Four Time Scales for Technology Development and Deployment

Rodney Brooks identifies four distinct time scales for technology development: research ideas typically require 10-20 years to mature into solid lab demonstrations (exemplified by neural networks taking 60 years from initial concept to modern LLMs), followed by rapid innovation cycles once established. Hype cycles often accelerate visibility of technologies dramatically, with ideas like "AI agents" going from obscurity to ubiquitous marketing in months, only to fade as new hypes emerge, as seen with blockchain and the metaverse. Brooks argues that people frequently conflate these different time scales, leading to inaccurate predictions about when technologies will actually be deployed.