When asked about how AI can reduce humanity’s existing and future energy demands, former Google CEO Eric Schmidt said during the Special Competitive Studies Project AI+Energy Summit that the demand for AI computing (this is its power requirement) is infinite and that the key point is “we’re not going to get there through conservation.”
The host then followed up with, “Do you think we can meet AI’s energy without total blowing out climate goals?” and Schmidt answered with, “We’re not going to hit the climate goals anyway because we’re not organized to do it — and the way to do it is with the ways that we’re talking about now — and yes, the needs in this area will be a problem. But I’d rather bet on AI solving the problem than constraining it and having the problem if you see my plan.”
It seems that the former Google chief favors dropping climate goals to ensure that AI companies will have enough power to drive their AI ambitions. Incidentally, this is happening now at Google, as its greenhouse gas emissions have jumped by 48% since 2019, primarily driven by its data center energy demands. Nevertheless, Schmidt recognizes our climate problem but believes that we shouldn’t let targets shackle AI development as we could use it to solve that problem. Besides, he says that we will not be able to meet the targets we’ve set anyway.
The host then closed the interview by asking Schmidt about his top three action steps that the industry needs to take now, and he said, “More power that is predictable, that is base load or base load-equivalent, as soon as they can. They [AI companies] need more places to site, they need more ways of getting things built, and they need more things to get them connected.” In return, Schmidt promises that these AI companies will make energy generation systems at least 15% more efficient or maybe even better, telling the audience that “that’s a lot of money for a utility.”
While we’re achieving many gains in our clean energy and energy efficiency goals, it sounds reckless to abandon our greenhouse gas emissions targets to push AI development to its peak. After all, English economist William Jevons observed in 1865 that steam engine improvements, which made coal use more efficient, did not reduce fuel use. Instead, these advancements, which made steam engines cheaper to operate, increased demand for coal even more.
This occurrence, called Jevon’s Paradox, could still happen today with energy consumption and AI. If AI could make energy production at least 15% more efficient, demand could increase as energy prices drop. Even Schmidt said it himself, “The economics will drive it anyway. No large company wants to have a huge power bill.” He adds, “Most of the people I’ve talked with say [the] power bill is becoming a very large component of their expenses.”
Furthermore, are we sure we should rest our future on AI decisions? Hollywood has shown us many times how leaving the fate of humanity in the hands of AI (or even another form of intelligence) might not be a good idea.