Four Time Scales for Technology Development and Deployment

Rodney Brooks distills decades of observing AI, robotics, and technology into a four-scale framework that explains why predictions of imminent transformation are almost always wrong. Research takes decades, hype burns fast, deployment at scale takes longer than anyone expects, and reshaping the economy takes a human lifetime. The framework is simple enough to fit on an index card and devastating enough to dismantle most AI hype with a single question: "Which time scale are you on?"

Key Quotes

"I have come to understand four very different time scales for development of technologies and their deployments. And I think people often jump between them and end up making outrageously wrong, and sometimes damaging, predictions of when in the future a technology is going to be able to do what."

The thesis in two sentences. Brooks is not arguing that technologies don't develop — he's arguing that the category error between time scales is the root of both bad predictions and wasted capital. The person who sees a lab demo and predicts economic transformation in five years hasn't misunderstood the technology; they've misunderstood which scale they're looking at. #concept

"The ratio of extraordinary hype events to actual extraordinary technologies is way too high."

Brooks' deadpan summary of Time Scale 2. Blockchain, metaverse, IBM Watson, nanotechnology chinos, expert systems — each was going to change everything. Each faded. The pattern is the point: hype is not a bug in the technology development process, it's a feature of how humans respond to new research. The question is whether you can tell which is which before the chinos go on sale. #pattern

"Getting things to work at scale is orders of magnitude harder than getting them to work at first and having your first few dozen satisfied customers. Scale has always taken Herculean effort."

The transition from Time Scale 1 to Time Scale 3 is where most technology predictions die. Brooks' own Waymo ride — promised his house, delivered to the wrong location after 20 minutes with customer support — is the perfect illustration. The technology works (Time Scale 1). It's deployed (Time Scale 3). But at scale, edge cases aren't edge cases anymore — they're Tuesday. #pattern

"People think that when they hear about a new research result it is going to change everything. And then they believe the hype about how soon it will do so, as hype-notists manipulate capital markets through their promises."

"Hype-notists" is the coinage that deserves to stick. Brooks distinguishes between honest researchers, delusional true believers, and the people who knowingly use hype to manipulate capital markets. The third category is the smallest and the most damaging. #concept

"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."

The most important sentence in the piece and the one most likely to be ignored. The "causal correlation" parenthetical is doing heavy work: Brooks is suggesting that the 50-year timescale isn't a coincidence — it takes roughly one generation for the people who didn't grow up with the old way to become the people running things. The technology doesn't just need to work; the old guard needs to retire. #concept

Key Themes

#concept The Four Time Scales

Brooks' framework in full:

1. Research Ideas (10–20+ years): From first concept to solid lab demonstrations. Neural networks took 60 years; many technologies die and are resurrected multiple times along the way.
2. Hype Generation (months to years): Sudden explosion of press coverage, marketing, and re-branding. Everyone claims they've been doing it all along. Dies down when the next hype arrives.
3. At-Scale Deployment (20+ years): From engineered product to mass adoption. Software is faster than hardware, but even software takes two decades. "Scale has always taken Herculean effort."
4. Reshape the Economy (50+ years): From mass adoption to genuine economic restructuring. Domesticated animals, sailing ships, electrification, commercial air travel, containerization — every one took half a century or more.

The framework's power isn't in the individual scales — it's in the gaps between them. A technology can be solidly in Time Scale 1 (working in the lab) while sounding like it's in Time Scale 3 (ready for deployment) to anyone who can't distinguish a demo from a product. #concept

#pattern The Hype Graveyard

Brooks' catalog of dead hype cycles functions as a mnemonic device. If you're 30, you remember blockchain and the metaverse. A bit older: IBM Watson and nanotech. Old enough for social security: expert systems. Each generation gets its own hype cycle, and each generation believes this one is different. The graveyard is the argument: look at the track record. #pattern

#comparison Software vs. Hardware Deployment

Software has zero marginal cost to manufacture — you don't need supply chains and raw materials to go from one copy to thousands. And even so, "software typically takes 20 years or more to scale up." Unix → Linux → Microsoft adoption in 2012: 43 years from invention to running on billions of devices. Hardware takes even longer — Brooks' self-driving car timeline runs from 1987 to a still-incomplete 2026. The asymmetry is smaller than most technologists assume. #comparison

#concept The Hype-notist

Brooks coins a useful distinction: there are honest researchers doing the work, delusional true believers who "really believe things that they say, but which are impossible due to such little problems like fundamental physics," and hype-notists who knowingly manipulate capital markets. The economic damage comes from the third category, but the intellectual damage — the erosion of the public's ability to distinguish real progress from marketing — comes from all three. #concept

#person Rodney Brooks

Brooks is worth understanding as a source. He's not a skeptic by disposition — he co-founded iRobot (Roomba), Rethink Robotics (Baxter), and Robust.AI. He's spent his career building the future and has the scars to prove it. When he says "scale takes Herculean effort," he's not speculating — he lived it. His authority comes from having been inside the machine, not from throwing rocks at it from outside. #person

Critical Analysis

This is the most useful single framework for evaluating technology predictions I've encountered. It's simple enough to apply in real time — "which time scale is this claim operating on?" — and specific enough to falsify. Brooks' four scales don't tell you whether a technology will work; they tell you when to expect it, and the answer is almost always "later than you think."

The framework's strength is also its limitation: it's descriptive, not predictive. Brooks tells you that economic transformation takes 50+ years, but he doesn't give you a mechanism for determining which technologies will eventually achieve it. The framework is a filter against false precision, not a crystal ball. This is fine — most technology predictions need filtering more than they need precision — but it means the framework can't distinguish between "this will never work" and "this will work in 2040." Both produce the same observable: it's not working now.

The "causal correlation" between human lifetimes and economic transformation is the most intriguing and least developed idea in the piece. Brooks tosses it off in a parenthetical, but it's doing the philosophical heavy lifting. If the mechanism is generational turnover — the technology doesn't win until the people who grew up without it are no longer in charge — then the bottleneck isn't technical, it's demographic. That has implications Brooks doesn't explore: what if generational turnover is accelerating? What if AI itself changes the rate at which old-guard decision-makers are displaced? The parenthetical is almost certainly correct and almost certainly under-argued.

Brooks' Waymo anecdote is perfectly chosen and perfectly timed. After laying out a 39-year timeline for self-driving cars (1987→2026), he reveals that his own Waymo ride failed at the most basic task: getting him home. The technology that took four decades to reach production can't reliably navigate a city it's licensed to operate in. This isn't a gotcha — Brooks isn't saying self-driving is a failure. He's saying scale is harder than anyone admits, and the edge cases that are invisible in demos become the product at scale.

What Brooks doesn't address: AI's recursive potential. The framework assumes technologies develop on independent tracks — neural networks took 60 years, self-driving cars are taking 40. But if AI accelerates AI research itself (the recursive scenario), the time scales could compress. Brooks would likely respond that this is Time Scale 2 thinking — the hype of the moment mistaking itself for a structural break in the pattern. He might be right. But the recursive possibility deserves more engagement than the essay gives it, if only because it's the strongest counterargument to the framework's universal applicability.

The essay pairs naturally with AI Will Not Make You Rich — Neumann's containerization thesis maps directly onto Brooks' Time Scale 4, and containerization is Brooks' own canonical example of 50-year economic reshaping. The Market for Doom provides the institutional-precondition side of the argument: Brooks explains when transformation happens; Partridge explains what makes it possible. AI Mania Is Eviscerating Global Decision-Making documents what Time Scale 2 looks like from inside a large organization — the hype-notists have won, and honest discussion of timelines has become professionally fatal. The AI Productivity Paradox is the inevitable result of confusing Time Scale 2 (AI tools exist) with Time Scale 3 (AI tools are deployed at scale in transformed organizations).

Sources: raw/four-time-scales-for-technology-development-and-deployment, Four Time Scales for Technology Development and Deployment Last updated: 2026-08-06