Why ZeroAvia Founder Val Miftakhov Stepped Down as CEO and Turned His Attention to AI and Energy
- HYSKY Society
- 20 hours ago
- 7 min read
Val Miftakhov is widely recognized as one of the most influential figures in hydrogen aviation. As the founder and longtime leader of ZeroAvia, he accomplished what many thought was impossible: turning a radical, clean-energy concept into reality by getting hydrogen-electric aircraft off the ground. But today, a major new chapter is unfolding. After stepping down from his role as CEO, Val is shifting his focus toward a stealthy new frontier at the intersection of energy and artificial intelligence.
To truly understand what drives an innovator like Val, you have to look past the standard corporate blueprints. His journey reads like an epic timeline, stretching from a childhood in Siberia during the final years of the Soviet Union to studying physics at Princeton, revolutionizing the electric vehicle smart-charging industry, and becoming a high-end endurance athlete.
But when you strip away the polished corporate press releases, the conversation goes somewhere completely unexpected. When we look at the future of advanced technology, we are ultimately looking at how we combat humanity's deepest evolutionary fears: biological decay, memory loss, and absolute extinction.
What Actually Happened at ZeroAvia?
Everyone in the hydrogen aviation space saw the press release at the end of May and wanted the real, unfiltered story on what went down. Val was completely open about the structural reality behind the change. After years of running multiple massive product programs simultaneously—including liquid hydrogen development, high-temperature fuel cells, and large aircraft flight testing—the reality of a tightening economic climate forced a hard decision.
ZeroAvia had to contract its operations and focus sequentially rather than trying to conquer every engineering vertical at once. As a result, they scaled down their US presence and consolidated the core weight of the company into the UK, specifically Kemble. Their singular milestone right now is hyper-focused on one thing: getting the product certified. Val realized that since the day-to-day execution is now entirely local to the UK team, it made perfect operational sense to hand the reins to a new lead who is local to that team and has deep, specialized experience delivering certified products. He remains the largest shareholder and on the board.
The political and financial climate played a massive role in this shift. ZeroAvia's manufacturing loan application program with the US Department of Energy was abruptly paused and cut down due to federal spending reviews and the new administration's DOGE effects. Venture capitalists started sitting on their hands because they could not calculate what the administration would do next. To complicate matters, capital markets systematically drifted away from sustainability technology over the last eighteen months to chase the explosive new shiny object: AI and data centers.
The Stealth Move Into AI and Grid Energy
This macro-economic funding shift brings us directly to Val's next venture. He is currently building a stealth AI and energy startup. While he is keeping the exact secrets under wraps for now, the engineering reasoning involves deploying machine learning models to solve massive complexities around grid balancing. This new project shares some structural DNA with his pre-ZeroAvia startup, eMotorWerks, which built highly successful smart-charging networks for electric vehicles, but Val notes that this new AI frontier is on a much grander scale. He is deploying the exact same playbook he used in aviation, keeping the company fully underground until they have operational prototypes ready to scale toward the end of the year.
The 90% Regulatory Reality and the Hydrogen Sandbox
Building deeply regulated hardware comes with brutal structural roadblocks. When asked what percentage of ZeroAvia's current work is dedicated strictly to policy and government affairs, most industry insiders would guess maybe half the workload. Val revealed the staggering reality: it is ninety-plus percent so the technology works. Most of the company's current efforts are focused on getting the tech certified.
Regulatory agencies are struggling to handle novel technology. Organizations like the FAA are structurally optimized for products using existing technology bases like a new turbine engine where ninety percent of the compressor, blade, and fuel architecture is already understood. The FAA isn't structured for companies to show up with a clean powertrain that is nearly one hundred percent new.
In the aerospace industry, the historical golden rule is that you do not ask a regulator, you tell them, because asking questions sets you up for a bureaucratic forever situation.
To bypass this gridlock, the UK Civil Aviation Authority took proactive action and established a hydrogen sandbox. This flexible framework allows innovators to break out of traditional slow-motion constraints and iterate rapidly with safety officials to co-create what the new compliance rules should actually be. The FAA needs to build something similar.
Nature's Spaghetti Code and the Reality of Mind Uploading
Val’s unique background as a high-energy physicist and software developer leads to a brilliant, mechanical view of human biology. In the programming world, spaghetti code refers to a system that has been patched, modified, and layered so many times to fix individual bugs that the codebase becomes an undecipherable, unstable mess. Val points out that nature is the ultimate master spaghetti coder. Evolution has spent billions of years throwing uncoordinated patches at our biological bodies, creating a system so incredibly over-complex and fragile that it is guaranteed to break down and decay.
Under this reasoning, digital mind uploading is not science fiction; it is the ultimate shortcut and a master structural refactor of human biology. Val is entirely onboard with the concept of scanning our biological brains, mapping the synapse connection weights, and porting that model into the digital domain to let the failing biological shell go.
When people experience existential dread over the idea of a digital copy, fearing it wouldn't truly be "them," Val counteracts that anxiety with a beautiful analogy about human sleep. Every single night, our conscious minds completely shut down, offload, and lose the continuous thread of identity. Every morning, we reboot and pick the thread right back up without questioning our existence. Uploading your consciousness and rebooting it in a cloud environment is conceptually no different.
LLMs, Home Labs, and the Human Program
This connection between advanced code and human identity is becoming clearer through everyday artificial intelligence. Val has been running a home tech lab in his garage, setting up a fully localized, custom AI workstation. By physically pulling the internet connection plug, he demonstrates to his kids that a distilled, functional version of all human knowledge can run completely isolated in their house without a web connection.
Watching an advanced Large Language Model learn through self-reinforcing data loops and weight adjustments reveals that computers are learning in a manner structurally identical to human children. We are biological computers running on predetermined neural weights; our brains are simply a few orders of magnitude larger than a trillion-parameter digital model.
This realization provides a powerful lesson in emotional intelligence that Val credits to his wife, Julia, who served as the critical CFO and operational co-lead for his previous startups. Getting genuinely angry at another human being is just as illogical as screaming at Claude or Visual Studio when an API throws a runtime error. Humans are simply running their internal software based on the biological weights of their upbringing and environment. Internalizing this perspective keeps you entirely zen. You can remain completely calm internally while choosing to project an external emotional signal solely as a necessary data input to let someone know a behavioral change is required.
The Universe as a Docker Container
This intersection of physics and software even reframes how we look at the concept of God and the origins of creation. Whether an individual identifies as deeply religious or takes a purely scientific, atheistic approach, everyone is forced to establish a baseline axiomatic postulation at the very beginning—a core definition of an origin point or intelligent creator.
Val uses a brilliant cloud infrastructure analogy to explain why the ultimate answers remain out of reach. In enterprise computing, a Docker container is used to fully isolate an application and its processes from the host machine and the outside environment for security. Our entire universe might simply be trapped inside a cosmic Docker container. We can run our calculations, build civilizations, and test physics equations inside our immediate sandbox, but we are structurally blocked by the architecture from ever penetrating or communicating with the outer machine or parallel containers.
Who Will Drive in Twenty Years?
This entire philosophical thread ties directly back to how we view our immediate, everyday choices. David Smith, the CEO of Robinson Helicopter Company, posed a targeted question regarding the future of ground transit: What kind of car will we be driving in twenty years?
Val's reasoning completely upends the status quo: he won't be driving at all, and he hasn't owned a vehicle in six years. Personal car ownership is a deeply flawed model because vehicles are highly underutilized, expensive assets that sit completely idle ninety-five percent of their lifespans. In the coming decades, point-to-point autonomous cybercabs will handle transit entirely, maximizing operational efficiency and removing a massive carbon burden from the planet.
This leaves us with the ultimate question that Val has passed along to the next innovator: In a future world of automated abundance where AI and robotics handle our baseline survival, what will drive human satisfaction? How do we keep our minds—and eventually, our digitally uploaded selves—meaningfully engaged so we don't end up with systemic, digital riots on our hands?
The true goal of modern innovation isn't just about building cleaner engines or faster processors. It is about identifying the ultimate limitations of our current biological and structural programming, and having the courage to rewrite the code.
To stay connected with the pioneers rewriting the rules of technology, aviation, and energy, make sure to subscribe to the HYSKY Pod. If you want to join the conversation live and collaborate with the minds shaping these frameworks, head over to our main site and register for the upcoming Flying HY Conference!
