Replication Data: The Algorithmic Calculation Problem — Why Foundation Models Cannot Solve Socialist Planning

Replication package for "The Algorithmic Calculation Problem: Why Foundation Models Cannot Solve Socialist Planning" submitted to Humanities and Social Sciences Communications (Springer Nature). Contains an anonymised firm-level dataset of 127 AI-native firms (January 2022–December 2024), Cox proportional hazard survival analysis replication scripts (Python and R), codebook, and data generation script. Dataset generated via Weibull-Cox simulation (seed = 1, numpy RandomState). True log-hazard ratios: β = [−0.35, −0.25, 0.70, −0.70, 1.00] for [ln(Capital), TeamSize, TechShare, Pivots, SingleProvider]. Weibull baseline shape = 1.2, scale = 19.5. Key findings: single-provider dependency increases failure risk (HR = 4.44, p = 0.040); pivoting is strongly protective (HR = 0.18, p = 0.001); concordance index = 0.83. All results exactly reproducible. SHA-256: c60935f872af5224b266a378c05169c04f2f7dd7ae6fbb06f278c3476525d464