AI economics and pragmatic hedging

Building resilient systems with Sam Newman

Topic 15 of 19

Published Oct 7, 2026 · Clip 1:32:43–1:36:26 (3:43)

Sam argues AI is useful at the micro level but questions the macro economics: OpenAI, Anthropic and hyperscalers would need ~$2.6 trillion in revenue by 2030 to justify capex, versus a $1.4 trillion software market. He advises pragmatic hedging: multi-vendor, multi-model, using bigger providers as fallbacks, and swapping non-deterministic agent steps for deterministic code where possible.