AI 'productivity pollution': the hidden tax on everyone
One person's AI-fueled productivity causes extra work for everyone else. It's time to stop the delusional thinking about AI.
SILICON VALLEY, CALIFORNIA (AUGUST 12, 2026) — Early refrigerators ran on ammonia, methyl chloride, and sulfur dioxide. These gases were toxic, flammable, and deadly to families that owned refrigerators. So in 1928, Thomas Midgley Jr. solved the problem by inventing chlorofluorocarbons to replace the toxic gases used in refrigerators.
Over the ensuing decades, chlorofluorocarbons from refrigerators destroyed the Earth’s ozone layer, the ultraviolet shield protecting every living thing on Earth. The depletion of the ozone layer dramatically increased UV-B radiation reaching Earth’s surface, leading to higher rates of skin cancer, cataracts, and immune suppression in humans, severe damage to marine food chains and terrestrial ecosystems, and a big shift in Southern Hemisphere climate patterns. (The ozone layer’s depletion was reversed by the 1987 Montreal Protocol, a global treaty that phased out the production and use of chlorofluorocarbons and other ozone-depleting halocarbons, allowing stratospheric ozone concentrations to gradually recover. Full recovery is not expected until roughly 2050.)
Millions of consumers buying chlorofluorocarbon-based refrigerators to solve their individual, personal home pollution risk caused global pollution that harmed the entire planet.
Now, something comparable is happening with AI.
Call it “productivity pollution.” One person’s AI-fueled productivity increase is lowering everybody else’s productivity.
How “productivity pollution” works
A podcast from The Economist magazine called “The Intelligence” reported on a phenomenon in Britain that’s surely global: People are using AI to cope with the confusion and complexity of dealing with government bureaucracy.
Great, right? Well, not if you’re the bureaucracy.
The podcast described the crisis like this: Government procedures for filing employment tribunal claims, submitting planning objections, applying for benefits, appealing parking tickets, and challenging rent increases under the Renters’ Rights Act are complex and confusing. So citizens are using AI to not only deal with the complexity, but also find and take advantage of obscure rules and loopholes.
As a result, government workers are being overwhelmed by the number and complexity of citizen inputs. Because they’re overwhelmed, fewer requests or claims are being processed. And so the massive productivity gains of citizens aren’t resulting in the outcomes they wanted because their individual productivity reduced the productivity of the system.
This is happening in business, too.
READ: The Attachment Economy, Computerworld, Gastronomad newsletter, Gastronomad book
LISTEN: Superintelligent, TWiT, Gastronomad podcast
JOIN: The Gastronomad Experience
FOLLOW: Gastronomad on Surf Social, Machine Society on Surf Social, Bluesky, Reddit, Notes, Mastodon, Threads, X, Instagram, Facebook, Linkedin!
LOOK AT: Mike Elgan Photography
The time and trouble of creating a business report, proposal, plan, slide deck, or budget proposal used to limit their size, complexity, and frequency.
Now, thanks to AI, people can churn out incredibly complex business communications, ideas, and proposals in a few minutes. Using AI makes the person generating such documents super productive. But then it burdens everyone else who has to sift through those documents, teasing out hallucinations, problems, and irrelevancies, and struggling to grasp ideas that even the so-called creator hasn’t taken the time to understand. One person’s productivity is everyone else’s information overload, leading to lower productivity for the company.
In other words, AI creates positive personal results, but wide negative externalities.
The “productivity pollution” dynamic is playing out elsewhere:
Job applicants can use AI to apply for far more positions, burdening hiring managers and slowing the hiring process. (LinkedIn data shows applications per job posting have more than doubled since since spring 2022.) AI that’s supposed to make it easier to find a job actually makes it harder.
AI is emboldening more people to represent themselves in court and also speeding up the work of lawyers. A 2026 MIT study of 4.5 million federal civil cases found that around 18% of civil complaints now contain AI-generated text (up from essentially zero in 2023). That burdens courts and judges, who have to now check long, complex filings to see if they contain nonexistent cases or other hallucinations, slowing down justice. AI that’s supposed to make court cases quicker makes them slower.
Students use AI to write long, complex essays, forcing professors to use AI tools to process and grade them. Neither the student nor the teacher is reading the essays. The student isn’t learning. The teacher isn’t teaching. AI that’s supposed to make it easier to get an education makes students less educated.
The AI-pilled chatbot enthusiasts who believe that AI is solving all their problems are seeing at the impact of the technology through their own narrow, personal lens.
Even worse, some analysts and and AI companies are looking at the apparent productivity gains of a single person, extrapolating that across thousands in an organization, and falsely concluding massive productivity gains.
Instead, the overuse and misuse of AI are burdening organizations with productivity reductions, as the total size, quantity and complexity of business communications explodes.
The idea that productivity-enhancing AI might reduce productivity sounds paradoxical, so consider this oversimplified thought experiment. Suppose AI enables you to write three times as many emails as before, say, 30 a day instead of 10. Your email-writing productivity has tripled, making you more valuable to the company. The problem is that every additional email you send lands in someone else’s inbox. Your colleagues, who once got 10 emails from you daily, now get 30. Multiply that across an organization: if 10 people triple their email output, the team gets 300 emails a day instead of 100. If 100 people do the same, that figure blows up to 3,000, three times the reading burden. The AI that made your email writing easy makes email reading hard for everyone else.
This is a classic case of the Tragedy of the Commons.
How many cows?
We’ve all heard the phrase “Tragedy of the Commons” thrown around. The concept originates with William Forster Lloyd, the Drummond Professor of Political Economy at Oxford, who in an 1833 pamphlet titled Two Lectures on the Checks to Population used a hypothetical example of shared pastureland to argue that commonly held resources would be overexploited compared to privately owned ones.
He explained the reason. When medieval farmers acted rationally, they limited grazing on their own land to preserve the soil and grass for the next season. They balanced the maximization of their herd against the cost of overgrazing.
But in the shared pasture lands, the “commons,” they maximized their herds without having to pay the price of over-grazing. Other farmers made the same choice. And so the commons became tragically over-grazed. The rational self-interest of each farmer resulted fewer acres of grazing land for all.
With AI, the time and attention of fellow employees is the “commons.” An ambitious worker can churn out massive work for those employees at little cost to themselves, and so they do it. But the result is a degraded field.
