I'm not sure how you arrive at this conclusion. If you read the paper being criticized, the authors clearly intend to claim that heat exhaust is the cause.

> "The electronic components in AI hyperscalers are characterized by extremely high-power densities, often reaching magnitudes on the order of 107 W/m^2

That's an actual passage from the paper. Masley shows that AI, data centers, and computation writ large have nothing to do with the effects measured. All they discovered was that parking lots and roofs were hot. None of your priors, or mine, or any of the authors, matter here. If Masley is correct, the paper is junk science.