Javier Saldana on Why Energy Transition Fails After Generation | Interview

The energy transition no longer fails at clean power generation. It fails at storing and transporting energy in forms existing infrastructure can actually use.

In this interview, PromEX founder Javier Saldana explains why molecules, not electrons, are the missing link between renewable energy and scalable, reliable power systems.

1. Most energy transition conversations focus on producing more clean electrons through solar and wind. Promex took a different path by focusing on molecules instead. Can you take us back to the moment you realized the real bottleneck wasn’t generating clean power, but storing and transporting energy in a form existing infrastructure can actually use?

Javier Saldana: Looking back, dealing with electrons wasn’t at any time part of my journey because I started in the world of biogas. But as I became fully immersed in the broader energy industry, I realized that if we want to build something that scales globally, the primary challenge isn’t clean power generation, we already know how to do that and there are many options, that keep getting better and better.

The real bottleneck is what happens after generation, specifically how we store energy at scale and move it across vast distances, especially when the sun isn’t shining or the wind isn’t blowing. That’s when it clicked for me. While electricity has revolutionized our quality of life, it remains a difficult medium for long-distance transport and long-duration storage.

Batteries are incredible, but they don’t easily scale to support the grid for weeks or months at a time. Meanwhile, the world already has a massive, trillion-dollar infrastructure built for molecules: pipelines, tanks, engines, and turbines. I realized that instead of trying to force the entire world to adapt to electrons, we should be focused on adapting electrons to the existing world.

2. Many climate technologies treat CO₂ as a liability to be buried through carbon capture and storage. Promex treats it as an input. What convinced you that CO₂ could be economically upgraded into a valuable energy carrier rather than permanently disposed of?

Javier Saldana: For me, it came very naturally, and it mostly comes from my background. I’m not an energy engineer by academic background. I was trained in biotechnology and molecular medicine,and my cofounder is a biochemical engineer and in that world carbon is not seen as a problem to get rid of. It’s the element that makes life possible. More importantly, it’s nature’s preferred way of storing and moving energy. Biology overwhelmingly runs on C–H bond and not by accident. Over hundreds of millions of years, evolution tried many different ways to store energy, and it converged very strongly on carbon–hydrogen chemistry because it simply works better than anything else.

So when people act surprised that our economy still runs on carbon-based fuels, I find that odd. Of course it does. The C–H bond is extremely energy-dense, stable, easy to transport, and easy to store. We still haven’t found anything that beats it at scale.

What always felt wrong to me was not carbon itself, but how we treat it. We extract it, burn it once, and then act as if the only responsible thing to do is to bury it forever. That’s completely opposite to how nature works. Nature loops carbon continuously.

I never saw upgrading CO₂ as a clever business idea (which it is) but I rather saw it as something that had to exist. If humanity wants a world with abundant, reliable energy, carbon cannot be a dead end, it has to be part of a closed loop. The problem was never physics or chemistry; it was that there wasn’t a cheap, compact, practical way to do it. Once you solve that, everything else follows.

That’s what we’re obsessed with at PromEX: closing that loop and turning CO2 back into a usable energy carrier, instead of pretending that permanently disposing of it is a long-term solution.

3. Plasma-assisted catalysis has existed in labs for decades without large-scale deployment. What was the specific technical barrier that finally made this leap possible now, and why couldn’t previous generations of engineers cross it?

Javier Saldana: Yes, plasma-assisted catalysis always worked. That was never the issue. The issue was that it couldn’t be controlled, repeated, or stabilized economically outside a lab.

Previous generations were missing three things simultaneously. Power electronics weren’t good enough meaning you couldn’t run stable high-voltage plasma continuously without destroying the reactor or the catalyst. The systems were effectively blind leaving previous systems with no real-time sensing, no feedback, no visibility into what was actually happening. And computation was too slow and too expensive, which meant no adaptive control, no optimization, and no learning during operation. So plasma stayed a science experiment.

What changed is that all three bottlenecks disappeared at the same time. High-voltage power electronics became cheap and reliable. Sensors became fast, compact, and affordable enough to live inside the reactor. And AI-based control now allows continuous, real-time adjustment of power and flow to keep the plasma and catalyst in a productive regime. Once those constraints collapsed, industrial plasma finally became possible.

4. You describe the M41 reactor as AI-assisted, but plasma systems are inherently chaotic. Is AI in your system acting as passive optimization, or is it making real-time, millisecond-level control decisions to actively stabilize plasma behavior?

Javier Saldana: The goal is not to control plasma. Plasma is inherently chaotic, and trying to micromanage it directly is the wrong abstraction. From first principles, our goal is to combine carbon and hydrogen in the cheapest, most reliable way possible. So the AI is applied where it actually matters, at the system level, not inside the chaos itself.

The AI operates in real time, making decisions on the variables that determine whether the plasma–catalyst system stays productive: gas composition, flow ratios, power input, and operating regime. By keeping those boundary conditions tightly controlled  the plasma naturally stays in a stable, useful state without needing to be “tamed” directly.

5. Traditional chemical plants are largely fixed once built. With Promex, software appears inseparable from the physical system. Does this mean deployed M41 reactors can materially improve over time through data and algorithm updates?

Javier Saldana: Yes, and they have to. If PromEX is going to become the backbone of the world’s energy system, the reactors can’t be static machines.

From day one, the M41 is designed to learn. Every deployed unit generates data on plasma behavior, gas composition, catalyst performance, degradation, and efficiency. Over time, that data reveals very specific hotspots, operating regimes, control strategies, and design parameters, where small improvements create outsized gains in how efficiently we store energy in carbon.

6. Catalyst degradation and coking are long-term failure modes in these systems. How does your AI balance short-term yield optimization with long-term reactor health? Are physical constraints embedded directly into the learning process?

Javier Saldana: We never treated coking as a surprise failure mode, it’s one of the important  things we design around. At the hardware level, the M41 is built to minimize the conditions that promote coke formation. But we understand that we should not rely on design alone. The AI is explicitly trained to treat reactor health as a first-class constraint.

We instrument the system at multiple points, before, inside, and after the reactor, so we can detect early signatures of catalyst stress, carbon buildup, and efficiency drift long before failure occurs. From that data, the system can infer with high confidence which operating conditions are driving degradation.

Physical constraints like temperature, residence time and gas composition are embedded directly into the control logic. The AI actively looks to preserve catalyst life while maintaining the highest sustainable efficiency.

7. Energy economics has long favored scale through massive centralized plants. Promex is betting on modular, distributed production. What gives you confidence that mass-manufactured, decentralized reactors can outperform fully depreciated centralized infrastructure?

Javier Saldana: Large centralized plants dominated historically for good reasons, like high capital intensity, custom engineering, manual control, and the need to amortize infrastructure over long lifetimes. Under those constraints, scale reduced unit cost.

What has changed is not the physics of energy, but the economics of manufacturing, control, and deployment. Mass-manufactured systems follow predictable learning curves, while centralized infrastructure is largely static once built. Software-defined operation allows performance to improve over time without rebuilding the asset.

Even fully depreciated centralized plants still carry significant ongoing costs from moving inputs and products over long distances. I’ve seen this firsthand when I used to run a trucking company, it’s expensive, lossy, and operationally complex. Distributed reactors placed directly at the source of carbon and energy demand eliminate much of that overhead.

PromEX’s approach scales through manufacturing repetition rather than asset size. Thousands of identical reactors can be produced, deployed, and iterated faster than a single bespoke facility can be designed and permitted. Over time, learning effects, reduced logistics, and software-driven optimization outweigh the traditional advantages of centralized scale.

Centralization was the correct solution for past constraints. Under today’s constraints, modular, distributed systems are economically competitive, and in many cases, like ours,  superior.

8. While much of the industry is betting on hydrogen, Promex chose synthetic methane. Is this a rejection of the hydrogen narrative, or a pragmatic response to infrastructure reality over the next decade?

Javier Saldana: We’re not rejecting the hydrogen narrative. On the contrary, hydrogen will likely be one of the next trillion-dollar markets.

But every time I talk about hydrogen with experts, I hear the same thing over and over: it has enormous potential, but the moment you need to store it, transport it, or handle it at scale, it becomes extremely difficult.

As I’ve said before, nature has been pointing to the solution all along. Hydrogen becomes truly useful when it’s bonded to carbon. Every form of life on Earth is essentially hydrogen attached to carbon. That’s why I’m convinced our contribution to the hydrogen economy is to make storage, transport, and use dramatically easier than hydrogen could ever be on its own.

9. You are building a decentralized energy vision while working closely with utilities like Alabama Power. How do you reconcile decentralization with centralized grid operators, and where does M41 ultimately live in that system?

Javier Saldana: Centralized grid operators exist to guarantee reliability, safety, and coordination at scale. That role doesn’t go away. What changes is where and how energy is converted and stored. M41 lives at that conversion layer.

M41 was designed from day one to be flexible. It can live inside a utility-scale deployment, fully integrated into a centralized operation, or it can live at the edge, next to a biogas facility or an industrial carbon stream. Same machine, same logic, different scale.

So decentralization here doesn’t mean bypassing utilities or fragmenting the grid. It means pushing energy and carbon conversion closer to where the physics and economics make sense, while centralized operators continue to run the system as a whole. M41 ultimately lives as a modular building block inside a grid that is still centrally coordinated, just far more adaptable.

10. AI itself is now facing power constraints as data centers scale. Have you considered a closed-loop future where excess renewable energy is converted into synthetic fuel on-site and later used to power AI infrastructure?

Javier Saldana: Yes, absolutely. And that use case is one of the most exciting ones for us.

PromEX is serious about enabling a future where humanity lives better and more, and AI is clearly part of that future. From the beginning, we recognized that data centers are becoming one of the most constrained energy systems on the planet and the M41 is a natural fit for that problem. It allows excess or stranded renewable energy to be converted on-site into a dense, storable fuel. That energy can then be used to power AI infrastructure reliably. 

11. If we fast-forward to 2030, where should we expect to physically encounter Promex technology first? Industrial sites, utility infrastructure, remote communities, or something closer to everyday life?

Javier Saldana: By 2030, PromEX operates as a global, distributed network of micro fuel factories. M41 reactors are deployed across industrial sites, utilities, and energy-rich locations, forming a physical energy layer, we like to explain it much like how Starlink created a new orbital network layer for connectivity. Individually, each unit is small. Collectively, they function as a large, coordinated system that converts CO2 and energy into storable fuel at scale.

At that point, PromEX is a global mesh of micro-factories that turn carbon into energy wherever it exists, quietly running in the background. Most people won’t see it, but they’ll benefit from the stability, flexibility, and abundance it enables, much like NVIDIA operates today: not as a consumer-facing company, but as the invisible infrastructure powering nearly every major advance in computing and AI.

That’s the vision: a planetary-scale energy network built from modular machines, not monolithic assets, doing for fuels what distributed networks did for communication.

12. Promex inevitably evokes Prometheus, the figure who gave fire to humanity. If your technology succeeds, it doesn’t just sell fuel. On a personal level, what does it mean to you to be building something that could permanently alter how civilization accesses energy?

Javier Saldana: It’s funny to compare us to the myth of Prometheus. If we could magically sit down and have a conversation with him, I think we would immediately aggree that expanding humanity’s ability to live better and more always comes down to energy.

Prometheus understood it, when energy becomes more abundant, more reliable, and more accessible, progress accelerates across every dimension of society.

Where we would disagree with the him is in strategy. Prometheus had to steal fire from the gods and was punished for it forever. That’s a zero-sum story. What we’re building at PromEX is the opposite. We’re not trying to break the system or take something away from anyone. We’re trying to improve the system so that everyone benefits, large players, small players, and the system as a whole.

On a personal level, I don’t think much about what this means for me as an individual. When PromEX succeeds, it means society has more energy to work with, and that outcome will be the result of many talented people working together over time. That’s the only part that really matters.

If energy stops being the bottleneck, human potential skyrockets. That’s the mission.

Editor’s Note

This interview argues that energy abundance depends less on producing clean electrons than on closing the carbon and energy storage loop at scale.