Unscarcity
Listen to the Preamble free. No sign-up, no email. Listen free Sign in for free: Preamble PDF & ebook + Forum access + Direct purchases · Sign In By signing in, you get our daily podcast email. Unsubscribe any time.

Unscarcity Research

How Will Governments Change in a Post-Scarcity World?

Government doesn't wither away, and AI doesn't take the throne. Its job narrows, machines take the whistle, and power gets an expiry date.

6 min read 1385 words Updated November 2026 /a/blog-governments-change-post-scarcity

How Will Governments Change in a Post-Scarcity World?

How Will Governments Change in a Post-Scarcity World?

Of all the post-scarcity questions, this one splits a thread fastest. A composite of how it usually arrives:

“If machines do the work and nobody earns a wage, what is the government even for? Does it just fade away? Or does some AI end up running everything, and we all just hope it likes us?”

Those are the two pictures people carry: the state withering away, or a machine taking the throne. Unscarcity argues for neither. It describes three changes, and each one is already visible in rough form today. The job of government narrows and hardens. Machines take over enforcement while people keep the rulebook. And power, by design, comes with an expiry date.

Change One: The Job Gets Smaller and Harder to Skip

Much of what governments do today is a way of managing scarcity: taxing people who work, deciding who qualifies for help, rationing what there isn’t enough of. Take away the scarcity and a large share of that machinery has nothing left to do.

What remains becomes non-negotiable. In the book’s architecture, the Foundation handles material survival as a universal right: food, housing, healthcare, energy. It is meant to work like infrastructure, closer to a power grid than a welfare office. Nobody applies for the grid, and nobody has to prove they deserve the lights.

Running it is mostly logistics, and logistics is where machines are already good. The companion research describes a Civic Layer that operates as “a referee and registrar, not a ruler.” When a community needs 50,000 more kilowatt-hours, it routes the energy. When someone asks for shelter, it finds housing. But it “never decides values.” Whether your community prizes density or sprawl, silence or celebration, stays a human question.

So the first change is a government that does far fewer things and cannot stop doing the few it keeps.

Change Two: Machines Enforce, People Decide

The closest working model is not a government at all. It’s tennis. After a run of bad calls against Serena Williams at the 2004 U.S. Open, the sport turned to Hawk-Eye, a camera system that reconstructs where the ball landed. Three of the four Grand Slams have since removed line judges altogether; Wimbledon retired them in 2025, after 147 years. Nobody has proposed letting Hawk-Eye rewrite the rules of tennis.

Singapore runs about 2,700 traffic intersections through regional computers, with human operators who can override them and a designed fallback: lose confidence in the algorithm and the lights revert to fixed timing. Wikipedia lets bots make a large share of its edits, logs every bot action in public, and lets any human reverse any bot instantly.

The pattern holds each time. Enforcement is cheap, fast and automated. Rule-making stays with people. And the line between the two is guarded by one thing: an override that anyone can use and everyone can see.

From the Podcast: A Front Door With a Back Office

The September 30, 2026 episode of Minds, Bodies & Terawatts, our daily podcast, caught a government trying the first half of this and skipping the second. Its headline: The Government’s New AI Front Door Runs on Google’s Gemini and Musk’s Grok. Its First Answers Contradicted the President, and Before He Finished Speaking It Stopped Taking ‘Political Questions’.

The White House had launched America.gov, a chatbot that by its own description folds 29,000 government websites into one box, and an executive order gave every agency 90 days to plug its public services into it. Asked whether Trump won in 2020, the bot first answered that official results show Biden won. While the president was still on stage, according to AP, the answer changed to a refusal to take political questions. The White House’s only comment was that the site “will continue to be updated.” Mikey Dickerson, who once ran the government’s in-house tech team, told FedScoop the launch was a demo that “solved like the easiest five percent of the problem.”

The episode’s argument is the one this post is about. Machines answering for the state is the easy part, and it is arriving. The hard part is who can quietly change what the machine says. As the guest put it: “The clerk is fine. The hand rewriting the clerk’s instructions is the problem.” A steered answer about a benefit or a deadline, the episode points out, is exactly the one you’d never notice.

The book’s second law, Truth Must Be Seen, reduces this to four words: “No decision in the dark.”

Change Three: Power Gets an Expiry Date

Smaller and more transparent government still leaves the oldest problem in politics. In 1911 the sociologist Robert Michels studied the German Social Democratic Party, close to a million members obsessed with internal democracy, and watched it drift into the hands of a small professional leadership. He called the pattern the Iron Law of Oligarchy, and nothing since has repealed it. US federal lobbying hit a record $5.08 billion in 2025. The blockchain experiments meant to decentralize decision-making ended up with under 1 percent of token holders controlling about 90 percent of the voting power.

The book doesn’t try to beat that gravity with better people. It answers with decay, on the principle that “power that cannot accumulate cannot ossify.” Emergency powers in the MOSAIC “auto-destruct at 90 days,” renewable only after 30 days of normal governance or with 90 percent consensus across diverse communities. Any community may leave or fork the federation if its rules stop serving members, and because the Foundation is guaranteed everywhere, leaving doesn’t mean losing food or healthcare. “Term limits a ruler can ignore,” the chapter says. “Code a ruler cannot.”

On unscarcity.ai: Three Articles That Go Deeper

AI as Referee, Humans as Conscience covers the enforcement and decision split in detail, from Hawk-Eye to Wikipedia’s 2,855 approved bot tasks, plus the trap that comes with it: humans who sit in the loop too long get rusty. The Iron Law of Oligarchy lays out how organizations get captured and four mechanisms meant to resist it, and closes with an honest caveat that no system can guarantee the outcome. The Chapter 3 companion summarizes the whole governance argument, including the Diversity Guard and the Five Laws.

In the Book: Two Neighbors, One River

Chapter 3 of Unscarcity, “The MOSAIC of Commons,” opens in Philadelphia in 1787 with fifty-five men who nailed the windows shut and spent four months on one problem: “how do you build a government strong enough to stop the chaos, but too weak to become the chaos?”

The section “Two Neighbors, One River, Three Ways to Fail” turns that into a test. Upstream, a campus rinsing computer chips in ultra-pure water. Downstream, rice paddies fed by the same streams for four hundred years. “Go ahead. Govern them.”

The Pyramid puts one office in charge. It is fast, and it is blind: it “cannot see the paddies from the capital, and the engineer who noticed the problem first has learned not to mention it.” The Islands let the neighbors split and compete for residents, “except the river does not split,” and exit only works for people who can afford it. The Federated Empire builds a water ministry and a court, and its center never shrinks; the European Union began in 1951 as a coal-and-steel deal and now “legislates the chemistry of shampoo.” The chapter’s verdict on that third shape: “a trellis that grows into a cage, one regulation at a time.”

The section ends by naming the alternative, “a network of self-governing communities with no permanent center to drift toward, held together only at their edges by a handful of shared rules.” The next section, “A Trellis That Cannot Close,” builds it out as the MOSAIC, with Switzerland as precedent: four language groups and twenty-six self-governing regions held together for seven centuries without a king. Nobody owns the internet either, the chapter notes, and more than five billion people use it every day.

The book then takes the river all the way to a settlement in 2047, where a mediator AI with no decision-making power puts the hydrology on a public ledger, and both sides learn that their own numbers were wrong.


More questions like this one are collected on the blog page. Follow the series on Substack.

Share this article: