Tech oligarchs bet on Trump and lost: the 2026 midterms were a democratic rebuke of big tech, but the industry will not go quietly

Source: https://www.theguardian.com/profile/jonathan-liew. "Elon Musk and the tech bros banked on Trump and lost. Humbled, humiliated: what will they do next? | Jonathan Liew | The Guardian." November 9, 2026. www.theguardian.com

The Gist

The author argues that tech billionaires bet heavily on Trump and the hard right and got badly beaten in the midterms, as voters rejected datacentres and the politicians backing them. That should make us doubt their reputation as visionaries. But the author warns the industry has plenty of money and influence and will likely keep resisting democratic oversight.

Conclusion

The 2026 midterm results were a rebuke not just of Trump but of the Silicon Valley oligarchs' ideology, which should prompt a sober reassessment of their supposed visionary status. Big tech is losing its grip, but this setback is unlikely to end its influence, and it may respond by pursuing less democratic accountability.

Premises

1. Tech elites such as Musk and Huang aligned themselves closely with Trump and envisioned an AI industry free of regulation and safeguards, yet failed to foresee the electoral defeat of the unpopular leader they backed.
2. Voters rejected big tech and the politicians who enabled it: a Democrat won Texas on a datacentre moratorium despite being outspent roughly eight to one by tech donors, a Democratic challenger in Ohio successfully branded his opponent 'the face of datacentres', and 46 of 47 datacentre-related ballot measures were defeated.
   • Needed for the conclusion
   • Unresolved objection: The ballot-measure count may reflect the content or sponsorship of the measures, not opposition to the industry, and no baseline is given.
   • This is the main direct evidence for voter rejection of tech, but it is ambiguous and lacks a baseline.
3. Musk spent nearly half a billion dollars trying to tip the electoral scales toward the hard right and failed.
4. Financial markets reacted to the progressive landslide with a sharp sell-off in tech stocks, just as Brazil's stock exchange fell after Lula's victory, suggesting capital is uneasy when hard-right autocrats are repudiated.
5. AI and datacentre expansion advance through the undemocratic manufacture of consent and circumvention of scrutiny; for example, the UK's AI growth zones removed planning decisions from local councils, ministers met almost no community or environmental groups, and officials moved into AI company jobs.
   • Needed for the conclusion
   • The process facts about planning and access are plausible, though they are UK-specific and do not show how industry will respond to this setback.
6. Tech and political elites are increasingly funding the hard right (e.g. Reform UK's £72m from crypto billionaires), and Goldman Sachs expects datacentre moratoria to be short-lived, suggesting industry will keep pushing back.
   • Needed for the conclusion
   • Large donations and an analyst forecast support continued pushback, but not that the pushback will be less democratic.

Assumptions

• The midterm results reflect voter opposition to big tech and its agenda, not merely partisan or economic factors.
  • Needed for the conclusion
  • Whether the results reflect opposition to tech and not only partisan or economic factors is the central unresolved question.
  • Would be settled by: Exit polls or surveys linking vote choice to views on tech and datacentres, compared against the normal midterm swing and economic drivers.
• Datacentre ballot measure outcomes and campaign messages are representative of broader public sentiment toward the tech industry.
  • Needed for the conclusion
  • Datacentre opposition is often about local costs, so it may not stand in for views on the industry or its ideology.
  • Would be settled by: Evidence on what drove the datacentre votes (costs, land use or attitudes to tech) and whether those contests were representative.
• Market sell-offs are caused by the election result rather than other economic factors.
• Tech billionaires' political spending is driven by a desire to avoid regulation and preserve tax breaks.
• Failure to predict an election outcome is relevant evidence against claims of visionary status.
• Electoral setbacks for industry will lead to evasion of democratic accountability rather than accommodation.
  • Needed for the conclusion
  • The prediction of evasion is plausible but is asserted and not shown, and accommodation is an equally available response.
  • Would be settled by: Tracking of post-election industry behaviour against clear indicators separating accommodation and ordinary advocacy from bypassing scrutiny.
• Small, local victories can be generalized to a shift in national political power over big tech.
  • Needed for the conclusion
  • Generalising from a few local results to national political power over the industry is not demonstrated.
  • Would be settled by: A representative sample of races and measures across the country showing a consistent tech-specific pattern.

Analysis

Reasoning breaks down

A step the conclusion needs doesn't follow. This says nothing about whether it is true.

Argument type: Inductive.

Potential Fallacies

• Hasty generalization (Premise 2) — A few selected races and a ballot-measure tally are treated as a national verdict on the whole industry. Even if these results reflect anti-tech feeling (as A1, A2 and A7 allow), they concern datacentre siting. They do not show rejection of an entire worldview or a lasting loss of power.
• Equivocation / slide between meanings (Conclusion) — Opposition to local datacentres, defeat of Trump-aligned donors, and rejection of a deregulatory ideology are treated as one thing. Each needs separate evidence, and the argument gives it only for the narrowest of the three.
• Post hoc / false cause (Premise 4) — A market sell-off after the election is attributed to the election's meaning, supported by a single foreign analogy. A sell-off could just as easily reflect expected tech regulation, interest rates, earnings or valuation worries. It also sits awkwardly with the claim that tech is losing its grip.
• Hindsight bias / irrelevant evidence (Premise 1) — Failing to foresee one election result is judged with knowledge of the outcome. It is also a different skill from the technological and commercial foresight that 'visionary' status concerns, so it offers only weak support for reassessing that status (A5).
• Cross-jurisdiction conflation (Premise 5) — UK planning reforms and UK party donations are used to forecast how the industry will react to a US election. The institutions, rules and actors differ, and the evidence predates the midterms.
• Guilt by association / blended villain (Premise 6) — Tech leaders with different motives, crypto donors, the hard right and Trump are bundled into one bloc, so criticism of one transfers to all. Lawful lobbying and donations are then read as evasion of accountability.
• Unfalsifiable framing (Conclusion) — If industry loses, it is a rebuke, and if it keeps influence, it will not go quietly. Accommodation could be reread as tactics. With no outcome that would count against the claim, it is hard to test.

Counterarguments

The step from a Democratic win and a few datacentre results to a rebuke of tech ideology is the heart of the conclusion, and nothing in the argument rules out the ordinary midterm swing or local-cost explanations. If it fails, the conclusion shrinks to a loss for Trump-aligned donors and some local datacentre resistance, and the reassessment and the forecast lose their trigger.

Without this step, nothing else in the argument reaches the conclusion.

• Strongest challenge: Argument as a whole (High impact) — The same evidence is expected from an ordinary midterm swing against the president's party, cost-of-living concerns, and local worries about power bills, water and land. Nothing in the argument separates a tech-specific rebuke from these simpler causes, and the figures are given with no baseline for comparison.
• Premise 2 (High impact) — The ballot-measure tally does not say whether the defeated measures were pro- or anti-datacentre, who sponsored them, or how often such measures fail anyway. Two named races are anecdotes picked from many, and party loyalty or candidate quality could have decided them.
• Assumption 6 (High impact) — Lobbying, donations and litigation are lawful democratic participation, and industry may instead accommodate by negotiating community benefits, funding grid upgrades or giving to both parties. The prediction of evasion rests on UK examples and a bank forecast, not on any observed response to this setback.
• Premise 4 (Medium impact) — A tech sell-off after the election fits expected regulation, rate moves or an AI valuation correction at least as well as unease at a repudiated autocrat. If markets expected tech to be hurt, that shows sector exposure, not voter opinion.
• Premise 1 (Medium impact) — Election forecasting is a different competence from technological or business vision. Judging the elites' foresight by one outcome is hindsight reasoning, and the same test would strip nearly every pundit and investor of credibility.
• Premise 3 (Low impact) — Campaign spending has small and uncertain effects on outcomes. A large spend that did not win shows limited marginal impact of money, not public rejection of an ideology, and any wins from the spending are not counted.

Suggested Improvements

• Premise 2: The ballot-measure statistic and the two named races are given without context, direction or baseline. — Name the measures and races, state what each asked and who sponsored it, and compare the results with typical outcomes for similar measures and races. Readers cannot tell whether the figures show opposition to tech or something unrelated, and may even read the direction wrongly.
• Assumption 1: The link from vote outcomes to opposition to tech is assumed without separating it from partisan and economic drivers. — Cite exit polls or surveys that connect vote choice to attitudes toward tech and datacentres, and compare against the usual midterm swing. A general anti-incumbent swing and cost-of-living concerns would produce similar results.
• Premise 4: The sell-off is tied to the election by a single analogy and given a contested interpretation. — Drop the market claim or support it with a comparison against market and sector benchmarks, and treat it as evidence of expected regulation, not of voter sentiment. Markets move for many reasons, and the reading offered sits in tension with the claim that tech is losing power.
• Premise 5: UK planning and funding examples are used to forecast US industry behaviour after a US election. — Show US examples of evasion of scrutiny after setbacks, or explain explicitly why the UK cases carry over. Different institutions and a different timeline weaken the connection.
• Assumption 6: Continued lobbying and funding is equated with seeking less democratic accountability. — Define what counts as evasion, as opposed to ordinary advocacy, and state what behaviour would count against the prediction. These are lawful forms of participation, and accommodation is also a plausible response.
• Premise 1: A missed election forecast is used to question visionary status. — Tie any reassessment of visionary status to technological or commercial claims, such as the track record on promised safeguards or capabilities. Political forecasting is a different skill from technological vision.

Scenario Tests

• Post-election surveys show that voters who gave datacentres or tech companies as a reason for their vote swung to the winning side by more than the usual midterm swing, even after controlling for party. (Supports) — This would separate a tech-specific effect from the ordinary anti-incumbent baseline and give the main claim the evidence it lacks.
• The 46 defeated ballot measures turn out to be anti-datacentre or moratorium proposals that voters rejected. (Challenges) — The evidence would run against the thesis and remove its most concrete data point.
• Tech companies respond by funding grid upgrades, signing community-benefit agreements and donating to both parties. (Challenges) — This would show accommodation rather than evasion and undercut the forecast, while still being consistent with the claim that influence persists.
• Tech firms lobby for federal preemption of local datacentre rules and expand special-zone planning after the midterms. (Supports) — This would give direct evidence for the predictive part of the conclusion, which is currently inferred from UK examples.
• Analysts attribute the tech sell-off mainly to rates and an AI valuation correction, and sector moves match the broader market. (Neutral) — The market premise would drop out, but the main case does not depend on it.

Coherence & Relevance

The argument tells a coherent story but its parts are of mixed kind (US elections, UK planning, Brazilian markets, Reform UK funding) and rest on tacit analogies. The load-bearing claim of a tech-specific rebuke has the thinnest support, while the more modest claim of persistent influence is plausible. The framing also makes almost any outcome confirm the thesis, which weakens its testability.

• Tech elites such as Musk and Huang aligned themselves closely with Trump and envisioned an AI industry free of regulation and safeguards, yet failed to foresee the electoral defeat of the unpopular leader they backed. (Weak) — Political forecasting is not the competence behind visionary standing, and a single miss is weak evidence.
• Voters rejected big tech and the politicians who enabled it: a Democrat won Texas on a datacentre moratorium despite being outspent roughly eight to one by tech donors, a Democratic challenger in Ohio successfully branded his opponent 'the face of datacentres', and 46 of 47 datacentre-related ballot measures were defeated. (Moderate) — Directly on point for datacentre backlash, but not for rejection of an ideology. The ballot-measure direction and baseline are missing.
• Musk spent nearly half a billion dollars trying to tip the electoral scales toward the hard right and failed. (Weak) — No counterfactual. Failed spending shows limited effect of money, not what voters think of tech, and it overlaps with the other premises as one election result.
• Financial markets reacted to the progressive landslide with a sharp sell-off in tech stocks, just as Brazil's stock exchange fell after Lula's victory, suggesting capital is uneasy when hard-right autocrats are repudiated. (Weak) — No control for other causes of the sell-off. The interpretation pulls against the 'losing its grip' claim.
• AI and datacentre expansion advance through the undemocratic manufacture of consent and circumvention of scrutiny; for example, the UK's AI growth zones removed planning decisions from local councils, ministers met almost no community or environmental groups, and officials moved into AI company jobs. (Moderate) — Relevant to the concern about evasion of scrutiny, but UK-specific and pre-dating the midterms. It does not show a response to this setback.
• Tech and political elites are increasingly funding the hard right (e.g. Reform UK's £72m from crypto billionaires), and Goldman Sachs expects datacentre moratoria to be short-lived, suggesting industry will keep pushing back. (Moderate) — Supports continued pushback but not that the pushback will be less democratic, and the Goldman item is a forecast, not conduct.

View this argument on LogicFirst.ai