TokenFootprint
A minimal mathematical framework for total AI energy: training + inference at scale. Closed-form break-even time, growth models, and sensitivity analysis.
AI Externalities Lab
Research track
A math-first research program on the external costs of AI at scale: energy, carbon, and the dynamics of adoption. Focus: modeling, theory, and clear quantitative claims.
Projects
Coming next
- Inference Dominance under logistic adoption
- Rebound effects (efficiency → demand → footprint)
- Carbon intensity as a time-dependent control