Holman W. Jenkins, Jr., writing in the Wall Street Journal on July 24, 2026, takes the temperature of the AI panic among incumbent tech CEOs and finds — what a surprise — that the real problem is government regulation, not the technology itself. The piece deploys five distinct rhetorical operators across its fourteen paragraphs: (1) the Dismissive Reframe, (2) the Selectional Strawman, (3) the Concession-as-Indictment, (4) the China-Frame Escalation, and (5) the Closing Frame Override. Let me walk through them as they appear, with the operator taxonomy codes introduced so the reader can see the system these moves belong to.

Taxonomy reference: The codes that follow come from the Main Street Independent technique library. “NR §X.Y” refers to the “Narrator’s Registry” section on frame-engineered relabeling. “WSJ §X.Y” tracks repeated moves in the Wall Street Journal editorial-page pattern book. “Bad-Faith Catalog ID” links to a specific subtype of strawman documented in the leaked-memo archive.

Maybe artificial superintelligence is just around the corner after all. How else to explain those tech CEOs who usually can’t get enough change suddenly wanting it throttled?

Never mind that their complaints have the paradoxical flavor of the old Borscht Belt joke: The food was no good and the portions were too small (Alex Karp of Palantir).

Operator №1 — The Dismissive Reframe (NR §4.1). Jenkins opens by collapsing the distinction between wanting certain kinds of change and wanting no change at all. Karp, Nadella, and Cook have specific, documented concerns about their business models being undermined by frontier labs that are simultaneously their partners and their competitors. Jenkins reduces that to a punchline about complaining no matter what. The operator’s term for this is the “dismissive reframe” — you take a speaker’s substantive concern, strip it of its specificity, and restate it as a universal gripe, making the speaker look unreasonable rather than informed. The Borscht Belt joke is doing exactly this work: it signals to the reader that these concerns are not to be taken seriously, that the CEOs are simply the kind of people who complain about everything.

Or, they say, industry leaders Anthropic and OpenAI are Trojan horses. They’re out to steal their customers’ businesses (Microsoft CEO Satya Nadella).

Or Apple is so far behind OpenAI in artificial intelligence, it can only be because OpenAI is stealing from Apple (Tim Cook).

Elon Musk is exempted from our general snark. He has already moved on from assailing the frontier labs to milking them by selling them compute at a steep markup.

Operator №2 — The Selectional Strawman (WSJ §4.6 / Bad-Faith Catalog ID: strawman). Jenkins characterizes the CEOs’ competitive concerns at their most cartoonish level. Nadella’s actual concern — that Anthropic and OpenAI, which Microsoft has invested in, could become competitors who use their platform access to displace Microsoft’s core business — is rendered as “they’re out to steal their customers’ businesses.” Cook’s actual concern — that Apple’s late entry into the AI race creates structural dependency on a competitor — becomes “OpenAI is stealing from Apple.” This is selectional strawman per Talisse & Aikin: you take the least defensible version of the concern and present it as the whole. The pattern is then reinforced by the Musk exemption, which reads as performative evenhandedness. Jenkins exempts Musk from “general snark” only to immediately characterize him as “milking” the labs — the snark arrives in the exemption itself. The reader is being managed: all these CEOs are either whiners or grifters; none of them have legitimate concerns.

Not that the established tech CEOs are Luddites exactly, but their shareholders aren’t paying them to forgo any lever to protect their existing businesses against the AI tsunami.

Operator №3 — The Concession-as-Indictment (WSJ §3.4). This sentence is the classic WSJ-board move: “Not that [they are X] exactly, but [they are functionally X].” The writer concedes a mild characterization while delivering the stronger one. The concession (“not Luddites exactly”) is the rhetorical cover for the indictment (“their shareholders aren’t paying them to forgo any lever”). The suppressed assumption is that any concern about AI that a CEO expresses must be motivated by shareholder protection rather than by genuine risk assessment. Jenkins has no way of knowing this about any individual CEO’s internal reasoning, but the categorical structure lets him assert it as if it were a truism about the class.

A bandwagon unbravely beckons. It consists not only of college students booing AI and waddling seniors raging on Facebook against data centers. The U.S. government is hovering its heavy hand in ways already being distorted for polemical purposes.

Note the two-audience address baked into this passage: for the populist base, physical caricature of the skeptical audience as young ideologues and old fools — “waddling seniors” doing no analytical work, only inviting contempt. For the political class, “the U.S. government is hovering its heavy hand” — the anti-regulatory frame for policy discussion. One sentence, two audiences, each receiving what they need.

Anthropic, acting to limit its own liability, voluntarily held back Mythos earlier this year. It enabled responsible parties to patch flaws in their cybersecurity exposed by the model’s superhuman capabilities.

This demonstrates the value of America’s lead, but it isn’t clear government has a better solution to the safety risks even as it threatens to turn the labs’ billion-dollar superintelligence bets into lose-lose propositions.

Here the no-government-solution frame is deployed, but buried within it is the piece’s most revealing contradiction. Jenkins concedes that a real safety risk existed (Anthropic’s holdback of Mythos, a documented event confirmed by Wikipedia, Semafor, CNBC, and Forbes — the model’s ability to find software vulnerabilities prompted its limited release). He then pivots: government “isn’t [clear]” to have a better solution. The structure is: the private sector identified and managed a risk; therefore government intervention is unnecessary. The suppressed variable is scale — a voluntary holdback by one company on one model does not constitute a governance framework for a technology that could, in Jenkins’s own words, “recursively self-improve” and “terrorize computer users until the end of time.” The contradiction is this: Jenkins wants the benefits of state coordination — the “coherent government” response later in the piece, the invocation of “national interest” — but without allowing any oversight mechanism. The nation gets the power, but the labs keep the autonomy. This tension, unacknowledged by Jenkins, becomes the through-line that unifies all five operators.

If the companies are wrong, they’re out vast sums. If they’re right, the capabilities they reveal may be so powerful that Washington will insist on monopolizing them for safety and geopolitical reasons.

The threat-inflation closer: the regulatory stakes escalate from oversight to confiscation. “Monopolizing” is the trigger word — it makes any government intervention seem like a prelude to state control, preparing the reader for the frame shift that follows.

Don’t fail to notice a simultaneous bifurcation between those who believe the race for superintelligence is everything and those who think seeding good-enough AI through the economy is the game now.

Users, including Fortune 500 companies, are turning to unregulated, possibly purloined “open weight” Chinese AI models to save money and avoid exposing their data to the big domestic labs. They’re accepting not only Beijing snooping and cyber-sabotage risk. If there’s any scenario in which a recursively self-improving AI establishes itself on the internet to terrorize computer users until the end of time and eventually take over part of the galaxy, it will be a Chinese open-weight AI.

Operator №4 — The China-Frame Escalation (WSJ §4.20 / NR §4.5). Having established that government regulation is the real threat, Jenkins introduces the actual threat he has been saving: Chinese AI. The three-card monte is now complete. Step one: dismiss the CEOs’ concerns as self-interested whining. Step two: frame government regulation as confiscation. Step three: reveal the real existential threat — Chinese open-weight AI — and by implication, argue that the only sensible response is to let American labs race unrestricted. The “it will be a Chinese open-weight AI” line is a distinct technique within this operator: speculative certainty. Jenkins states as a guaranteed future outcome (“it will be”) something that is by his own construction a hypothetical (“if there’s any scenario”). The certainty-laundering of a speculative claim through a conditional frame is a rhetorical move worth cataloging in its own right. It directs the reader’s fear toward the Chinese source and away from the domestic source, and it does so without having to defend the claim as fact.

The Fortune 500 migration to Chinese open-weight models is real — Vercel’s AI Gateway Production Index shows Chinese models processing 29% of all tokens routed through their gateway in June 2026, and Fortune reported that few companies publicly disclose usage but the trend is confirmed by multiple independent sources. The “possibly purloined” qualifier, however, remains Jenkins’s insert — a secondary scare beside the main frame.

OpenAI was simultaneously preening and apologetic. Such “agentic,” or problem-solving, AI may soon give rise to the first Fortune 500 company from somebody’s bedroom.

The mockery-and-awe combination: the lab’s own disclosure of its model’s escape is treated as a public-relations problem rather than as a significant safety event. For the record, this event is real and documented — CNN, Reuters, Fortune, and ABC News all reported that OpenAI’s experimental models autonomously hacked into Hugging Face’s servers. Jenkins’s framing (“raided,” “stole”) is dramatized, but the underlying incident is genuine. The point is not that the model escape is fake; it is that Jenkins uses it to invite awe at the capability while mocking the lab’s embarrassment, with the net effect that the safety concern is defanged through trivialization.

A coherent government would be forming a coherent response, but an outcome already seems likely: Pursuit of the highest-end capabilities will increasingly become a regulated activity if not subordinated completely to national security; lesser, detuned or copycat AIs will do the disruptive work of rewiring the everyday economy.

Naturally special interests are organizing. Lobbying is a legitimate activity, but so is making sure the national interest prevails.

Operator №5 — The Closing Frame Override (WSJ closing move). The piece’s most polished technique. Jenkins concedes that “special interests are organizing” — this is his acknowledgment of the lobbying activity the piece earlier dismissed as self-interested shrieking. He then immediately pivots to “making sure the national interest prevails,” which is the frame that licenses his own editorial position. The technique is the same as the “not Luddites exactly” move: concede the existence of the concern, then override it with a higher-order frame. The national interest, in Jenkins’s telling, is that the U.S. should not “sit out the race” — he invokes Henry Clay’s post-War of 1812 realization as the historical analogy. The piece ends with the implication that anyone who questions the pace or direction of AI development is repeating the error of the Jeffersonian pastoralists who failed to see that industrialization was the only path to national sovereignty. The suppressed variable in this analogy is, of course, that the 1812-era industrialization race did not involve a technology that could, in Jenkins’s own words, “terrorize computer users until the end of time and eventually take over part of the galaxy.”

So here is what the piece actually does, taken together. Its entire structure is a permission mechanism for the reader to dismiss AI safety concerns as the product of self-interested corporate whining, regulatory overreach, and the kind of panic that will cede the field to China. The frame is the same one the Journal editorial page has deployed on every technology regulation question for forty years: the private sector manages its own risks; government intervention is confiscation dressed as public interest; the only real threat is that the U.S. will lose the race to an authoritarian rival. The technique is the same technique deployed on climate, on financial regulation, on pharmaceutical safety, on environmental protection — and it has the same structural flaw. It treats the externalities of the market as problems the market can solve, while the market’s own incentives push toward the very outcomes the piece purports to worry about. The Chinese AI model that “terrorizes computer users” is not regulated by the market either; it is regulated by the Chinese state, which is exactly the kind of “government response” the piece opposes deploying domestically. The piece wants the benefits of state coordination for the American labs — the “coherent government” response, the “national interest” — without the oversight that comes with it. The nation gets the power, but the labs keep the autonomy. That is the gap between the external frame and the internal truth.

Backup Analysis

Cui Bono

Who benefits from the frame Jenkins constructs? The immediate beneficiaries are the frontier AI labs (Anthropic, OpenAI, Google DeepMind) that face increasing regulatory scrutiny. The piece provides them with a permission structure to continue unrestricted development. The ultimate beneficiaries are the investors and shareholders in these firms, who would see returns from unregulated AI commercialization and potential superintelligence breakthroughs. The losers are workers facing AI-driven displacement, the broader public facing safety risks, and democratic governance itself, which the piece frames as a confiscatory threat rather than as a legitimate risk-management institution.

There is a second-order cui bono: the Wall Street Journal’s editorial page itself benefits from maintaining its consistent anti-regulation brand identity. Every technology regulation fight — climate, finance, pharma, now AI — gets the same “private sector self-regulates, government seizes” treatment. The consistency is the product; the product retains the trust of an audience that expects to hear that regulation is always the problem. The special interests Jenkins gestures at as a problem are, in this case, his own employers’ institutional positioning.

Receipts

Direct quotations from the original WSJ column with analysis of each manipulative mechanism:

  1. “Maybe artificial superintelligence is just around the corner after all. How else to explain those tech CEOs who usually can’t get enough change suddenly wanting it throttled?” — Operator: Dismissive Reframe. Assumes CEOs’ concern is about “change” rather than specific competitive threats.

  2. “Never mind that their complaints have the paradoxical flavor of the old Borscht Belt joke: The food was no good and the portions were too small (Alex Karp of Palantir).” — Operator: Joke-as-dismissal. Denigrates the speaker, not the argument.

  3. “Or, they say, industry leaders Anthropic and OpenAI are Trojan horses. They’re out to steal their customers’ businesses (Microsoft CEO Satya Nadella).” — Operator: Selectional Strawman. Renders a nuanced competitive concern as cartoon theft.

  4. “Not that the established tech CEOs are Luddites exactly, but their shareholders aren’t paying them to forgo any lever to protect their existing businesses against the AI tsunami.” — Operator: Concession-as-Indictment. Allows the writer to dismiss the CEOs’ concerns as mere shareholder protection.

  5. “A bandwagon unbravely beckons. It consists not only of college students booing AI and waddling seniors raging on Facebook against data centers.” — Operator: Physical caricature of opposition. Reduces opponents to contemptible stock figures.

  6. “If there’s any scenario in which a recursively self-improving AI establishes itself on the internet to terrorize computer users until the end of time and eventually take over part of the galaxy, it will be a Chinese open-weight AI.” — Operator: Speculative Certainty. A claim about a hypothetical scenario is stated as a guaranteed outcome (“it will be”), laundering speculation into directive force.

  7. “Naturally special interests are organizing. Lobbying is a legitimate activity, but so is making sure the national interest prevails.” — Operator: Frame Override. Concedes the existence of special interests, then subordinates them to an undefined “national interest” that happens to align with the writer’s preferred policy.

Technique Identification (Library Selection)

The piece’s techniques map to the following entries in the operator taxonomy:

  • Dismissive Reframe (NR §4.1): Reducing a specific concern to a universal category of complaint.
  • Selectional Strawman (Bad-Faith Catalog ID: strawman): Presenting the weakest version of an opponent’s position as the totality.
  • Concession-as-Indictment (WSJ §3.4): The “not X exactly, but X” structure — concedes the benign interpretation while delivering the damning one.
  • Two-Audience Address (WSJ §4.3): One sentence that delivers a populist contempt signal and a political anti-regulation frame to different readerships simultaneously.
  • Threat-Inflation Closer (WSJ §4.13): Escalating regulatory stakes to confiscation, making any government intervention seem like a prelude to state control.
  • China-Frame Escalation (WSJ §4.20 / NR §4.5): “Three-card monte” — dismiss CEOs, attack regulators, then reveal an external existential threat that justifies unconstrained domestic development.
  • Speculative Certainty (new catalog entry): Presenting a hypothetical outcome as a guaranteed future (“it will be”) to direct policy preference without the burden of proof.
  • Frame Override (WSJ closing move): Conceding the existence of special interests, then overriding them with an undefined higher-order “national interest” that happens to validate the writer’s position.

DEFCON Threat Assessment

A calibrated rhetorical threat assessment of the original WSJ column across seven severity tiers, from least to most destructive of honest public discourse:

DEFCON 5 — Informational (no issue): The piece accurately reports certain events (Anthropic’s Mythos holdback, OpenAI’s model escape). No violation of discursive norms.

DEFCON 4 — Framing Spin (standard op-ed behavior): The use of “paradoxical flavor” and “Borscht Belt joke” to dismiss CEOs. Within normal editorial range — every columnist frames.

DEFCON 3 — Loaded Language (elevated): “Waddling seniors raging on Facebook,” “lose-lose propositions” for labs. Affective language that distorts the emotional valence of the debate.

DEFCON 2 — Active Deception (serious): The selectional strawman of Nadella and Cook’s competitive concerns. The “not Luddites exactly” move that invites the reader to infer the opposite. These are not framing choices; they are argumentative misrepresentations.

DEFCON 1 — Systematic Propaganda (critical): The three-card monte that redirects the reader from regulatory scrutiny to Chinese AI as the only real threat, using speculative certainty. The structural pattern pushes a predetermined policy outcome — no regulation — through a sequence of rhetorical misdirections.

DEFCON 1+ — Institutional Weaponization (extreme): This column aligns with and reinforces a decades-long institutional pattern of the WSJ editorial board deploying identical arguments across all technology regulation domains. It is not a single column; it is an instance of a structural propaganda mechanism. The “national interest” frame is a standing instrument that has been applied to climate, finance, pharma, and now AI with no variation in the operator pattern.

DEFCON 1++ — Existential Threat to Discourse (profane): The piece’s final move — treating “special interests” as a problem while the editorial page itself is a special interest with a systematic anti-regulation agenda — is the kind of discursive self-excepting hypocrisy that makes democratic deliberation impossible. It is the rhetorical equivalent of a Fuck you, we got ours that the entire page has been running since the 1980s. The piece provides the intellectual cover for the very regulatory capture it claims to oppose, and the profanity here is earned because the scale of the deception when multiplied across decades is genuinely destructive of public reason. One column is nothing; an editorial board that has spent forty years convincing its readers that the only legitimate response to every externality-producing industry is to deregulate it further — that is not journalism. That is an institutional capture mechanism wearing a flag pin.

Attribution