Computable Fairness: Boltzmann-Softmax Control for AI Resource Allocation

In three linesCFD (Computable Fair Division) proposes a framework for fairly allocating GPU and bandwidth resources among AI agents. The Boltzmann-Softmax β parameter controls the efficiency-fairness tradeoff. AHC++ adjusts β in real time to maintain diversity without major throughput degradation. Simulations: 100x agents = +5.5x execution time.

Summary generated by Claude — human-verified