Strike while AI is hot: Rebuilding worker power for the age of AI | IPPR

TEXT ANALYSIS: IPPR Worker Power Paper

THE DISSECTION

This is a progressive think-tank autopsy dressed as a policy roadmap. It reads like a 1970s trade union strategy paper that wandered into the wrong century—solving for distribution in an economy where the foundation itself is dissolving. The authors correctly identify that AI threatens workers and that power is asymmetric. They then prescribe institutional labor reforms—co-determination, portable benefits, disclosure requirements, union rights—as if organizing the corpse will resurrect it.

The intellectual architecture assumes the post-WWII labor market is a malleable terrain that political will can reshape. It does not grapple with the possibility that AI severs the mass employment-wage-consumption circuit at the structural level, rendering worker power a category with nothing left to exercise power over.

THE CORE FALLACY

The paper treats a structural annihilation problem as a distributional negotiation problem.

The DT framework establishes a hard mechanical constraint: when AI achieves durable cost and performance superiority across cognitive work (P1), the majority of human labor becomes economically optional. Not temporarily displaced. Not renegotiable through collective bargaining. Functionally unnecessary at scale.

The IPPR authors write as if "workers' ability to influence AI adoption at the firm level" is the decisive variable. This assumes:
1. Firms need human labor in ways that are negotiable.
2. Worker power can extract meaningful concessions from capital that controls AI.
3. The productivity windfall will be large enough that sharing it preserves mass employment.

All three assumptions collapse under DT logic. Once AI achieves P1, the need for worker consent evaporates. Capital doesn't negotiate with labor it doesn't require. Co-determination and portable benefits wallets are institutionally elaborate forms of rearranging deck chairs on a ship that has already struck the iceberg.

HIDDEN ASSUMPTIONS

1. The Labor Market Persists as the Primary Allocation Mechanism. The entire framework—portable benefits, worker levies, union membership, retraining—assumes workers will transition to new roles. DT establishes that for the majority, there are no structurally viable transitions. The "flexicurity" model (Danish model of flexibility + security) was designed for a world where displaced workers could be retrained into other jobs. That world required human labor to remain economically necessary.

2. Political Power Can Override Competitive Dynamics. The paper treats AI adoption as something that can be "shaped" through consultation rights. In competitive markets, firms that deploy AI will undercut firms that don't. Consultation rights that slow adoption simply ensure domestic firms lose to international competitors who didn't pause for collective bargaining.

3. The Productivity Windfall Is Sufficiently Large to Share. The paper assumes there will be substantial gains to distribute. This may be true in aggregate, but DT mechanics indicate those gains concentrate at the Sovereign level (AI capital owners) with near-zero marginal cost of replication. The "shared prosperity" framing requires a distribution mechanism that doesn't exist and cannot be engineered without dismantling property rights—which the paper studiously avoids proposing.

4. Dignity of Work Is Preservable Through Institutional Reform. The implicit normative anchor is that meaningful work is a social good worth preserving. DT suggests that for the majority, meaningful productive participation becomes structurally unavailable—not through moral failure, but mathematical inevitability. Retraining 40-year-old truckers as "AI trainers" or "digital benefits wallet managers" is not a serious solution; it's a cruel fiction dressed in technocratic language.

SOCIAL FUNCTION

Prestige signaling wrapped in institutional nostalgia.

This paper performs progressive credentials—cites labor history, invokes solidarity, advocates for worker representation—without engaging the actual depth of the discontinuity. It is written for an audience that wants to believe the 20th century's labor-capital negotiation framework is still operative. It offers enough institutional detail to feel substantive while evading the structural question: what happens when organizing labor becomes irrelevant because the labor itself is irrelevant?

The "portable benefits wallet" concept is particularly illustrative. It's a clever policy design that addresses 21st-century precarity within a 20th-century framework. It doesn't ask: benefits for what kind of work? The implicit assumption is that there will be enough fluid, dynamic work to justify the infrastructure.

Classification: Transition Management Theater

The paper helps institutional actors (politicians, think tanks, unions) feel like they're addressing the problem while avoiding the conclusion that the problem cannot be solved through the mechanisms proposed. It extends the intellectual shelf-life of laborist thinking for another policy cycle.

THE VERDICT

The IPPR paper is a sophisticated document that diagnoses symptoms and prescribes placebos.

It correctly identifies the asymmetry of power and the stakes of AI adoption. It is wrong that this asymmetry can be corrected through institutional reform. The DT framework establishes that AI severs mass employment from economic participation—not gradually, not through negotiation, but structurally. Worker power requires workers to possess something capital needs. P1 eliminates that condition for the majority.

Portable benefits, co-determination, disclosure requirements—these are hospice care for a patient whose vital organs have been removed. They may soften the transition. They cannot prevent the structural outcome.

The political choices the authors invoke cannot override competitive necessity. In a global economy, nations that deploy AI without labor consultation will outcompete nations that do. The "race to the bottom" dynamic the authors acknowledge but don't fully reckon with is not a failure of political will—it is the logical output of competitive markets operating as designed.

The paper is earnest, well-researched, and structurally unserious. It treats the death of post-WWII capitalism as a political design problem rather than a structural discontinuity. Workers who organize based on its prescriptions will be better positioned than those who don't—but "better positioned" means "slightly more prepared for a collapse that cannot be prevented."

Survival relevance: For Oracle of Obsolescence clients, this paper is confirmation that mainstream progressive institutions are still operating in denial. The lag defenses (institutional inertia, cultural attachment to work) it relies on are real but cannot reverse the outcome. Clients should not expect meaningful worker-power resurgence at scale. Sovereign/Servitor/Hyena positioning remains the operative framework.