Google has announced that it’s helping to financially support the electrical training ALLIANCe (etA), an organization formed by the National Electrical Contractors Association and the International Brotherhood of Electricians. The goal is to train “100,000 electrical workers and 30,000 new apprentices in the United States” to meet the growing power demands of AI.
Using AI will unlock unspecified, but positive economic opportunities, Google’s new white paper, “Powering a New Era of American Innovation,” claims. In order to take advantage of them, though, the US power grid needs to become more capable and efficient. That’s largely because the data centers used to run and train AI models require vast amounts of energy. The white paper claims that new data centers could demand an additional “15-90 GW” of energy by 2030, something that more efficient chips and model training can’t account for. For a sense of the scale, the US Department of Energy says 1 Gigawatt is the equivalent to 103 offshore wind turbines.
Google’s paper calls for investments in alternative energy sources like nuclear power, but also notes that expanding the electrical workforce is necessary. “McKinsey estimates that 130,000 additional electricians will be needed by 2030 to build out data centers and manufacturing facilities,” the company writes. Currently, though, retiring electricians outnumber newly trained ones. “Nearly 10,000 American electricians either retire or change careers each year, while only about 7,000 new entrants join the field.”
Investing in electrical training is Google’s attempt to help change that. It would be nice if it was paired with a clearer explanation around what that AI will be doing with all that extra power — Google notes that there’s a “causal relationship between energy consumption and economic growth” — but skilled job training isn’t a bad thing.
This article originally appeared on Engadget at https://www.engadget.com/big-tech/google-is-funding-electrician-training-to-help-meet-the-power-demands-of-ai-221320678.html?src=rss
