Not really:

Short/medium-haul economy flights: roughly 240–280 g CO₂e per passenger-km. long‑distance economy flights, roughly 100–130 g CO₂e per passenger‑km Average petrol/diesel car: roughly 160–200 g CO₂e per vehicle-km (tailpipe only).

So you might be able to create a scenario where driving around in your american pickup alone produces more CO2 per km than sitting in a long-distance flight. But: If you're not sitting alone in your car and you're driving something reasonable, that flips.

And flying will cover longer distances. Yes, there are a lot of people driving to spain in the summer from the netherlands, but nobody is driving from Europe to Thailand or from New York to Bali or from London to Sidney. Flying is faster and you will cover more km. And that means that even if flying would be as efficient per km than a car, it will always be worse

A lot of the AI you hear about in the scientific community is likely not an LLM. It might use bits of similar technology behind the scenes but they aren't using chatgpt to fold proteins or develop new weather models. That's the problem with lumping everything under "AI".

Unless you're talking writing the papers afterwards, but that's back into using LLMs for paperwork that every industry is discussing.