Critique: Workers Have a Secret Weapon Against the AI Build-Out

Source: jacobin
Authors: ByBen Carroll
URL: https://jacobin.com/2026/05/workers-ai-power-plants-south

What the article reports

The piece argues that concentrated domestic manufacturing of gas turbines and large power transformers — clustered in the Carolinas, Virginia, and Georgia — gives workers structural leverage over the AI data-center build-out. It documents the scale of energy-infrastructure investment by companies such as GE Vernova, Siemens Energy, and Hitachi Energy, cites demand-growth figures from industry analyst IIR, and describes organizing efforts by the Southern Workers Assembly (SWA), where the author works, including a "Rank and File Program" placing worker organizers in strategic plants.

Factual accuracy — Mixed

Several figures are specific and traceable. GE Vernova's "25 percent of the world's electricity" and "55 percent in the United States" are publicly available company claims; the $2.4 billion Q1 figure for data-center-related orders matches GE Vernova's reported earnings. The Siemens $1 billion investment announcement and the Hitachi Energy $37 million South Boston, Virginia expansion appear in press releases from early 2025. Jensen Huang's energy-bottleneck quote is attributed to a "late 2025" Joe Rogan interview, which is plausible but unverifiable from the text alone.

Two claims warrant scrutiny. First, "AI investment propped up an otherwise anemic economy, accounting for anywhere from half to 75 percent of GDP in the first quarter of this year" is an extraordinary macroeconomic assertion. GDP share figures of that magnitude would be front-page news globally; no source is cited, and no qualifier is offered. This is the article's most significant unsubstantiated claim. Second, the Dallas Fed statistic — "workers age 22 to 25 in the most AI-exposed occupations have experienced a 13 percent decline in employment since 2022" — is attributed correctly in substance, though the article does not note that "most AI-exposed occupations" is a defined subset, not the youth labor market overall, which risks misleading the reader. The Memphis pollution comparison (Colossus turbines vs. Memphis International Airport) is attributed to Memphis Community Against Pollution, an advocacy group, without independent verification or methodology.

Framing — Advocacy

1. "Left to their own devices, the likes of Altman and the other titans of the tech billionaire elite would like nothing more than to remake society into one where all else is subservient to their profit margins." This is an authorial characterization of motive, not a reported fact. No attribution is offered; it is stated as settled truth.

2. "buy up, hollow out, and implement AI" — The verb sequence "hollow out" is an editorial judgment inserted into a description of Jeff Bezos's Project Prometheus investment fund. The fund's own language ("manufacturing transformation vehicle") is quoted but then immediately reframed with connotation-heavy verbs.

3. "Elon Musk announced that he would build the largest supercomputing facility in the world in Southwest Memphis — the facility, called Colossus, was built with no public input whatsoever." The phrase "no public input whatsoever" is an absolute claim embedded in the news summary without sourcing or acknowledgment of any permitting process that may have occurred.

4. "the right-wing billionaire class" and "titans of the tech billionaire elite" — These are political characterizations deployed as descriptive nouns. An article that scores well on neutrality would either attribute such framings or reserve them for a labeled opinion section.

5. "With the stage set for a general strike on May Day 2028" — This is stated as established fact ("the stage set"), not as an aspiration of the SWA or any coalition. It presents an organizing goal as an imminent, agreed-upon event.

Source balance

Ratio: Voices framed critically toward the AI build-out (SWA, Memphis Community Against Pollution, the article's own authorial voice) heavily outweigh any substantive defense or alternative perspective. No hyperscaler, no economist skeptical of the supply-crunch thesis, no worker who favors the jobs these plants create, and no union official outside the author's own organization is quoted. The CEO quotes are used as scene-setting villains, not as genuine perspectives explored on their own terms. Effective ratio: approximately 5:1 critical-to-supportive among substantive voices.

Omissions

1. Author's organizational affiliation is disclosed but minimized. The SWA is mentioned once, near the end, in a subordinate clause: "where I serve as organizing coordinator." This is the most important disclosure in the piece — the author is describing an organizing campaign he runs — and it appears without any structural prominence (no author bio, no conflict-of-interest note at top).

2. No union perspective beyond SWA. The ILA 2024 strike is mentioned, but no ILA official, no AFL-CIO representative, and no mainstream labor economist is quoted. The reader cannot assess whether SWA's "general strike on May Day 2028" framing is a widely held goal or a fringe aspiration.

3. The AI-as-GDP-share claim goes unsourced. "Anywhere from half to 75 percent of GDP in the first quarter" is a dramatic figure that requires sourcing. Its omission is a material gap.

4. Worker perspectives from inside the targeted plants are absent. The piece argues workers at GE Vernova Greenville (2,500 workers) and Siemens Charlotte (1,500 workers) are a strategic lever — but no worker at any of these facilities is quoted, even anonymously. The reader cannot assess whether workers at these plants share the author's analysis.

5. Historical labor precedent is invoked selectively. The 1930s–40s organizing waves are cited as inspiration, but the many failed industrial-organizing campaigns of the post-1970s period — and the structural differences between that era and today's right-to-work South — are not acknowledged.

6. The environmental case for gas turbines is absent. The piece frames gas turbines solely as pollution sources, without noting that they are also enabling renewable-energy integration and grid stability. A reader wanting to understand the energy tradeoffs would not find that context here.

What it does well

• Specific industrial geography. The piece names precise plant locations and headcounts — "GE Vernova's main gas turbine production facility is located in Greenville, South Carolina, with over 2,500 workers" — giving readers verifiable, actionable detail unusual in labor advocacy writing.
• Demand-side data is well-marshaled. The IIR figures on load growth ("risen from roughly 23 GW of new load in 2023 to about 42 GW today") and the turbine-order-backlog explanation give the supply-crunch argument a grounded empirical base.
• Structural argument is coherent. The logic — concentrated manufacturing + capital intensity + long lead times = worker leverage — is laid out clearly and is internally consistent, even if the counterarguments are not addressed.
• Transparency about activist purpose. Unlike some advocacy journalism that disguises its orientation, the piece does eventually disclose "where I serve as organizing coordinator," and Jacobin's left-labor identity is apparent to regular readers of the outlet.
• The Altman 2021 quote ("Even more power will shift from labor to capital") is a genuine primary-source admission that strengthens the piece's central concern, and its use is fair.

Rating

Overall: 4/10 — A substantively informed but openly advocacy-driven piece that presents one side of a labor-strategy debate as reportage, with thin source diversity and unattributed interpretive framing throughout.