Food Production in Post-Scarcity Systems
A composite of the versions we keep hearing:
“Everyone says robots and fusion will make food free. But food isn’t a smartphone. It grows in dirt, it needs rain, it rots. Who actually farms in a post-scarcity world? Is it all lettuce from a skyscraper? And if feeding people is so easy, why are people still hungry now?”
Start with the last question, because it carries most of the answer. Unscarcity does not treat food as a production problem waiting for a miracle machine. It treats it as three problems. We already grow enough, and the failure is in getting it to people. The machines that would grow it more cheaply are less ready than the brochures say. And a food system tuned only for cheapness is one blight away from collapse.
The Food Already Exists
First, the waste. ReFED’s 2026 report puts the value of America’s 2024 surplus food at $380 billion, only a small share of it donated, while 47.9 million people lived in food-insecure households, 13.7 percent of all households. That is also the last official count: the USDA cancelled the 30-year-old survey in September 2025, calling it “redundant, costly, politicized, and extraneous.”
The raw ingredients are cheap. Bulk rice runs about five cents per 100 calories, beans about eight. What makes food expensive is everything between field and plate: the 1,500 miles an average American meal travels, the refrigerated trucks, the spoilage, the retail markup.
That is why the food line in a Free Zone, the framework’s proof-of-concept community, is budgeted at about $1,500 per person per year. The arithmetic is 2,000 calories a day, 730,000 a year, at two-tenths of a cent per calorie. That figure is a target that assumes automation and cheap energy, not a price anyone pays today. But the bet rests on removing distance and waste, not on a breakthrough.
Vertical Farms: The Honest Version
The book’s future has them. Maria Delgado’s pantry in Chapter 1 “overflows with food she didn’t pay for, grown in vertical farms rising forty stories above the old parking lots.” The present-day numbers are less kind.
In 2026, delivered greens cost roughly $5 a pound from a vertical farm, $4 from a greenhouse and $1 from a field, and that gap widened over the past two years. The industry’s flagship names are gone or shrunken: AppHarvest went bankrupt in 2023, Bowery Farming shut down in 2024, Plenty filed Chapter 11 in 2025, and Infarm pulled out of most of its markets. The post-mortems agree that the growing worked. The companies scaled before they were profitable, and energy ate the margin. Meanwhile the input everyone was counting on stalled: IRENA’s 2025 figures put the global average cost of solar power at $0.044 per kilowatt-hour, a tick up from 2024 and the first year in over a decade that it didn’t fall.
Singapore ran the experiment as national policy. In 2019 it set a goal of producing 30 percent of its nutritional needs locally by 2030. By late 2024, only 3 percent of the vegetables Singaporeans ate were grown locally, and in November 2025 the government formally replaced “30 by 30” with narrower targets: 20 percent of fiber and 30 percent of protein by 2035. The first scorecard, in May 2026, showed local farms supplying about 8 percent of fiber and 25 percent of protein. Malaysian farms can sell lettuce for half the price of Singapore’s vertical farms because they have the land, labor and energy.
So the defensible claim is narrow. Vertical farming does not beat dirt. It beats distance and weather: it wins where transport, water or the growing season is the binding constraint. Open fields keep doing most of the work, more and more of it with machines.
From the Podcast: The Farm Precedent
Farming is also the one sector that automation has already emptied, and the November 6, 2026 episode of Minds, Bodies & Terawatts, our daily podcast, used it as the yardstick. Its headline: The First American Humanoid Robot Stock Closed Up on a Day the Nasdaq Fell 5.5%. Its Co-Founder Says the Robot Rents at About What the Worker Costs, and the Filing Says It Will Get Five Times Cheaper to Build.
Asked what happens to the warehouse workers that Agility Robotics’ Digit robots replace, co-founder Jonathan Hurst pointed to the farm: most Americans once worked in agriculture, hardly any do now, and automation is the reason. The episode grants the point, then asks why that transition worked out. The machines made food cheap next to what people earned, and Agriculture Department figures show food takes a far smaller bite of family income than it did in our grandparents’ day. The host asked whether “the family that left the farm got paid back at the grocery store.” Through the price, yes.
A post-scarcity food system depends on that mechanism, and the episode warns that it isn’t automatic. Digit rents at a rate comparable to the human labor it replaces, while the company’s target build cost falls from about $150,000 in parts to $30,000. If prices stay pegged to wages, the savings stop at the warehouse. Automated food only feeds people if its savings reach the plate, through competition or, in the Foundation, by taking the price off entirely.
What a Kitchen Becomes
Once calories flow like tap water, the question moves from the farm to the kitchen. Chapter 9 follows Ines, Maria’s sister, who runs a small Puerto Rican restaurant on her grandmother’s recipes. In the Foundation economy she keeps cooking. “Ingredients arrive on the Foundation’s logistics, the same way they arrive at every kitchen,” and diners “come because they love the food, not because they need calories.” She earns Impact and standing. Rent used to be her first thought every morning; now it is the man at table four who sent the mofongo back for too much garlic.
The supply side is built for bad years, too. In the same chapter, every Mission Guild contributes surplus capacity to a solidarity fund, so “when a drought cripples agricultural output in West Africa, the Fund redirects industrial capacity from Guilds worldwide. Not as charity. As the system working as designed.”
On unscarcity.ai: Three Articles on the Plate
$15,474 on Healthcare vs. $8,000 for Everything lays out the $1,500 food line and the vertical-farm correction in full, including its own admission that “the economics will improve” is a forecast, not a plan. How Singapore Uses AI to Run a Country of 6 Million tells the 30-by-30 story and concludes that the technology was never the bottleneck: if food security is a right and not a market outcome, the question changes from “which lettuce is cheapest?” to “which food system ensures we can eat if shipping lanes close?” Inca Economy Without Money is the historical case, an empire that fed 12 million people with no currency, holding three to seven years of reserves in stone storehouses called qollqas. “Modern supply-chain optimization would call this ‘waste.’ The Inca called it ’no famines.’”
In the Book: Never Grow Only One Crop
Chapter 3 of Unscarcity, “The MOSAIC of Commons,” sets out five laws for a federation of very different communities. The fifth, “Law 5: Difference Sustains Life,” opens with a food disaster. In the 1840s Ireland grew almost exclusively one variety of potato. When a blight arrived that this variety couldn’t resist, the whole food supply collapsed at once. A million people died and another million emigrated, “not because there wasn’t enough food in the world, but because a colonial economy had left an entire nation’s poor dependent on a single crop.”
The chapter turns that into a rule, “Never grow only one crop,” and applies it mostly to how people think and govern: panels whose members share the same blind spots fail together. It ends the section with the line that matters most for food: “Monoculture is always efficient. Right up until the blight arrives.”
The book’s own picture of abundance follows the rule. In Chapter 1, the Foundation in Lagos “does not look like the Foundation in Detroit”: its vertical farms “grow egusi and ugu, two West African staples, instead of kale.” The guarantee is the same in both cities. What grows under it is not. In the book, that difference is treated as insurance, and the premium, in Chapter 3’s words, “was simply: let them be different.”
More questions like this one are collected on the blog page. Follow the series on Substack.
