being drawn into the Israel-Palestine debate). Primary turnout data. The 163x increase in billionaire spending. Relationship to thesis: This is the "so what" -- the moment bilateral capture stops being an AI story and becomes a democracy story. The reader should realize the scale of the pattern only here, when they're already inside the argument.

Close (~100 words)

Approach: The lingering question. Landing: The 2026 primaries are underway. The money is already spent. The ads are already running -- about everything except what the money wants. And here is the question that should keep you up at night: if bilateral capture works, how would you know? The election still happens. The voters still vote. The winner still gives a victory speech. Everything looks like democracy. The question is whether anything that happened before the ballot was printed still counts as a choice. Emotional register: Quiet discomfort. Not alarm, not hope -- the unresolved tension of a question that doesn't have a clean answer. The reader should leave unsettled, carrying the question, seeing the pattern in every primary ad that doesn't mention the industry that paid for it.

Architecture Notes

Breaking the pattern: The last three articles all introduced a named framework within the first 200-300 words and then applied it. This outline delays naming "bilateral capture" until the reader has already felt the mechanism through the juxtaposed realities of the opening. The term should arrive as recognition, not instruction. Similarly, the last three articles all closed by returning to their opening image. This one closes on a question that opens outward rather than circling back. The structural variety is deliberate and necessary.

Tone arc: The opening should feel like investigative journalism -- specific, factual, almost reportorial. "A voter in Chicago's northwest suburbs sees..." The middle sections should build analytical confidence as the framework proves itself against evidence. The Bores case is where the voice should get closest to personal anger -- not ranting, but the controlled heat of describing someone being attacked with his own biography by the industry he tried to hold accountable. The close should pull back to a quieter, more unsettling register -- not a call to action, not a punchy final line, but an honest question that lingers.

The "misdirection" framing needs care. The thesis's claim that AI PAC ads about immigration and healthcare constitute "misdirection" is analytically vulnerable -- every issue PAC runs ads about what voters care about, not about the PAC's policy agenda. The draft writer should frame this as evidence of what the money is and isn't buying rather than as deliberate voter deception. "The ads are about what voters care about. The money is about what the industry needs. That gap is the tell." This is more defensible than "they're hiding their agenda" and ultimately more damning.

The elections haven't happened yet -- own it. The article is naming a strategy in real time, before the outcome is known. This is not a weakness to hide; it's a strength to claim. The Fairshake precedent is the evidence that the strategy works. The AI cycle is the test in progress. Frame the article as a warning that has a window -- once the primaries are over and the candidates are seated, the mechanism becomes invisible. That urgency is the article's reason for existing right now.

AIPAC reference -- surgical, not editorial. The structural parallel (bilateral primary spending, ads about everything except the policy agenda) is real and important. Use it precisely: "AIPAC perfected this mechanism; AI companies adopted it." Do not get drawn into the Israel-Palestine debate, do not imply moral equivalence, and do not spend more than two sentences on it. The parallel is structural, not moral.

Anthropic as evidence, not exoneration. Anthropic's $20M in counter-spending through Public First is genuinely complicating and should not be dismissed. Frame it as proof that bilateral capture is a choice made by specific actors -- the anti-regulation faction could have engaged through democratic deliberation, and one major company chose exactly that. But the 6:1 spending ratio makes the "marketplace of ideas" framing untenable. A 6:1 ratio is not a debate. It's a purchase.

Word budget rationale: The opening gets 300 words because the juxtaposition needs room to land -- the reader must feel the gap, not just be told about it. "The Architecture of Both Doors" gets the most space (350) because it carries the most analytical weight: the framework definition, the harvest evidence, AND the patchwork counterargument. "The Proof of Concept" gets 300 because it carries both the Fairshake precedent and the Bores human story plus the political science counterargument. "The Real Election" gets 250 for the zoom-out. The close gets only 100 -- a question, not a speech.