President Trump right now introduced $92 billion in AI-related investments, certainly one of which is 10 new nuclear reactors to be constructed by Westinghouse in partnership with Google.
Building will start by 2030, interim CEO Dan Sumner instructed Trump right now at a convention at Carnegie Mellon College, CNBC experiences. However given the corporate’s historical past, that is simpler stated than executed.
The US has constructed simply three reactors up to now 30 years, in response to the US Vitality Info Administration (EIA). The primary is a 2016 undertaking in Tennessee. Two adopted, each on the similar web site in Georgia—one in 2023 and one in 2024. (The latter isn’t proven on the graph under, which incorporates information as much as 2023.)
Nuclear reactors constructed within the US from 1970 to 2023 (US Vitality Info Administratoin)
Each Georgia reactors are the identical Westinghouse mannequin the corporate plans to construct once more for its subsequent set of 10, known as the AP100. However that undertaking got here in $18 billion over finances and 7 years delayed. The federal authorities beneath President Obama issued an $8.33 billion mortgage to assist assure its completion, however Westinghouse nonetheless filed for chapter in 2017 largely resulting from its troubles in Georgia, Reuters experiences.
What’s totally different this time? Westinghouse is throwing AI on the downside. It is partnering with Google Cloud AI streamline manufacturing, making it “an environment friendly, repeatable course of.”
“By partnering with Google Cloud…we will speed up the deployment of recent AP1000 models whereas implementing highly effective AI applied sciences that may optimize the development and operations of nuclear energy vegetation,” says Sumner. The CEO says the nuclear trade is “reinvigorated” after Trump issued 4 govt orders to advertise the trade in Might.
Westinghouse AP100 reactors at Plant Vogtle in Georgia (Credit score: US Division of Vitality)
It stays to be seen if Google’s AI can save Westinghouse’s nuclear reactor enterprise, however the firm says it is already “efficiently achieved a first-of-its-kind proof of idea” utilizing Google Cloud tech. It is utilizing three Google companies: Vertex AI (a machine studying platform), Gemini (Google’s flagship massive language mannequin), and BigQuery (a cloud-based information warehouse). Collectively, they’re able to “autonomously generate and optimize AP100” building plans, Westinghouse says.
Beneficial by Our Editors
However utilizing AI to assemble nuclear reactors may be a recipe for catastrophe. Hallucinations stay a cussed subject with massive language fashions, so Westingouse might want to double (triple?) examine each output. Any points with these tasks might tank the trade, which has fallen out of favor after the disasters at Chernobyl (1986), Three Mile Island (1979), and Fukushima (2011). Might AI make issues worse, or create a black field through which even Westinghouse is not fairly positive how the reactor was constructed?
The tech trade is prepared to take the danger in its determined quest for extra vitality to energy its AI information facilities. Microsoft is reviving the Three Mile Island plant to energy its AI ambitions. Meta CEO Mark Zuckerberg this week introduced plans to create a number of information facilities to attain “superintelligence.” Nuclear energy might be a gorgeous, environmentally pleasant choice—if it really works.
Get Our Finest Tales!
Your Day by day Dose of Our Prime Tech Information
Join our What’s New Now e-newsletter to obtain the most recent information, finest new merchandise, and knowledgeable recommendation from the editors of PCMag.
Join our What’s New Now e-newsletter to obtain the most recent information, finest new merchandise, and knowledgeable recommendation from the editors of PCMag.
By clicking Signal Me Up, you verify you’re 16+ and conform to our Phrases of Use and Privateness Coverage.
Thanks for signing up!
Your subscription has been confirmed. Regulate your inbox!
About Emily Forlini
Senior Reporter
I am the knowledgeable at PCMag for all issues electrical automobiles and AI. I’ve written a whole lot of articles on these subjects, together with product critiques, day by day information, CEO interviews, and deeply reported options. I additionally cowl different subjects inside the tech trade, protecting a pulse on what applied sciences are coming down the pipe that would form how we stay and work.
Learn Emily’s full bio
