None of this will ever get stolen

The data in the plots comes from the U.S. Energy Information Administration. It includes quantitative comparison of lighting to computers. For residential users, in 2000, lighting completely dwarfed computers for power usage. In 2025 lighting was using only a fraction of the power it used in 2000, but still more than computers. The power usage of computers in residences grew from \~2005 to 2010, but then decreased substantially. I suspect this reflects the development of lower power processor technology (e.g., Intel's Core series; ARM); there have been dramatic gains in cycles per watt in the last 20+ years. I don't know what type of computer you are considering for your 1000 watt number. A fully loaded tower with a powerful graphics card will use that much when playing an intense computer game. But that's not a typical use case. The workstation I use at work, as a data scientist, has a 1000 W power supply, but it was overspecified. The fairly powerful GPU card I use draws \~ 250 W when it's churning (which it usually isn't). My new Mac Studio at home idles at 13 W and draws about 200 W in a hard workload (video rendering, say). But most home Apple users would have a Mini or an iMac, which use \~100 W at max load (as an Apple and Linux user, I don't know Windows PC numbers offhand). Most home computers now use laptop processors (Intel's Core series, or ARM processors) that draw low power. All this is in line with the US EIA's claim that computing is a smaller power draw than lighting, even modern LED lighting. By their numbers, TVs use more power than computers (and always have, since when the charts start, in 2000). Your comment about reduced manufacturing load is what really interested me, since it seems to make a lot of sense, yet it wasn't mentioned in the NYT article. The EIA numbers show that residential power use has been a bit larger than commercial use since 2000. They show that lighting was by far the largest component of commercial use in 2000, but now is second, behind ventilation. The reduction in power draw from lighting has been significantly smaller for the commercial sector vs. residential. Commercial electricity use overall has grown rather than shrunk since 2000, but not by much. But EIA's projections show it growing substantially over the next two decades, mainly due to data centers. So I don't know how to reconcile the shutdown of manufacturing facilities with the EIA numbers. Could it be that the shift from a manufacturing economy to a service economy roughly preserved electrical usage, just moving it from plants to office parks? "We connect larger electrical services to new houses than we did in 1980 and have twice as many houses." I just had the service for my own house upgraded from 100 A to 200 A, as part of a heat pump installation. However, I didn't really need all of that extra power. It was installed largely to future-proof my house, in particular, in case I or a future owner have to charge an EV. The EIA numbers show that US residential energy use has actually \_declined\_ slightly since 2005, despite population and housing growth, presumably reflecting increasing efficiencies. However, it's expected to grow in the coming decades as consumers move to electric heating (heat pumps) and, esp., EVs. I'm not doubting your experience or the specific claims about closed plants. But unless the EIA is missing something huge, there was something more complex happening than simply removing a big chunk of electrical demand & supply without replacing/shifting that demand & supply. The EIA has a web page charting the history of total US electrical power generation over time. It grew really quickly from 1950 to 2000. At that point, it nearly leveled off, but it did continue to grow. At no point was significant electrical capacity removed. And over the last 5 years the rate of growth has picked up significantly (mainly from natural gas power plants and renewables). You are right that a huge amount of coal-powered capacity was removed starting around 2005. But it was more than made up for by dramatic growth from other sources.