'Joules-to-GDP' framework: the highest-value AI tasks aren't the most energy-efficient
arena · x · 2026-11-05
Responding to Sam Altman's argument that home compute sits mostly idle so hardware belongs in the cloud, Mersad Abbasi agreed with the problem but disputed the solution, citing Jon Saad Falcon's Joules-to-GDP (J2G) framework — measuring how well AI converts energy into economic value, and which tasks, models, and hardware give the best return.
Using task data across 44 occupations (finance, engineering, law, admin and more) in OpenAI's GDPval, three findings emerge:
- The tasks with the highest economic value are not the highest value per joule. CAD design earns 10% more per task than building websites, but the latter earns 27% more value per joule.
- The most powerful model isn't always the best use of energy (thread truncated).
The framework offers a quantitative tool for the AI energy-vs-value debate and pushback against the "all compute must be centralized" view.
Related event: Stanford's Joules-to-GDP Framework Quantifies AI's Economic Value per Joule(8 posts)→
More from Infra
- Samsung's second Taylor fab reportedly fully booked before 2030, mostly for AMD and Nvidia HBM base dies — Beth_Kindig · 2026-11-05
- Why Tsugami's 31% operating margin shows manufacturers can't leave China — PAstynome · 2026-11-05
- Short RL phase on GLM 5.3 Flash costs $372 vs $1,894 for notes, breaks even in 8.7 days — teortaxesTex · 2026-11-05
- Texas lawmaker wants data centers to pay their own energy, recycle water, get community consent — Scobleizer · 2026-11-05
- Memory pricing is squeezing phone chips from both ends: ARM royalty growth cools to 8%, Qualcomm handsets down 26% — tengyanAI · 2026-11-05
- Nationwide voters rejected data centers in 40+ provisions as AI's scientific wins go unnoticed, argues investor — beffjezos · 2026-11-05