Midterms: a rejection of AI or a punishment of MAGA? Commentators split

Democrats won big in the US midterm elections, and American commentators are split two ways over what it means. Writers who have long criticized the tech industry say voters rejected AI, data centers and tech billionaires. Economic commentators say voters simply punished MAGA, President Trump's movement, and that AI was not a central issue. Neither side has squarely answered the other. And the AI accelerationists, who argue that development should move faster and who spent money on the election and lost, have yet to explain anything.

For Japanese companies: Which reading wins out in Congress will shape who pays for the electricity used by US data centers and how strict AI regulation becomes. Companies with investments in US data center or power businesses, and companies whose work depends on US AI services, should watch what Congress does after it returns on November 9.

The "AI was rejected" camp

Brian Merchant, a journalist who covers and criticizes the tech industry, wrote "A generative election" in his newsletter Blood in the Machine. The title plays on generative AI. In it, he casts the election as a verdict on AI, on the data centers that use up electricity, water and land, and on the tech billionaires pushing them forward.

Shakeel Hashim, editor of the AI policy newsletter Transformer, took a similar view, writing about what the Democrats' "blue tsunami" means for AI policy and legislation. Tech Policy Press, a policy publication, ran an essay arguing that voters had handed Congress "a mandate to confront corporate power and technofascism." Technofascism is a critical term for an alliance between tech billionaires and authoritarian politics.

What this camp shares is the view that data center construction and the AI industry's political donations reached voters as local, everyday issues.

The "MAGA was punished" camp

Economics blogger Noah Smith put a very different reading in the title itself.

In other words, the winner was not AI criticism but ordinary voters who had turned away from MAGA. On this reading, the result was a very traditional punishment of the party in power over prices and political turmoil. Tyler Cowen, an economist at George Mason University who writes the blog Marginal Revolution, also did not treat the result as a verdict on AI, reading it instead as a story about politics.

Talking past each other

What stands out is that the two camps are not actually debating. The AI-rejection camp has not rebutted the MAGA-punishment explanation, and the MAGA-punishment camp has not said how it accounts for districts where data centers became an issue. Both are looking at the same result and using it to back up what they were already saying.

The truth may not be one or the other but both. One possibility is that worries about data centers and AI were layered on top of existing discontent with MAGA. Sorting out how much each factor mattered would require laying out the numbers district by district. So far, no essay appears to have done that work.

The accelerationists who lost are silent

The side that most owes an explanation is the side that lost. Leading the Future, a political group funded by large donations from the AI industry, and others who have opposed regulation and called for faster development poured a lot of money into the election. That side has not yet explained why it lost or what it misread.

The silence means something. If this was a punishment of MAGA, the accelerationists could say "AI had nothing to do with it." Their silence can be read either as reluctance to admit the loss was a policy failure or as a sign that they have nothing to argue back with.

Congress will give the answer

Which reading is right will be tested not in essays but in Congress and the states in the months ahead. The new Congress returns around November 9. The first test will be whether AI bills are introduced, such as House Democrat Greg Casar's proposal to ban the development of superintelligence (AI that far surpasses humans). In Texas, Hinojosa was elected governor, and attention will turn to how 13.3GW of planned data centers is handled.

AI companies are already moving. Axios reported that after the election, OpenAI will begin lobbying for a mandatory "speed limit" on the most powerful AI. The company is also reported to be rebranding data centers as "AI compute campuses" and backing agreements between states and local communities. If the AI-rejection reading gains traction in Congress, these industry concessions will spread and local conditions for data centers will get tougher. If the MAGA-punishment reading prevails, AI regulation will proceed on its own track, separate from this election.

In this writer's view, both camps are reading into the result what they want to see. What will settle it are the first bills introduced in Congress and the decisions made in the communities where data centers are planned.