Unscarcity
Sign in for free: Preamble (PDF, ebook & audiobook) + Forum access + Direct purchases Sign In

Unscarcity Research

Robots That Build Abundance: The Compounding Buildout

Gritt's robots let 8 workers install 4,000 solar panels a day, not 800. When machines build the energy machines, abundance starts compounding on itself.

9 min read 2038 words Updated July 2026 /a/robots-building-abundance

Note: This is a research note supplementing the book Unscarcity, now available for purchase. These notes expand on concepts from the main text. Start here or get the book.

Robots That Build Abundance: The Compounding Buildout

Or: What Happens When the Machines Start Building the Machines That End Scarcity

On July 21, 2026, a San Francisco startup called Gritt walked out of stealth with roughly $32 million and a claim that sounds like a rounding error until you sit with it. Its AI-directed robots — arms that bolt onto the excavators and pile-drivers a construction crew already owns — let an eight-person team install 3,000 to 4,000 solar panels a day. The same eight people, working by hand, place about 800. Gritt says it is already contracted to help build 2.8 gigawatts of solar in the next 18 months, and that three of the ten largest power-construction firms in America are customers.

Four hundred percent more panels per worker per day. Hold that number, because it is not really a number about solar panels. It is a number about the thing underneath solar panels, and fusion reactors, and data centers, and every other object in the abundance economy: the buildout. The unglamorous, sweaty, permits-and-pile-drivers work of turning a cheap design into physical infrastructure at planetary scale.

Most stories about abundance stop one step too early. They tell you the marginal cost of energy is collapsing toward zero — true — and imply the rest takes care of itself. It doesn’t. Somebody still has to build the collapsing-cost machine, and until last week that somebody was a human with a torque wrench and a sunburn. Gritt is the first company where the three ingredients of the abundance thesis — AI as the brain, robots as the body, and cheap clean energy as the fuel — converge on a single job site to build more of the fuel. That is not one more automation story. It is the moment the flywheel grabs its own axle.

The Buildout Was Always the Real Bottleneck

Here is the paradox that Gritt walks straight into. The solar panel is already one of the cheapest manufactured goods in human history. Module prices fell roughly 90% across the 2010s and kept sliding; the silicon is nearly free. So why isn’t the desert already carpeted in the things?

Because the panel was never the expensive part. The expensive part is everything around the panel: the graded land, the driven piles, the racking, the wiring, the interconnection queue, the labor. On a utility-scale project, a large and stubborn slice of the total cost is exactly this — the human hours of assembling cheap parts into a working plant. Learning curves crushed the cost of the component and barely touched the cost of the construction. You can’t 3D-print a substation, and until now you couldn’t automate a field crew.

That is the wall the terawatt buildout keeps hitting. Everyone quotes the falling cost-per-watt of the module and forgets that the United States needs to roughly triple its grid to run the AI it is busy building, and that the binding constraint on that tripling is not the price of glass — it’s skilled electricians, interconnection paperwork, and daylight hours. Our note on the electron gap makes the case that energy, not chips, is the true ceiling on the AI era. Gritt’s insight is one layer more specific: the ceiling isn’t even energy in the abstract. It’s the rate at which we can physically install energy. Attack that rate, and you move the one variable everyone else treated as fixed.

The Flywheel Nobody Priced In

Now watch what happens when you close the loop.

An AI-directed robot builds a solar field four times faster. That field produces cheaper electricity. Cheaper electricity lowers the cost of running the data centers that train the next, better AI — the AI that directs the robot. It also lowers the cost of charging the robot itself, and the cost of forging the steel and casting the silicon that the robot assembles. Cheaper, smarter robots then build the next field faster still. Each turn of the wheel lowers the cost of the next turn.

This is what economists call a compounding loop, and it is the same mathematics that made solar itself unstoppable — Wright’s law, where every doubling of cumulative production knocks a fixed percentage off the price. What Gritt adds is a second flywheel bolted to the first. It isn’t just that we build more panels so panels get cheaper. It’s that the builders get cheaper and faster too, and the two curves feed each other. Cheap energy makes cheap robots; cheap robots make cheap energy.

The book has a name for the pivot this loop crosses: the Terawatt moment, the point where energy stops being a line item and becomes the substrate of everything, because once energy is nearly free, food, water, compute, and manufacturing follow it down. The quiet argument of Gritt’s press release is that we reach the Terawatt moment sooner than the electron-gap pessimists think — not because the physics changed, but because the construction rate just became programmable. Abundance was always a supply-chain problem wearing a physics costume. Automate the supply chain of the supply chain, and the timeline compresses.

That is the genuinely new idea here, and it is worth saying plainly: the cost of abundance is falling faster than a straight line, because the thing that produces abundance is now producing itself.

The Part Where We Argue With the Optimism

If this note stopped there, it would be a press release with footnotes. It doesn’t, because a compounding buildout answers the how-fast question and sharpens two harder ones the book refuses to skip: who does the work and who keeps the winnings.

Start with the work. Notice that Gritt did not eliminate the crew — it kept eight people and made them four times as productive. This is the pattern we documented in Autonomy Theater: “autonomous” systems are almost always a spectrum, not a switch, with a human still in the loop holding the controls. Today Gritt is a power tool that makes electricians superhuman. But the honest reading of its own roadmap — “solar plants, then everything else” — is that the eight becomes six, then four, then a remote operator watching forty sites. The construction trades were supposed to be the safe side of the Labor Cliff: physical, dexterous, outdoors, hard to offshore, hard to automate. Physical AI is the news that the safe side was never safe. It was just next. The employment statistics that already show AI eating white-collar work are about to grow a blue-collar chapter.

Then the winnings, which is the question the book cares about most. A compounding buildout throws off an enormous, growing surplus — the gap between what abundance costs to produce and what it’s worth. Somebody captures that surplus. If the robots that build the machines that end scarcity are themselves privately owned, then abundance doesn’t automatically become a commons. It becomes a product, metered and priced by whoever holds the robots — the same capture risk we traced when hyperscalers started buying equity in fusion reactors instead of just buying their power, and the same logic that decides where value lands once intelligence itself is commoditized. Cheaper to build is not the same as free to everyone. A falling cost curve tells you the pie is growing. It says nothing about who gets a slice.

And there’s a third body already circling: whoever owns the compounding buildout also owns a veto. When one firm can install energy four times faster than anyone else, the communities and regulators who decide where the buildout is allowed to happen are negotiating with a much stronger counterparty. Speed is leverage.

Why the Book Saw This Coming

Unscarcity is built on a simple, uncomfortable claim: the technology that produces post-scarcity arrives before the institutions that distribute it, and the gap between the two is where the whole century is won or lost. Gritt is that claim in miniature. The Foundation — the guaranteed floor of housing, food, healthcare, energy, and compute that every Resident receives unconditionally — is only affordable if the marginal cost of the underlying infrastructure falls off a cliff. A self-compounding buildout is precisely the mechanism that makes the Foundation payable rather than utopian. No cheap terawatts, no Foundation; and now, no cheap robots to build the terawatts, no cheap terawatts. The recursion is load-bearing.

But the same recursion that makes the Foundation possible also makes capture possible, and faster. This is why the book spends its entire second half not on the technology — which it assumes will arrive — but on the plumbing of predistribution: structuring who owns the compounding assets before the concentration hardens, not taxing it apologetically afterward. It’s why Transition Trusts exist, to route the surplus of the buildout into public infrastructure rather than private moats. It’s why the framework measures contribution in Impact — a currency that decays — rather than in a capital stock that compounds forever into a permanent aristocracy of whoever owned the first robot fleet. When the machines start building the machines, the compounding is going to happen no matter what. The only open question is whether it compounds into a commons or into a cage.

Gritt is not a villain in this story. Robots that quadruple the rate at which we install clean energy are unambiguously good news for the timeline — they pull the abundance future closer, and they do it on a decarbonizing grid. But they are also the clearest signal yet that the abundance economy will be built by machines, at a compounding pace, throwing off a surplus large enough to fund a Foundation or a dynasty. The book’s whole wager is that we notice the difference in time to choose.

The panel got cheap. Then the reactor. Now the builder. Each layer of the abundance stack is automating the layer beneath it, and the stack is starting to lift itself. The compounding has begun. The distribution has not.

Further Reading

References

The machines are now building the machines that end scarcity, and each one makes the next one cheaper. The only thing still done by hand is deciding who owns the result. Argue it out at unscarcity.ai/forum.

Share this article:

Share Share Share