Building the
pickup truck
of the sky.

The Core Thesis

The next generation of great companies will deliver on the sci-fi promises of the mid-20th century: humanoid robots and general intelligence freeing human effort to be expended on higher value work; colonizing the solar system; energy too cheap to meter; and – most stereotypical of all – flying cars. Obviously, many of these promises are beginning to be fulfilled. SpaceX has made access to space cheap enough that a multiplanetary civilization is within sight. Tesla, Figure, and AI in general are capturing the value of humanoid robots and general intelligence. Valar Atomics is on the road to “energy too cheap to meter”. Flying cars remain conspicuously absent - the promise of the next step-function improvement in human mobility has yet to be realized. The first part of this story is, then, a tragedy.

Why No Flying Cars?

The first manned flight occurred in 1783 when two men flew in a hot air balloon in Paris. 120 years later, on December 17, 1903, Orville Wright piloted the first heavier-than-air flight. Wright’s flight sparked a pace of innovation almost unimaginable today. Orville flew the first powered aircraft and lived long enough to see Chuck Yeager pilot the X-1 through the sound barrier in 1947. Barely more than two decades after that, Neil Armstrong stepped onto the moon. In a single man’s lifetime, aviation progressed from twelve seconds over the dunes at Kitty Hawk to supersonic flight.

Shortly thereafter, humanity reached the moon.

And then, for no apparent reason, progress in personal aviation ground to a halt. In place of fundamental innovation, incrementalism took court, with engines, materials, and manufacturing only getting slightly better.

For today’s pilots, modern aircraft are nearly identical to the machines their grandfathers would have bought and flown.

The Common Explanations

There are familiar explanations for this stagnation:

•Increasing regulatory burden and longer certification timelines.

•Product-liability litigation.

•The exhaustion of the obvious technological breakthroughs—the “low-hanging fruit.”

•An industry-wide shift from maximizing performance toward maximizing safety.

•The feedback loop of decreased units sold requiring higher price per unit, further decreasing demand and unit count sold.

All of these are defensible reasons, but none get at the root – and some of them are just symptoms of the real problems.

The First Structural Mistake: Optimizing for Mechanical Efficiency

The conventional airplane and helicopter are extraordinarily refined solutions to the problem they were designed to solve. Fixed-wing aircraft are an extraordinarily efficient way to travel through the air. Helicopters are arguably the most mechanically efficient way to create a vehicle capable of replacing the point-to-point utility of a car: one large lifting rotor, mechanically driven from one combustion engine. If the objective is maximizing mechanical efficiency, this architecture is difficult to beat. However, mechanical efficiency is not the sole variable of merit in a transportation system.

Helicopters concentrate enormous amounts of flight-critical functionality into relatively few components. The rotor system, transmission, controls, engine and associated driveline contain failures that cannot simply be tolerated.

They therefore require those failures to be made extraordinarily unlikely.

The industry consequently adopted the following philosophy: engineer away failure. Use better materials.

Inspect more frequently. Add redundancy where possible.

Improve manufacturing quality. Certify more rigorously. Maintain obsessively. This has produced remarkably safe aircraft, but the fact remains: some failures remain catastrophic. No matter how rigorous the engineering, manufacturing, maintenance or regulation becomes, machines eventually fail. The Titanic was not in fact unsinkable. Occasionally, the component designed never to fail does. If the goal has shifted from maximizing performance to maximizing passenger safety, the more fundamental question should be: Why should the architecture itself not assume that things will fail?

The alternative philosophy is not merely to engineer away failure, but to engineer to accommodate failure. A transportation architecture designed around degradation should be capable of losing motors, controllers, energy sources, sensors or other individual components without losing the aircraft. This philosophy shift requires reconsidering the architecture from first principles.

sidebar: the nuclear lesson

Nuclear energy went through the exact same failure mode as aviation – by optimizing for energy efficiency by mass, the adoption of water moderated reactors became ubiquitous. These reactors are still incredibly safe – so are helicopters – but only at great cost.

The fundamental lesson I learned at Valar is that the architecture shift to a fundamentally more safety resilient design, despite its mechanical efficiency sacrifices, enabled rapid iteration and realized meaningful cost savings – even on a first of a kind.

I want to apply the same lesson to aircraft, and believe it will spark the next century of rapid innovation in American aviation.

The Second Structural Mistake: Aviation Stopped Being Transportation

Just as aviation’s architecture froze, its market collapsed. General aviation increasingly became a product for people who wanted to fly rather than a product for people who wanted to go somewhere.

A conventional light aircraft typically requires the customer to:

•Travel to an airport.

•Hold a pilot certificate and maintain proficiency.

•Accept substantially greater weather and operational limitations than a car.

•Own or rent a second vehicle at the destination.

•Pay acquisition and operating costs dramatically higher than those of the automobile it is theoretically replacing.

The result is a product with extraordinary capability but poor substitution value. The automobile won not simply because it moved faster than a horse. It created an overwhelmingly better transportation system. For aviation to become mass transportation, the relevant comparison cannot be against another aircraft – it must measure up against and clearly win against the car. The customer does not care about lift-to-drag ratio, rotor efficiency or aircraft category. The customer cares about how long it takes to get from where they are to where they want to be, what it costs, how safe it is, how difficult it is to operate, and whether it works reliably enough to become part of ordinary life. A flying car cannot just be a better aircraft; it must be a better transportation product.

The Rise of the eVTOL

The modern multirotor eVTOL industry begins from a much better architectural premise. Distributed electric propulsion makes it possible to design an aircraft in which the failure of a single motor does not imply the loss of the aircraft. Software can control many propulsion units simultaneously. Electric motors are mechanically simple, have a high thrust ramp rate, are easy to isolate and hence naturally lend themselves to redundancy.

The promise of the eVTOL: simplicity, flexibility, and accessibility is well-known: Billions of dollars have been invested, many crafts can fly, and a couple public companies continue to chase the dream of the 1950s. But no company has meaningfully broken through the adoption curve. Our problem today is not simply whether an eVTOL can fly; it is whether the resulting aircraft is actually the right product.

Most contemporary eVTOL programs are fully electric. This design unlocks an elegant propulsion system but introduces severe constraints in range, payload, and utilization. Most eVTOLs compensate by targeting an air-taxi market that reproduces much of the friction of conventional aviation: dedicated infrastructure, professional pilots, fleet operations, low margins, and an experience still fundamentally separate from everyday transportation. Others focus on the ultralight market but have quickly realized the market for flying machines that requires no training is de minimis, especially given the “toy” nature of these aircraft - ultralight aircraft are restricted to 254 pounds, meaning these aircraft lack endurance beyond roughly a dozen miles range.

A handful of companies have recognized the energy-density problem and moved toward hybrid architectures.

These efforts are important because they expose the next layer of the problem: just adding fuel as energy storage does not result in the right flying car.

Zapata’s AirScooter is a clear example – it combines electric propulsion with hydrocarbon power to achieve dramatically greater endurance than a battery-only multicopter. But it still depends on combustion for a substantial portion of its lift. The electric system is not designed to sustain the aircraft in normal hover following the loss of thermal propulsion. AirScooter solves the energy-density problem but at the cost of robust failure tolerance.

Other approaches

Others, such as Skyways and Elroy Air, have similar issues. Both companies use hybrid propulsion for their long range autonomous cargo aircraft. Each demonstrates that hydrocarbon energy and distributed propulsion can be combined into a practical, modern aircraft. Skyways is already flying cargo missions, while Elroy’s Chaparral has demonstrated autonomous hybrid-electric flight, point-to-point cargo delivery, and is now being developed under a U.S. Army program.

But their products are optimized towards different endpoints than our mission. Skyways’ aircraft is a relatively small UAV, designed for long-endurance logistics; Elroy’s Chaparral is a large autonomous cargo platform carrying hundreds of pounds. Each may become successful products in those markets, but neither product provides a navigable path from their aircraft to a simple, affordable personal transportation vehicle. Simply put, neither market nor product instructs each company how to build the car of the sky.

Archer’s Halo represents a final technical approach. It combines a series-hybrid powertrain with tiltrotors, giving it the range advantage of hydrocarbon fuel while retaining electric propulsion. It is an elegant solution to the efficiency problem. But tilt introduces an additional mechanical degree of freedom into the propulsion system: the same rotor must perform one job in vertical flight, another in cruise, and reliably transition between the two.

That creates an architectural question more fundamental and hence more difficult. If a propulsion unit becomes trapped between flight modes – e.g., if the aircraft cannot complete the transition as intended – the failure affects not merely one component, but the orientation and usefulness of an entire source of thrust.

This design reintroduces exactly the kind of flight-critical mechanical state that the distributed-electric architecture allowed us to eliminate.

The central design principle should be that no single failure mode jeopardizes the aircraft’s ability to remain airborne. Motors can stop. Controllers can die. Generators can shut down. Sensors can disagree. Components can seize. The aircraft should degrade around those failures rather than asking every critical mechanism to work correctly until touchdown. The lesson of the eVTOL era is therefore not simply that batteries are insufficient.

Battery eVTOL discovered the right propulsion topology but paired it with the wrong energy storage. Some hybrid aircraft solved the energy problem but compromised the failure mode philosophy. Others built credible aircraft around markets that do not converge toward personal aviation. The remaining opportunity is to combine all three: distributed electric propulsion designed for inevitable failures; hydrocarbon energy used without making the combustion system flight-critical; and a product and market roadmap that compounds directly toward personal mobility. Our combination – not just eVTOL, and not just hybridization – is how we achieve our grandparents' dream of the flying car.

The Three-Legged Stool

For our dream of flying cars to become a real mass-market product, three tenets must be true simultaneously.

It must begin with a real market. It must provide real utility. And it must reach a reasonable cost. A failure on any one of these dimensions means the product will not mount the adoption curve.

Leg 1: Real Market

Facebook started with Harvard students. Tesla started with a sports car. Amazon started with books. Do not despise the day of small beginnings – be comfortable building meaningful market penetration in a directionally aligned real market – for us, that means the following unique market penetration strategy: We will design and build a Light Sport (2 seat manned aircraft) to be certified under Part 22 regulations, published in July 2026 via the Mosaic Program. This regulatory path is unique, because it eliminates the requirement for a Production Certificate (PC) to manufacture the aircraft. This drastically cuts the time to market for a Part 22 Type Certified aircraft.

Along the way, we will apply this manned aircraft design to unmanned applications via Part 108 – the Beyond Visual Line of Sight / BVLOS application for a near identical architecture restriction found in Part 22. This intermediary step will enable reps in the arena while minimizing risk of an early product to human occupants, similar to the Falcon 9 development path.

The Light Sport category provides a meaningful beachhead market for the company. There are half a million individuals in the US licensed to fly in at least this category, and it represents the minimum barrier to entry for an FAA pilot’s license.

The Light Sport Aircraft is to us what Harvard was to Facebook, what books were to Amazon, what the Roadster was to Tesla.

Leg 2: Real Utility

The ultimate product must replace meaningful automobile trips. Only 3% of car trips today are greater than 50 miles, but those 3% of trips account for nearly a third of all miles driven - this is the first place to target. That requires useful range, useful payload, point-to-point operation, high dispatch reliability and a refueling or recharging experience compatible with ordinary life and existing infrastructure. Pure battery-electric propulsion currently imposes an unacceptable trade between those requirements.

The flying car should therefore be series hybrid. Initially, it will be a hybrid between piston powered level flight and electric lift, and eventually a hybrid built on an in-house design turbogenerator coupled with pure electric thrust.

Ultimately: Electricity should move the aircraft. Fuel should carry the energy.

The result preserves the safety and controllability advantages that made multicopter eVTOL possible without accepting the range and payload constraints that have prevented it from becoming a true car substitute.

Leg 3: Reasonable Cost

A mass-market aircraft cannot be designed like a conventional aircraft and simply manufactured at greater scale. Cost must be attacked architecturally. Every moving part creates manufacturing cost, maintenance cost, failure modes, inspection requirements and supply-chain complexity. Every unique component increases tooling and inventory. The aircraft should therefore be relentlessly simple – crucially, this simplicity is only possible when combined with an inherently failure-tolerant architecture from the beginning.

Ford did not create mass mobility by building the world’s finest handcrafted car. He redesigned the product and the manufacturing system around scale.

SpaceX did not make launch cheaper by marginally improving the existing expendable-launch model. It changed the architecture around reuse and then built the manufacturing and operating system around that premise.

Valar’s analogous insight is that cheap nuclear energy is not achieved merely by making conventional nuclear projects execute somewhat better; the system itself must be reconsidered around cost and repeatability – similarly enabled by reconsidering the architecture and choosing a system that does not have catastrophic single-point-of-failure modes.

The flying car requires the same kind of architectural reset – you must make an aircraft architecturally safe enough that it can tolerate relentless simplicity.

The Roadmap

The path to the flying car looks something like the following, with each step informing the next and with reality and rapidly developed test data being the final arbiter. The roadmap is one product architecture moving progressively toward the full vision, with only small and calculated compromises made along the way.

Stage I – Prove the Architecture

Build and fly the simplest aircraft capable of proving the distributed-electric, hybrid-electric, controls and failure-tolerance architecture. Demonstrate that the core technical thesis works.

Accumulate flight hours aggressively. Break things. Learn what fails. Design the system so those failures are survivable.

Stage II – Build the Beachhead Product

Turn that architecture into a commercially useful aircraft operating in a defined initial market via Part 108. Use autonomous missions to generate utilization and credibility in the market. Develop manufacturing, maintenance, logistics, software and operational capability around an aircraft that is already converging toward the eventual passenger vehicle. The objective at this stage is not primarily revenue, it is market capture and pace of data accumulation.

Stage III – Deliver the Two-Seat Aircraft

Certify and sell the architecture as a two-seat light-sport aircraft. This is the first manifestation of the actual mission. The aircraft should begin collapsing the historical friction of personal aviation:

•Lower acquisition cost.

•Lower operating cost.

•True point-to-point travel for numerous real trips.

•High redundancy.

•Useful range.

At this point, the customer begins buying transportation rather than aviation.

Stage IV – The Turbogenerator

While a piston internal combustion engine will enable the company to rapidly progress through the first three stages, it will not deliver on the ultimate goal. The only powerplant capable of delivering the flying car will be an in-house developed turbogenerator. Integrating this product and rebuilding the aircraft around it will unlock the final stage of personal air mobility.

Stage V – Mass Adoption

The flying car will progress through a similar adoption curve that Tesla’s full self-driving worked through.

Initially, operating the aircraft will require the appropriate pilot training. Hardware and software capable of autonomous operation will work its way in early, but will not supersede or replace human pilots for many years. Eventually, however, the aircraft will fly itself.

Solving the self-flying problem for the masses is a far technically simpler problem set than self-driving.

Proximity and total number of nearby obstacles is radically reduced by vertical space and the elimination of reliance on road infrastructure. Air travel will be orders of magnitude safer than today’s 2 lane highways, and someday we will look back on the 2 lane state highway like we look back on open sewers in cities a few centuries ago – an insane and intolerable risk when the sky has enough space to reduce the risk of collisions between passenger-carrying vehicles to near zero.

The mid term measure of success for our flying car will be the number of miles flown that otherwise would have been driven. The long term measure of success will begin with miles travelled that used to be impractical – the world becoming smaller.

The elevator’s biggest consequence was not that people stopped climbing stairs. It made the skyscraper feasible. The automobile’s greatest achievement was not replacing the horse buggy - it radically expanded the geographic and economic radius of human life. This expansion was not linear (once road infrastructure caught up to the car’s needs), and this next step function will not be linear either. An effective travel radius expands your accessible area with the square of the radius, meaning just a 4x increase in radius yields 16x the accessible area. The economic benefits of this change are as meaningful as the upgrade from horse buggies to the early automobile.

The elevator changed the vertical geography of cities. The automobile changed the horizontal geography of cities. Personal aviation will eventually change the geographic radius of civilization itself.