The compact machine at the center of Benj Conway’s working life makes its point in a fraction of a second. A pulse of electricity enters a stream of plasma. The current generates a magnetic field. The field squeezes the plasma into a filament hot and dense enough for atomic nuclei to fuse. There are no cathedral-sized external magnets in Zap Energy’s version, no laser complex, no cryogenic architecture. The apparatus is still ferociously difficult. It simply has the manners to look modest.
That distinction explains Conway better than a standard founder biography does. He is not the plasma physicist in the story. Uri Shumlak developed the scientific idea at the University of Washington; Brian Nelson built generations of machines around it. Conway’s contribution was to recognize that a simpler device could change the economics of discovery. A fusion company needs exquisite science, but it also needs to learn before its capital and calendar expire.
Conway arrived at that view by a route almost comically unsuited to a neat résumé. He studied medicine at Gonville and Caius College, Cambridge, where he earned a first-class result, and at University College, Oxford. Then he joined the British Foreign and Commonwealth Office. Much of the next decade was spent posted across Asia and the Middle East. Medicine teaches diagnosis; diplomacy teaches that the stated problem and the actual problem are often different. Both habits would prove useful.
The useful education of a difficult market
After government came frontier-market investing. Conway co-founded InFrontier, a private-equity firm that put capital into small and medium-sized businesses in Afghanistan. The pitch met a predictable welcome. One large institution, according to an account of the fund’s early years, answered the word “Afghanistan” with “over my dead body.” Conway and his partners began deal by deal anyway, investing in businesses that included a pharmacy chain and an LPG importer. A track record eventually helped them bring state-backed investors into a $22 million fund.
It is tempting to turn that episode into a parable about appetite for risk. The sharper lesson is about separating a category from the evidence inside it. “Afghanistan” could end a conversation before a business was examined. “Z pinch” could do much the same in fusion. The concept had charmed researchers since the field’s early years: run enough current through plasma and its own magnetic field will pinch it inward. It was elegant, compact and famously unstable. The plasma buckled and broke apart before it could do useful work.
Shumlak and his collaborators found that a carefully arranged difference in flow speeds could suppress those instabilities. Picture traffic moving at different speeds in neighboring lanes, except the road is a column of electrically charged gas and the desired destination is a fusion reaction. Years of University of Washington research led to the Fusion Z-pinch Experiment, or FuZE. Conway joined Shumlak and Nelson to spin that work into Zap Energy in 2017.
“There is a graveyard littered with over-promises, missed milestones and false starts.”Benj Conway on fusion’s burden of credibility
Conway’s public language about fusion is strikingly short on victory laps. He has said the industry must deliver more than promises and concepts. In a 2023 survey of fusion founders, he predicted that even if everything went well, only one or two companies would build a fusion power plant in the 2030s. Most, he said, would fail to reach scientific energy breakeven, much less an engineering result and a price that could compete in electricity markets. This is not pessimism dressed for a conference. It is an operator’s refusal to let possibility masquerade as a schedule.
A pulse, a pinch, a possible power plant
A product cadence for a physics problem
The founder’s favorite argument is not that Zap has escaped hard physics. It is that the company can confront hard physics more often. In a 2025 interview, Conway asked listeners to imagine spending a billion dollars on an iPhone prototype and building one phone every ten years. No consumer technology could travel from Windows 3 to the modern smartphone at that cadence. Yet long design and construction cycles are normal in major fusion programs.
Zap’s machines are designed to cost single-digit millions and take less than a year to build, Conway said. That does not guarantee the answer. It buys more questions. Hardware can be altered, instruments improved and another campaign run while the last failure is still fresh in the room. Conway calls this rapid iteration a superpower. It is also a very old industrial virtue: make the cost of being wrong survivable.
The company’s 2024 results made that case more concrete. Measurements from its FuZE device, published in Physical Review Letters, showed electron temperatures between 1 and 3 kiloelectronvolts - as high as 37 million degrees Celsius. Only a small group of fusion concepts had crossed the 1 keV threshold. Zap’s machine did so without external magnetic coils. Conway’s response paired satisfaction with commercial discipline: compelling innovation economics, he said, are vital if a fusion product is to arrive on a timescale that matters.
That same year Zap closed a $130 million Series D, taking announced funding beyond $330 million. The roster ranged from climate investors to strategic energy companies and international financial groups. Capital of that scale is flattering, but it also turns an elegant experiment into a crowded obligation. Payroll grows. Subsystems multiply. A brilliant plasma must eventually heat something, survive repetition, produce fuel and submit to maintenance by people who would prefer not to work beside a neutron storm.
The unromantic machinery around the star
Century, Zap’s engineering test platform, is an answer to those prosaic demands. It combines repetitive pulsed power, a chamber lined with flowing liquid metal and measures to protect electrodes. It does not make fusion reactions: ordinary hydrogen stands in for fusion fuel. Its job is to rehearse what happens around a future reaction, repeatedly, until the surrounding systems behave less like a laboratory curiosity and more like plant equipment.
In one campaign, Century fired 1,080 plasma shots over three hours without failure, each at a specified current and into a flowing liquid-metal environment. The Department of Energy certified the milestone in February 2025. Later that year, Zap reported that the machine had reached one shot every five seconds, with 39 kilowatts delivered to the cables leading to its chamber. A plasma pulse had become a rhythm.
“Fusion is not just a plasma problem,” Conway said when the result was announced. “It’s a systems integration problem.” This sounds obvious only after someone pays to build the systems. The glamour of fusion belongs to the instant when nuclei join. The business belongs to every stubborn component that must work before and after that instant, all day, at a tolerable price.
Conway once described the conventional commercialization path as a triathlon: science, then engineering, then commercialization. Zap, he said, was attempting to swim, cycle and run at the same time. The metaphor cheats athletically but works managerially. Plasma teams push performance while plant engineers handle liquid metals and power electronics, and commercial teams cultivate regulators, investors and possible sites. Sequential purity would be tidier. The atmosphere and the power market are not waiting for tidiness.
The founder changes chairs
In April 2026, Conway gave up the CEO title and became Zap’s president. Zabrina Johal, whose background spans nuclear technology, large projects and the U.S. Navy’s Nuclear Power Program, took over as chief executive. Conway’s new brief covers strategy, partnerships and long-term technology development. For a founder, surrendering the principal noun can be a test of whether the institution matters more than the biography. In this case, the move came with an even larger institutional bet.
Zap announced that it would develop compact modular fission alongside fusion, arguing that the two share materials, liquid-metal systems, neutron environments, supply chains and regulatory work. Fission could address nearer-term demand; fusion remains the longer horizon. “Fission and fusion are two expressions of the same underlying physics,” Conway said. He insisted that the addition was not a pivot but an integration.
The strategy creates fresh ways to be wrong. Two nuclear technologies contain more engineering, licensing and capital risk than one. They also create shared work and a nearer path to sold power, especially as data centers swell electricity demand. Conway’s career suggests that he is rarely put off by a proposition merely because its category has acquired an alarming reputation. He wants to inspect the machinery inside the label.
A month after the leadership change, the Department of Energy approved Zap’s preconceptual design milestone for a fusion pilot plant. The report covered an approximately 50-megawatt net-electric module and the long list that transforms a plasma concept into infrastructure: power conversion, tritium, controls, safety, maintenance and a site. It was a design approval, not a working plant. Conway’s own standard demands that the difference remain visible.
The interesting wager is not that fusion can occur. It is that a company can learn quickly enough to make fusion useful.
That is the quieter unity in his biography. A medical student studies a whole system rather than a symptom. A diplomat learns that coalitions determine what ideas can travel. A frontier investor discovers that neglected markets must be judged beneath their headlines. A fusion founder learns that the reactor is only one component of the power plant, and that the power plant is only one component of an electricity market.
Zap’s machines may yet expose a fatal assumption. Conway says as much when he talks about the field’s graveyard. But his company has made a specific, testable claim about how difficult technology progresses: remove unnecessary complexity, put economics beside physics, and keep the learning loop moving. There are grander ways to describe bringing a star to Earth. Few are as useful on Monday morning.