AI, work & competition
Computational game-theory study
A computational study of how competing firms adopt AI, and why the best choice changes with the objective.
- Status
- Unpublished working study
- Kind
- Data & geometry
- Stack
- Python · Game theory · R · Sensitivity analysis
BUILD NOTES / Unpublished working studyAI, work & competition
Project overview · not an interface capture
Computational game-theory study
Python / Game theory / RProject overview · not an interface capture
The problem
Productivity, employment protection, and welfare can favor different adoption decisions. I wanted to make those tradeoffs explicit rather than hide them inside one metric.
What was built
- A two-vendor model with routing decisions, labor displacement constraints, and competing objectives.
- Best-response and equilibrium checks alongside scenario comparisons.
- Sensitivity analysis across 22 assumption profiles to examine which conclusions depend on the model’s inputs.
What I learned
An optimum is conditional on its objective and assumptions. Reporting both is as important as solving the optimization problem.