Hannah Ritchie of Our World in Data writes data-driven posts on energy, climate, electric vehicles, and the environmental footprint of AI.
Commenters cite it as solid data, such as AI energy use per query and EV versus petrol car emissions. One calls her work fantastic.
> Nuclear plants probably kill less wildlife and people per delivered kWh than lithium extraction
Uranium extraction has not negligible effects. Recent case: https://www.rfi.fr/en/africa/20230124-french-uranium-miner-leaves-20-million-tonnes-of-radioactive-waste-in-niger
> Even as we ramp lithium production as fast as we can, we’re facing production shortages
Maybe, however the big picture seems clear: https://hannahritchie.substack.com/p/lithium-electric-vehicles
> We should pick the top 5-10 technologies
Globally we did so during the last 20 years, as many nations (including the US) tried hard to build and then use nuclear reactors and also renewable sources of energy:
https://ourworldindata.org/grapher/nuclear-renewables-electricity?country=~OWID_WRL
The author Hannah Ritchie works on Our World In Data and also publishes the fantastic Sustainability by Numbers substack. It's in the same vein as the late, great David MacKay's Sustainable Energy Without the Hot Air.
This tool has its own recent substack post. See the comments too, especially the one by Chris Preist that contextualizes the energy usage of streaming video (a topic that has also been discussed on HN before).
https://hannahritchie.substack.com/p/does-that-use-a-lot-of
OP says one query uses 0.3 Wh. Driving an electric car for 10 miles = 3,000 Wh which is roughly 10,000 Wh per hour.
I'm not sure how many queries is equivalent to an hour of Claude code use, but maybe 5 seconds, which means an hour of continuous use = 216 Wh, or ~50x less than an electric car.
OP has a longer article about LLM energy usage: https://hannahritchie.substack.com/p/ai-footprint-august-2025
> The difference is that a lithium battery facility can be installed and functional in 12-24 months.
These things aren't comparable without putting it together as a system.
> A nuclear plant takes 12-24 years.
You're thinking about wild outliers there. It's around 6.5 years median [1].
[1] https://hannahritchie.substack.com/p/nuclear-construction-time
> the reactor in Belarus was built in 8 years
That’s slower than average [1]. Striking, considering how much where you build influences timelines as much as who’s building. (I doubt Minsk subjects Rosatom to community input, labor requirements or environmental reviews.)
[1] https://hannahritchie.substack.com/p/nuclear-construction-time
https://hannahritchie.substack.com/p/ev-fossil-cars-climate
This even compares cars built in an unrealistic 100% coal grid, and fuelled on 100% coal grid. Driving 14k miles a year for 10 years a Tesla Model 3 built under those extreme conditions beats a gas Fiat 500