AI Is a Rounding Error for Ordinary People: Diffusion Takes Fifty Years, Backlash Won't Wait
The Industrial Revolution gave ordinary people cheap clothes and sewing machines; AI so far gives them fun images and better search — the benefits are too indirect, and the political backlash is arriving too fast.
The video won't play here. Listen to the audio instead:
The argument · timestamps estimated from transcript position
AI is still a rounding error in daily life
The author spent a few weeks on his honeymoon without touching AI at all, and on returning realised: the points where ordinary people touch AI are either marginal or confusing. The positive impressions are making fun images and an enhanced Google search; the negative ones are addictive social media algorithms, a friend of a friend hooked on an AI chatbot, and various claims about data centers. Family, food, transportation, entertainment — the core of daily life — have barely changed directly. His conclusion: being fascinated by AI is the result of a very small number of people actively choosing it, and stepping outside the bubble reveals that what's happening today really isn't that important to most people.
The Industrial Revolution gave physical things; AI gives indirect benefits
The first Industrial Revolution brought cheaper clothes, cookware, reading material and new livelihoods; the second brought sewing machines, food preservation, indoor plumbing, photography, better light sources, bicycles and electrification — a long list, and a very physical one. The author worries that even if AI's benefits materialise (new scientific discoveries, treatments for rare diseases, sustained economic abundance), they are too indirect: ordinary people won't credit OpenAI or Anthropic with saving their life, because it was their family doctor who told them about the new treatment; and not many Americans will care that OpenAI solved the Navier-Stokes Millennium Prize Problem. He judges that fifty years from now, the daily life of the average American will still look a lot like today.
AI today is mainly a tool serving elites
Knowledge work is roughly half the US economy, and for that half AI is already (or soon will be) as foundational as electricity; rapid agent improvements over the next eighteen months will accelerate this. But a transformative, highly productive tool that only lifts half of society is highly unstable. The author cites the "Engels' pause": from 1790 to 1840 British workers' wages stagnated while GDP per capita grew rapidly amid technological upheaval. He argues that if the AI industry itself thinks this is the closest analogy, then backlash from those who don't benefit is legitimate. He also agrees with Doug O'Laughlin's judgment: knowledge-work output explodes while headcount in the tech industry very likely shrinks.
The backlash isn't rooted in AI, it's Big Tech's old debts
The author splits the political backlash facing AI into two layers: first, the early positive effects are too indirect; second, they are deeply intertwined with Western society's historical grievances against Big Tech. He stresses that the latter merely collided with AI in time — if AI's exponential growth had happened decades after the growing pains of platforms like Google and Meta, the data center issue very likely would never have risen to such a central political position. Solving either one would substantially relieve the pressure and give the industry more time to demonstrate positive reasons. And AI itself, through doom rhetoric and declarations of mass unemployment, has marked itself as a negative or unsafe technology, making both harder.
Robots might rescue the narrative for LLMs
The author raises an ironic possibility: if the intelligence explosion from mass-producing large language models spills over into robotics and accelerates the spread of robots into daily life, humans will quickly seize on AI's tangible benefits. This is ironic because many people have been working hard to convince the public that this LLM wave is very different from the general AI progress of the past ten or twenty years. If in the end the same dynamic saves (or largely drowns out) the LLM narrative, that would be funny. He also cautions that the connection between robotics, autonomous driving and the current AI revolution is, for now, more a matter of storytelling.
Diffusion takes fifty years; the fight against it won't wait
The author calls the present the first five years of a fifty-year diffusion process, and lays out two simple problems: the positive effects are too indirect, and the political backlash is entangled with Big Tech's history. He judges that the story of diffusion will last far longer than the struggle against it. Young people paying attention today will, within their lifetimes, see powerful AI go from effectively 0% to over 90% full adoption. What truly marks AI's evolution is the kind deeply integrated into enterprises and acting as a personal assistant; it is only just becoming viable, and adoption will be far slower than for easily grasped applications like ChatGPT. So continuing to push the technology forward is crucial — the benefits are astonishing but not guaranteed, and there is still a great deal of hard work to do on distributing them widely.
In their own words · checked verbatim
It’s a remarkable breath of fresh air to pop out of the bubble and realize how little what is happening really matters today .
Being obsessed with AI is a choice that a very few people have yet opted into.
It feels very likely in 50 years that the average American’s day to day life looks very similar.
Today’s AI is primarily a tool to serve the elite.
It’s highly destabilizing to have such a transformative, productive tool only bring half of society along.
The diffusion story will take a lot longer than the fight against it.
All of us younger folk following the story today will get to see powerful AI go from effectively 0% to 90%+ full adoption in our lifetime.
Figures
| Knowledge work's share of the US economy | about half | 3:47 |
| Expected window for rapid agent improvement | the next 18 months | 3:47 |
| Time span of the Engels' pause | 1790 to 1840 | 3:47 |
| The author's placement of AI diffusion | the first five years of a fifty-year diffusion process | 3:47 |
Glossary
- Engels' pause
- The period from 1790 to 1840 when British workers' wages stagnated while GDP per capita grew rapidly.
- Navier-Stokes Millennium Prize Problem
- The existence and smoothness of solutions to the fluid dynamics equations, one of the seven Millennium Prize Problems.
How to listen
Suited to founders and investors tracking AI narratives and policy risk, especially practitioners who want to understand why the public is indifferent or even hostile to AI.
The opening subscription-discount housekeeping can be skipped; the remaining eight minutes are worth hearing.