Four Time Scales Behind Why Tech Hype Is Wrong
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Neural networks research stretched roughly 60 years from McCulloch and Pitts's 1943 computational neuron models through Widrow's 1960 linear-threshold neurons to Hinton's 2012 deep-learning breakthrough before today's LLMs emerged, with the field "declared dead many times along the way."
- "AI agents" advertising went from absent on San Francisco buses in mid-2025 to now dominating nearly every AI-themed ad on the same buses, a compressed hype cycle that resembles blockchain, the metaverse, IBM Watson, and expert systems before it.
- Linux, first developed in 1991 as a free open-source Unix clone, wasn't adopted by Microsoft until 2012 — a 21-year lag that became the textbook case of how even zero-marginal-cost software typically takes 20+ years to scale.
- Self-driving cars span roughly 38 years from Ernst Dickmanns's 1987 Munich freeway demonstration through the 2007 DARPA Urban Challenge to Waymo's current fleet of about 4,000 vehicles licensed in San Francisco — "tiny compared to the number of cars in San Francisco, let alone the whole of the US."
- Hyperloop commercialization startups all shut down despite heavy hype, illustrating that many heavily promoted technologies fail completely in the commercial world while drawing capital and attention.
- Past economic transformations — domesticated animals, sailing ships, electrification, commercial aviation, and shipping containerization — each required over 50 years of continuous at-scale deployment before reshaping the world economy.
- Waymo dropped the author at the wrong destination despite more than 20 minutes on the phone with customer support, a single ride serving as the article's punchline on how far scale and reliability still lag the hype.
Why it matters: The article frames the two dominant capital-market narratives — LLMs displacing white-collar labor and humanoid robots displacing blue-collar labor — as conflating hype-cycle speed with the 50+ years of continuous deployment every past economy-reshaping technology has historically required. Workers planning careers and investors pricing labor displacement in the next decade are anchoring on timelines that have no precedent among prior economic transformations.
