7 Comments
User's avatar
Josh_Redwood's avatar

Thanks for this Michael! It's the best reporting on the Great AI Data Center Power Freakout that I've read so far. I love that you ended with utilities using data center-driven load growth as premise for more fossil fuel infrastructure investments--no surprise there. I'm hoping that one utility taking the bold step of meeting projected demand growth from data centers with *only* efficiency/VPP/etc. investments will get other doing the same. Great broad and deep coverage here, please revisit this story often for us!

Expand full comment
Michael Thomas's avatar

Thanks for saying that, Josh! It would be awesome to see more examples utilities could follow.

Expand full comment
Ben Fenton's avatar

great read thanks Michael…the incentives are quite perverse from a regulated utility standpoint as you highlight multiple times

Expand full comment
Michael Thomas's avatar

Thanks for reading, Ben!

Expand full comment
Max Dugan-Knight's avatar

This news may have broken after you finished writing, but curious if you have thought about how DeepSeek's innovation in model training plays into this. My sense is they have gone some way to revolutionizing cost-effectiveness, energy efficiency, and chip usage and their models are competitive with cutting edge US ones. Nvidia and energy stocks appear to be falling as a result.

Thanks for great article. I always had a hunch that projections of AI energy demand were overblown.

Expand full comment
Michael Thomas's avatar

Thanks for the comment, Max! This is a really interesting question. Jonathan Koomey, who I mention in this story, has argued that efficiency gains like this are what have prevented computing from blowing up the grid in the past. Good interview with him on that history here: https://heatmap.news/podcast/shift-key-episode-10-ai-data-centers

Seems a little early for anyone to make any accurate predictions based on the DeepSeek news. But definitely something I'll be watching.

Thanks for reading.

Expand full comment
Dr Sugandha Srivastav's avatar

The issue here is that rate of return regulation, which is still oddly prevalent in the US, creates severely mis-algined incentives and allows the transfer of risk from the utilities to ratepayers as you note. This isn't how utilities work in most of Europe though! It's a bit baffling why the US hasn't moved over to a system where utilities function like any other normal firm that has to think about the real profitability of investments

Expand full comment