Why should Moore’s law, at least the 120 year version, be exogenous to the economy? Is it actually? Is Kurzweil just BSing?
Claude: Key Question: Is Moore’s Law truly independent of economic forces?
Arguments Against Exogeneity:
-
R&D Investment
- Requires massive capital investment
- Depends on market incentives
- Intel/TSMC/etc. make explicit economic decisions
-
Market Demand
- Semiconductor improvements track where money flows
- Gaming/AI/mobile drive different optimization priorities
- Economic cycles affect research funding
Arguments For Exogeneity:
-
Physics-Limited
- Fundamental materials constraints
- Quantum effects at small scales
- Heat dissipation limits
-
Path Independence
- Different companies/countries achieve similar results
- Progress continues through recessions
- Multiple approaches converge on similar improvements
My Take: The 120-year version (including mechanical/electrical computation) seems more exogenous than the semiconductor-specific version. The longer timeline suggests deeper forces than just market dynamics.
But Kurzweil likely overstates the determinism. The pattern exists but requires massive economic coordination to maintain. It’s more like “potential progress given sufficient investment” than a pure physical law.
Most interesting insight: The distinction between short-term economic dependence and long-term technological potential. Markets determine timing but not ultimate possibilities.