Artificial intelligence is all the rage these days, and so is its energy consumption. From a game-changing technology that would usher in a new era, AI has recently turned into a problem for many—a problem that makes electricity more expensive while, apparently, not really living up to the promise of its developers. Big Tech is still pledging tens of billions in AI spend. Now, it may have to spend a bigger portion of those billions on securing its own energy, and on proving its worth the money that AI users would have to spend on the technology.
A growing number of commentators have been questioning the point of embedding AI into corporate operations recently. Bloomberg analysts this week argued the whole AI story shows signs of overhyping; Uber’s chief operating officer admitted the company’s investment in the technology has not led to the expected productivity gains. Meanwhile, cities are banning data centers across the United States. Because they are pushing up electricity prices, using a lot of water, and threatening the quality of life in the areas where they are built, according to a recent Gallup poll.
The energy issue seems to be particularly sensitive. The energy consumption of data centers housing artificial intelligence tech is notorious already. Analysts compare this consumption to whole nations. Yet some have argued that the higher electricity prices in some data center-heavy locations are not directly related to their consumption of electricity but rather to the local energy mix. If that mix features a lot of wind and/or solar, bills tend to be higher—as is the case in Europe—and the presence of data centers simply aggravates an already existing problem. Related: The Strait of Hormuz May Reopen, But the System Has Already Broken
This is not something that local communities appear to care much about, and with good reason. The International Energy Agency said in a report from this April that electricity demand from data centers generally last year jumped by 17% from the year before, with demand from AI-hosting data centers specifically rising even more substantially. The report noted that while the energy consumption per AI task is in decline, widening use of AI is offsetting this decline, and then some. In short, AI is an energy drain.
Wood Mackenzie this week reported that political opposition to artificial intelligence developers is on the rise across party lines, as politicians take up local communities’ concerns with their respective legislative authorities, the focus being on making Big Tech pay out of its own pocket for the investments necessary to secure the electricity and transmission infrastructure the data centers need.
“Growth in electricity demand, including demand from new data centres, has so far not had much impact on US power prices,” Wood Mac pointed out in its report. “However, it is starting to become a significant factor in some areas, including on the PJM grid that stretches from New Jersey to Tennessee.” It seems that whether the fears of higher electricity prices have a grounding in actual prices is irrelevant—those fears are driving action against data centers. Grid constraints are not helping.
The Wall Street Journal wrote this week that Big Tech is falling behind schedule with its new data center plans, because of “supply chain backlogs, permitting fights and availability of power supplies,” among other factors holding back the data center boom. The availability of electricity supply is directly related to the price of that electricity. Securing the necessary electricity supply for consumers as huge as data centers us also related to the price of electricity: PJM Interconnection last month said it would need to invest an additional $23.1 billion across the area it operates, which is the largest in the U.S.. Those billions will have to be “shared”, so to speak, the question being who they would be shared with—data center operators only or everyone in PJM’s area.
Big Tech, meanwhile, is addressing its growing electricity consumption problem in a way suggested by some in political and grid operation circles: generate your own electricity. Indeed, Big Tech is adding billions of dollars to its spending plans in order to build its own generating facilities. Of course, this is sparking doubts about the profitability of the whole AI rush, but it appears to be the most realistic approach to dealing with at least some of the local community and political opposition to the technology.
As for the question of what these generating facilities will be powered by, the answer, to the chagrin of many, would be oil and gas—in the U.S., mostly gas, but also nuclear. Without them, fears of an electricity crunch might become a reality, and the prices that go with an electricity crunch. It is doubtful anyone in a position of decision-making power would allow this to happen. And this means Big Tech would need to “bring its own electricity”.
By Irina Slav for Oilprice.com
More Top Reads From Oilprice.com
- Pakistan Turns to Russia and Venezuela as Middle East Oil Supplies Shrink
- India Pushes Refiners To Boost LPG Output
- Trump Extends Jones Act Shipping Waiver Through August