Workers have a secret weapon against the AI build-out - The Real News Network

TEXT START: "We're really going into what we believe is the early chapters of an investment supercycle in the US for electricity growth," Scott Strazik, CEO of GE Vernova, told Barron's during an interview at the World Economic Forum in Davos earlier this year.

THE DISSECTION

This piece argues that workers in gas turbine and large power transformer manufacturing occupy a "strategic chokepoint" capable of slowing or leveraging the AI build-out. It frames this as a practical organizing opportunity, draws historical parallels to 1930s labor upsurge dynamics, and positions the Southern Workers Assembly's infiltration-and-committee-building model as the tactical solution. The tone is militant, the analysis is earnest, and the optimism is structurally unearned.

THE CORE FALLACY

The article mistakes a lag for a leverage point.

The workers at GE Vernova in Greenville, Siemens Energy in Charlotte, and Mitsubishi in Savannah are not sitting on a secret weapon. They are sitting on a temporary chokepoint in a system that is actively engineering around every chokepoint. The piece correctly identifies that these facilities produce equipment with multi-year order backlogs and no viable near-term substitutes. It incorrectly treats this as a stable structural condition rather than a problem that capital is spending billions to solve. Jensen Huang himself saying "energy is the bottleneck" is not a gift to labor—it is a signal that every energy company, AI developer, and sovereign fund is racing to un-bottleneck it.

The Discontinuity Thesis does not say these workers have no leverage. It says the leverage is time-limited and structurally bounded—and that even a successful strike only buys time, it does not rewire the underlying mechanism of displacement.

HIDDEN ASSUMPTIONS

1. That labor action can meaningfully alter the pace of AI deployment. Even a total halt in US gas turbine production delays the build-out by years at most. Meanwhile: SMR development accelerates, battery storage scales, nuclear plants get permitted, and hyperscalers sign long-term power purchase agreements elsewhere. The AI economy does not need these specific workers—it needs energy, and energy has multiple supply paths.

2. That historical parallels to the 1930s hold. The CIO upsurge succeeded because manufacturing was intrinsically labor-intensive and scaled through human work. The capitalist class needed these workers in a way that has no structural equivalent today. AI capital does not need gas turbine workers to produce AI. It needs energy. The workers at these facilities are important to the transition, but not indispensable in the way auto workers or steelworkers were in 1936.

3. That the "hardware bottleneck" is a stable feature. It is a transient constraint being actively engineered away. GE Vernova is booking orders to 2030 not because the capacity is permanently limited—it's because capital is deciding to build capacity at a pace that suits its return-on-investment timeline, not labor's organizing timeline. When the pressure gets high enough, new facilities get permitted, automation gets fast-tracked, and the production timeline compresses. These companies have every incentive to break this bottleneck.

4. That organizing workers into strategic sectors is the primary variable. The piece treats this as a straightforward tactical problem: get organizers in place, build committees, link to May Day 2028. It ignores that the political economy of the moment—Trump administration ICE deployments in Memphis, redistricting to strip black political representation, class war in open form—means these workers are entering the fight under the worst possible institutional conditions.

SOCIAL FUNCTION

Lullaby for militants. This is a well-researched, earnest piece that offers genuine tactical information (the geographic concentration of transformer and turbine production, the SWA's organizing model, the potential leverage calculation) but wraps it in a narrative of power that the structural mechanics do not support. It tells activists they have found the Achilles heel of AI capital when they have actually found a symptom of the transition—the scramble for energy is itself evidence that the underlying displacement is accelerating, not decelerable.

Prestige signaling within left-labor circles. The Jacobin-by-way-of-Real-News-Network publication context frames this squarely in a certain strand of socialist labor organizing discourse. The piece does real intellectual labor in mapping the supply chain. But it performs militant optimism rather than delivering a clear-eyed assessment of what this leverage can and cannot accomplish.

THE VERDICT

The article identifies a real tactical opportunity—geographically concentrated, capital-intensive production that the AI build-out cannot rapidly substitute—and then dramatically overstates its significance as a lever of power. The workers in Greenville, Charlotte, and Savannah are not the new coal miners holding the grid hostage. They are a temporary chokepoint in a system actively working to eliminate chokepoints. Every month they delay the build-out is meaningful. But "significant leverage" is not the same as "transformative power," and conflating the two is how organizing movements burn themselves out on partial victories while the structural displacement continues.

The Discontinuity Thesis says: organize these workers, use the leverage, buy time for transition work—but do not mistake a lag defense for a survival strategy. The AI build-out is not stopped by gas turbine backlogs. It reroutes around them.