Four Time Scales for Technology Development and Deployment

Rodney Brooks — roboticist, former MIT CSAIL director, and co-founder of iRobot, Rethink Robotics, and Robust.AI — identifies four distinct time scales that govern how technologies develop and deploy, and argues that conflating them produces outrageously wrong predictions.

Time Scale 1: New Research Ideas (10–20+ years). Research takes one to two decades to progress from initial ideas to solid lab demonstrations, and sometimes much longer. Neural networks trace from McCulloch and Pitts (1943) through Widrow's linear threshold neurons (1960) to Hinton's deep learning breakthrough (2012) to today's LLMs — a sixty-year arc with many declarations of death along the way. Once a technology solidifies in the lab, a gold-rush phase of rapid tweaks can create the illusion of imminent transformation.

Time Scale 2: Hype Generation (months to years). Hype cycles erupt suddenly — a technology goes from obscurity to daily business-press coverage in months — and then die down as new hype arrives. Brooks catalogs the graveyard: blockchain, the metaverse, IBM Watson, nanotechnology (including nanotech-infused chinos), expert systems. The problem is that people not steeped in the technology confuse ongoing research with the hype about how it will change everything. The ratio of extraordinary hype events to actual extraordinary technologies is "way too high."

Time Scale 3: At-Scale Deployment (20+ years). Even software — with zero marginal cost to manufacture — typically takes 20 years or more to scale. Unix began in 1969; Linux arrived in 1991; Microsoft adopted it in 2012; now it runs most backend systems and billions of mobile devices. Hardware takes longer. Brooks first saw a self-driving car on a freeway in 1987; rode in a Waymo predecessor in 2012; rode in a production Waymo in 2026 — but it dropped him at the wrong location after 20 minutes with customer support. "Scale has always taken Herculean effort."

Time Scale 4: Reshape the Economy (50+ years). Every transformative technology — domesticated animals, sailing ships, electrification, commercial air travel, containerization — took over 50 years of continuous at-scale deployment to reshape the economy. People hear about a new research result, believe the hype about how soon it will change everything, and then think it can reshape the world economy in one or two or ten years. It doesn't. "It really does take decades, essentially a human lifetime (and there is a causal correlation), to deploy a technology at large enough scale that it reshapes the world's economy."

Brooks' central warning: the current AI-and-robotics hype — replacing white-collar labor with LLMs and blue-collar labor with humanoid robots, followed by a magical world of universal wealth — ignores all four time scales. Research is not deployment, hype is not reality, and economic transformation is not imminent.