Joules-to-GDP Framework Measures AI's Energy-to-Value Return

Key Info

Researchers introduced Joules-to-GDP (J2G), a framework that measures how well AI turns energy into economic value and identifies which tasks, models, and hardware yield the best return, evaluated across 44 occupations in OpenAI's GDPval.

Highlights

• The framework is called Joules-to-GDP (J2G) and links energy consumption to economic output.
• Evaluation covers 44 occupations spanning finance, engineering, law, and administrative work.
• Tasks with the highest economic value are not the tasks with the highest value per joule; CAD design earns ~10% more per task than building websites, but websites earn ~27% more value per joule.
• The most powerful model is not always the most energy-efficient choice for a given task.