Edition #3 · March 2026
THE BIG STORY
The Grid Gets a Brain
The transformer — that humming grey box on your street corner — hasn't fundamentally changed since 1886. That's about to end, and when it does, the entire bottleneck blocking renewable energy at scale disappears with it.
Everyone focuses on generation. Solar's dirt cheap. Wind's scaling. Battery storage costs keep dropping. But the actual constraint? The grid itself — a one-directional, mechanically-switched, stubbornly analog network designed to move electricity from big central plants to passive consumers. That architecture breaks completely in a world where millions of rooftops generate power, electric vehicles charge and discharge unpredictably, and data centers swing from consuming two gigawatts to offering grid services within the same hour.
Solid-state transformers fix this. Not incrementally. Structurally.
A conventional transformer is copper wrapped around iron. It steps voltage up or down using electromagnetic induction — the same physics Michael Faraday demonstrated in 1831. It has no intelligence. It can't route power dynamically. It can't respond to grid conditions. It rusts. It leaks oil. It sits there for forty years doing exactly one thing.
A solid-state transformer replaces that with power electronics built on silicon carbide and gallium nitride semiconductors. It actively manages power flow in real time. It integrates renewables bidirectionally. It enables vehicle-to-grid discharge. It gets better through software updates. According to IMARC Group, the global market hit roughly $242 million in 2024. That's tiny — barely a rounding error against the $900 billion global electricity infrastructure market. But the growth trajectory matters more than the current size. Mordor Intelligence projects the sector expanding at 12-16% CAGR through the early 2030s, and the real acceleration hasn't started yet because utility procurement cycles move at geological speed.
When you generate cheap renewable electricity and then shove it through infrastructure designed for coal plants, you're running modern software on a computer from 1995. The processor bottlenecks everything. Solid-state transformers are the processor upgrade for the grid — they turn passive distribution into active orchestration.
And the timing is perfect, because data centers just became the forcing function.
Calibrant Energy signed a deal in New Jersey where an Iron Mountain data center doesn't just consume grid power — it provides services back to the grid, acting as a distributed energy asset that smooths demand peaks and fills supply valleys. This is the model. AI infrastructure needs massive, reliable power. That same infrastructure, equipped with battery storage and intelligent grid interfaces, becomes the densest concentration of controllable load on the planet, sitting exactly where grid operators need flexibility most.
This convergence is the story nobody's connecting. Solid-state transformers make the grid programmable. Data centers, reimagined as bidirectional grid participants, provide the anchor load and flexibility that renewable-heavy grids desperately need. Together, they solve the coordination problem that has kept energy abundance theoretical even as generation costs collapsed.
The pattern is unmistakable. Generation got cheap first — solar's 90% cost decline since 2010 proved that. Storage is getting cheap now — lithium-ion prices dropped over 90% since 2010, and alternative chemistries are emerging. But the grid itself remained the dumb pipe connecting smart endpoints. Solid-state transformers are the moment the pipe gets smart too.
The Asia-Pacific region is where this breaks open first. According to Mordor Intelligence, Asia-Pacific already commands roughly 40% of 2025 revenue in the SST market and is growing faster than any other region. China, Japan, and South Korea are building new grid infrastructure from scratch in many areas, which means they don't carry the legacy replacement burden that slows North American and European utilities.
By 2029, at least one major national grid will run a fully programmable distribution network using solid-state transformer technology as its backbone, likely in East Asia. The West will be three to five years behind because our utilities are publicly regulated monopolies with thirty-year asset depreciation schedules and zero incentive to replace equipment that technically still works.
The abundance implication is profound. When the grid becomes programmable, energy stops being a commodity you consume and becomes a platform you participate in. Every building, every vehicle, every data center becomes both consumer and producer, orchestrated in real time by intelligent infrastructure. The generation problem is solved. The storage problem is being solved. The distribution problem — the one that actually blocked full decarbonization — just found its technical answer in semiconductors that cost less than the copper they replace.
The grid is getting a brain. Everything downstream of that changes.
TODAY'S EDITION IS BROUGHT TO YOU BY
Unscarcity
Every week, we break down the three forces reshaping the global economy: robots getting cheaper than labour, AI getting cheaper than thought, and energy getting cheaper than ever. No hype. No hedge. Just the signals that matter. If someone forwarded this to you, now's the time.
TRIFECTA UPDATE
ROBOTS BMW expanded its humanoid robot pilot to Leipzig in February 2026, building on Figure AI's work at its Spartanburg plant where Figure 02 moved over 90,000 components in an 11-month production run. Mercedes signed a commercial agreement with Apptronik in 2024 to pilot Apollo in manufacturing logistics, and has since invested over €100 million in the company. These aren't press releases — they're purchase orders with delivery schedules. Automotive manufacturing runs on thin margins and zero tolerance for downtime, so when BMW keeps a humanoid on the production floor for eleven months, they've already stress-tested the business case. The constraint now is production capacity, not market readiness. Figure, Apptronik, and the Chinese manufacturers are competing for a market that went from theoretical to operational in eighteen months. Labour-intensive manufacturing is about to experience what logistics experienced with automated warehouses — sudden, irreversible transformation.
AI Legal research that took associates three days now takes Claude thirty seconds. Marketing campaigns that cost $5,000 in creative fees cost $50. The knowledge work displacement everyone debated hypothetically is already here, running quietly inside firms that don't advertise the fact. What's emerging isn't job destruction — it's leverage compression. One person with sharp judgment and a suite of AI agents now produces what entire departments used to. According to multiple enterprise surveys, the companies adopting this model aren't cutting headcount so much as redirecting it — fewer people writing reports, more people deciding which reports matter. The scarce resource isn't output anymore. It's direction.
ENERGY Calibrant Energy signed a deal where an Iron Mountain data center in New Jersey doesn't just draw from the grid — it feeds back into it, acting as a distributed battery. Data centers consuming gigawatts while the grid strains is yesterday's framing. This is the new model. AI infrastructure needs massive, reliable power, but that same infrastructure, equipped with battery storage and intelligent grid interfaces, becomes the asset utilities need most — controllable load that charges when power is cheap and discharges when it's scarce. The Iron Mountain site isn't an experiment. It's a revenue-generating grid service contract. When data centers stop being consumption problems and start being grid solutions, the entire energy infrastructure equation shifts. The grid gets smarter at every layer — solid-state transformers at the distribution level, intelligent load management at the consumption level. Abundance isn't just about generating more power. It's about orchestrating the power we already have.
THE NUMBER
98.5%
That's the energy conversion efficiency of solid-state transformers at rack level, according to manufacturers like Amperesand and DG Matrix. Conventional transformers lose 2-5% of electricity as heat during every voltage conversion. In a grid that converts voltage six to eight times between generation and your wall outlet, those losses compound. At 98.5% efficiency per conversion, solid-state transformers recover enough wasted energy across the grid to meaningfully reduce total generation requirements. When generation is renewable, every percentage point of efficiency gained means fewer panels needed, less land used, less capital deployed. Efficiency isn't glamorous. But at grid scale, it's the difference between energy abundance that works on paper and energy abundance that works in practice.
WHAT I'M WATCHING
Commonwealth Fusion's magnet supply chain. Multiple private fusion companies are now fighting over grid connection slots for the early 2030s, but the constraint nobody discusses is tritium supply and high-temperature superconducting magnet manufacturing. Whoever solves magnet production at scale controls the fusion timeline.
Utility procurement cycles cracking open. Solid-state transformers only matter if utilities actually buy them. I'm watching for the first major US utility to announce a distribution-scale SST pilot program — that's the signal that the technology has crossed from investor enthusiasm to grid operator conviction.
Nio's swap network economics. Nio's battery swap network hit a peak of 177,000 swaps in a single day during the 2026 Spring Festival, with routine volumes running around 100,000 daily. We're approaching the scale where unit economics become visible. If Nio publishes per-swap cost data in their next earnings, that number will tell us whether battery-as-a-service is a viable global model or a subsidized Chinese experiment.
THE QUESTION
If the grid becomes fully programmable and every building can both consume and produce energy, does the concept of a centralized utility — and the regulatory framework built around it — survive the decade?
UNSCARCITY A weekly newsletter on the convergence of robots, AI, and cheap energy driving sustainable abundance.
Was this forwarded? Subscribe here.
© 2026 Unscarcity. All rights reserved.