Also, I am currently on paternity leave and had to write this in between cage matches with a screaming infant (highly recommend fatherhood, unironically) so if my writing is not up to snuff please have mercy.

That’s enough disclaimers. Without further ado…

The Electric Stack and Its New Layer

There is a revolution underway in human civilization, one that transcends any single invention or headline. Instead it accumulates, layer by layer, in the form of computer chips, cheap and high-capacity batteries, power electronics, and electric motors (or, more specifically, the rare earth magnets at their core). Taken together, these form what Sam D’Amico and Packy McCormick have termed the Electric Stack: an interlocking set of technologies that are redefining the industrial basis of modern power.

Just as digital technologies, from transistors to the internet, compounded in capability and strategic importance over decades, the electric stack is compounding now. Each layer enables and accelerates the next. New electric motors and cheap battery storage make electrified vehicles viable. EV’s create demand for fast charging, which requires advances in power electronics. The layers are not independent; they are mutually reinforcing, and the nations that hold structural positions within them will have massive influence over the economies and militaries of the twenty-first century.

This is why the current geopolitical contest over batteries, semiconductors, and EV’s is more than conventional industrial competition. It is a struggle over owning the stack, and to this point the United States has failed to act in time to definitively claim a single layer.

But the stack is not finished. Just as the Internet did not emerge immediately from the void after the first transistor was fabricated, the electric stack is still growing. In this piece I will argue that there is another, under-appreciated layer, one which is rapidly advancing and, crucially, is not yet dominated by any one nation: directed energy. Every layer of the electric stack to date has concerned itself with the flow of energy: through wires, through cells, through circuits. What no layer has yet addressed systematically is energy that aims: that can be concentrated, steered, and precisely delivered without a physical medium connecting source to target.

In short: the next frontier is Big Ass Lasers12. Hell yeah.

Why Big Ass Lasers Matter: The Use Case Landscape

The layers of the electric stack share a few notable features: they are general purpose technologies with both civilian and military applications, but are not in and of themselves end products. Instead, they are the building blocks underpinning nearly every modern device, from smartphones to televisions to drones. Lasers fit this description nicely: they are a general purpose technology with value in military and civil applications. Plus, while it may be fun to torment your cat with a laser pointer, lasers are generally an input rather than a final product. They provide critical capabilities in manufacturing, communications, medicine, and space, to name a few. Each of these applications shares a common thread: the ability to do work with light, at a power and precision that no other technology has matched. More specifically, lasers are good for:

Killing Things (Obviously)

The defense applications of high-power lasers tend to dominate the conversation for good reason. Modern warfare is facing a cost-exchange crisis that threatens to bankrupt even the most well-funded militaries on earth. The problem is straightforward: it currently costs vastly more to shoot something down than it costs to launch it in the first place. A Houthi drone assembled from consumer electronics for a few hundred dollars can compel the USS Carney to fire a SM-2, at roughly $2 million a shot. Multiply that across a contested theater and the numbers become catastrophic. There’s a reason the Pentagon has requested a $200 billion Iran War Supplemental after just a few weeks of conflict.

High-power lasers change the math. A laser engaging a target might spend only a few dollars of electricity per shot. It will not run out of ammunition as long as there is power. It engages at the speed of light, meaning no lead calculations and lower miss probability against fast moving targets. It can scale from a warning pulse to a disabling shot to a destructive beam by adjusting dwell time. The US Navy, Air Force, and Army are all actively fielding directed energy programs as responses to adversaries who have figured out how to destroy expensive, exquisite assets with cheap, disposable ones. And that’s actually one of the less important uses, compared to…

Making Things

Laser weapons are awesome, but in aggregate industrial lasers are far more economically significant. They are, in fact, already load-bearing infrastructure for the modern economy.

Consider semiconductors. Every leading-edge chip fabricated today depends on extreme ultraviolet lithography, a process that works by firing a laser at a tin droplet 50,000 times per second to generate plasma, which then emits the EUV light used to etch circuit features smaller than a virus. The machines that do this, built by the Dutch firm ASML, cost upward of $200 million each and are arguably the most complex manufactured objects ever made. Those machines are, at their core, very sophisticated laser delivery systems. When policymakers talk about semiconductor sovereignty, about who controls the commanding heights of the chip supply chain, they are talking (whether they know it or not) about laser capabilities.

While semiconductors may be the most dramatic example, they are not an outlier. Modern automotive bodies are laser-welded to tolerances that resistance welding cannot match. Electric vehicle battery packs are laser-welded because the geometry and thermal sensitivity of lithium cells makes it the only practical option at scale. Aerospace manufacturers use laser powder bed fusion to produce turbine blades and structural components with internal geometries that subtractive machining cannot reach; this is especially significant because these parts aren’t merely better, they literally could not be made otherwise. Medical device manufacturers cut coronary stents — sub-millimeter features in biocompatible alloys — with lasers, because nothing else is precise enough. The global industrial laser market sits at roughly $6–7 billion annually, growing rapidly, but that number dramatically understates the economic surface area it supports. Nobody buys a “laser-cut part.” They buy an aerospace bracket, a medical stent, a turbine blade. The laser is a silent prerequisite, invisible in the supply chain until it isn’t there.

This is exactly the pattern of every prior layer in the electric stack. Rare earth magnets don’t appear on any consumer invoice, but pull them out and the whole industrial economy stops. No normal person spends their day thinking about power electronics, but without them the grid collapses. Lasers are already in that category. Without them, the modern industrial economy ceases to function. Things get even more interesting when you push the power envelope.

Megawatt scale high-powered lasers represent a plausible step change in manufacturing. The barriers that currently make advanced manufacturing expensive are not primarily material costs, they are processing costs. Titanium is not expensive because titanium ore is scarce (it is in fact about 100 times more abundant than copper). It is expensive because titanium is extraordinarily difficult to extract, machine, form, and join with conventional tools. High-power lasers change that calculus: they could cut titanium faster, cleaner, and with less waste than mechanical tooling, and with stronger lasers powder bed fusion could print near-net-shape titanium components fast, increasing production rates for complex parts. Higher powered laser welding could reduce the number of passes needed to join parts. Carbon fiber composites, which currently require hours in an autoclave to cure — a capital-intensive bottleneck that constrains how widely composites can be used — could be directly laser-cured, enabling orders of magnitude higher production rates.

The nature of lasers as a general purpose technology means there are likely infinite more uses than anyone can imagine, but what can be said with some certainty is this: the countries that lead in high-power lasers will be able to make things that countries without those lasers simply cannot, and they will do so faster and at lower cost. As the power frontier advances, that gap will grow ever wider. Eventually we may even…

Ignite the Stars

In December 2022, the National Ignition Facility at Lawrence Livermore fired 192 laser beams at a target the size of a pencil eraser and produced more energy than it put in, the first demonstration of fusion ignition in a laboratory setting. It was, without exaggeration, one of the most significant scientific milestones of the century. The lasers were the entire mechanism: to over-simplify, NIF has turned fusion into a laser power delivery problem. As the field matures and the engineering challenges of commercial fusion come into focus, high-power laser capability could sit at the center of one of the most consequential energy technologies in human history. The countries that lead in lasers will have a structural head start in laser-driven fusion.

Beaming Power Across Distance

Finally, and perhaps the most straightforwardly science fiction-esque application: the ability to transmit energy without wires. Free-space power beaming (delivering electricity via laser to a receiver that converts light back to current) has already moved from theory to active research programs at NASA, DARPA, and several commercial ventures. The near-term applications are probably military: forward operating bases resupplied by drone, UAVs kept aloft indefinitely by a ground-based laser, remote sensors powered without batteries. The longer-term applications are where science fiction becomes reality: space-based solar power collected in orbit and beamed to ground stations, lunar surface energy logistics, and laser-propelled spacecraft.

The Unified Layer

Across all of these domains, the value proposition is the same: precise, efficient, directed energy delivery at speed and at scale. That consistency is what makes high-power lasers a stack-layer technology rather than a collection of unrelated applications. Just as the transistor turned out to matter for computing, communications, and consumer electronics simultaneously, the laser turns out to matter for defense, manufacturing, energy, and medicine. That is exactly what a foundational layer of the electric stack looks like.

The Electric Stack in Strategic Context

Like every other layer of the electric stack, ownership of directed energy technology has massive geopolitical implications. Understanding this in context requires, unfortunately, a tour of how the United States has handled the prior layers. It is a story of missed opportunities, belated recognition, and expensive lessons that we risk not learning.

The Scorecard So Far

The electric stack has been assembling itself for roughly three decades, and the ledger of who owns what layer is not flattering to the United States.

Batteries are the most visible and painful example. Lithium-ion battery technology was developed substantially in US research labs — John Goodenough, the Nobel laureate who did foundational work on the lithium ion battery, spent decades at the University of Texas. Today, Chinese firms CATL and BYD collectively produce more than half the world’s electric vehicle batteries, and dominate leading edge chemistries like Lithium-Iron-Phosphate. China refines roughly 70% of the world’s lithium, 60% of its cobalt, and nearly all of its processed graphite. China also produces the vast majority of the specialty chemicals that are needed for battery cathodes and anodes. The United States is, at present, structurally dependent on a strategic competitor for the storage layer of the electric stack. This did not happen because China out-researched us (at least not initially). It happened because China recognized the strategic value of the full manufacturing supply chain, and used it’s domination of the supply chain to build a world-leading battery innovation pipeline.

Rare earth magnets are arguably worse. The permanent magnets used in electric motors — like the neodymium-iron-boron magnets that make EVs drive, wind turbines spin, and actuators function — depend on rare earth elements that China has spent decades systematically cornering. China currently produces roughly 90% of the world’s processed rare earth elements and a similar share of the finished magnets themselves. This is not a matter of nature playing favorites; it’s almost cliche to point out that Rare Earths aren’t particularly rare. Rather, China’s dominance is an outcome of deliberate industrial strategy, executed patiently over decades while the United States treated rare earth processing as an unpleasant, low-margin business. The leverage this creates is massive, and China has already demonstrated willingness to restrict rare earth exports as a potent geopolitical tool.

Power electronics — the systems that convert, condition, and control electrical power in motors, chargers, inverters, and grid infrastructure — are a more complicated story. The US retains meaningful capability in the design of power electronics, and wide-bandgap semiconductors like silicon carbide and gallium nitride, which are critical to next-generation power conversion efficiency, remain areas of genuine American and allied strength. But manufacturing scale is increasingly Asian, and gallium — a key input for GaN power devices — is another material where China holds dominant refining capacity. China’s dominance in EV’s is increasingly resulting in a lead in high-voltage architectures, enabling its cars to charge far faster than foreign competitors.

Compute represent the most successful story for the US, and it is instructive precisely because it required a crisis to produce a response. The CHIPS and Science Act, passed in 2022, represented a belated but serious recognition that semiconductor fabrication was a strategic asset, not just a market outcome. I know from personal experience how real this investment is, as my wife helped open TSMC’s Fab 21 in Arizona (and … hoo boy do I have some stories from those days). This was major progress, but it came only after years of watching fabrication capacity migrate to Asia, after a global chip shortage exposed the fragility of the supply chain, and after a bipartisan panic about Taiwan’s role as the world’s most critical single point of failure for advanced chips. What’s more, it only addressed the leading edge, leaving trailing edge “analog” semiconductors to be dominated, increasingly, by China. We acted, but we did so late, at insufficient scale, and without the level of focus and rigor that the moment demanded.

The Pattern and What It Means

The through-line across all of these cases is not incompetence or malice (OK, maybe some incompetence). It is a structural difference in how the United States and China think about industrial technology. The United States, broadly, treats technology development as a market process and strategic competition as a matter of military procurement. China treats technology development as statecraft: a multi-decade project of identifying which layers of critical infrastructure it needs to own, and then deploying the full apparatus of the state to make it so. Made in China 2025, whatever its public reception, was an honest and detailed declaration of exactly this strategy.

The result is a consistent asymmetry: the US invents and researches, China manufactures and scales, and by the time the US recognizes that manufacturing is the strategic asset, China’s dominance is unshakeable.

The Next Layer Is Still Open

This is the context in which high-power lasers need to be understood. Unlike rare earths and batteries, no country has yet achieved structural dominance in this space. The technology is maturing, but there is still foundational innovation happening that could reshape the industry.

That will not be true forever. The question is whether the United States will approach this layer with the strategic intent the moment demands, or whether we will find ourselves, a decade from now, writing the same uncomfortable postmortem we are currently writing about batteries.

The Technology: CBC, PICs, and the Path to Practical High-Power Lasers

Having established that high-power lasers matter enormously across a wide range of applications, the natural next question is: if they have so much potential, why aren’t they ubiquitous? Why isn’t the Navy shooting down every incoming missile with a beam of light, why aren’t industrial lasers cutting through six inches of titanium at a fraction of current costs, and why isn’t anyone beaming power from orbit yet?

The answer is that scaling laser power is hard. Like, really hard. Understanding why, and the most promising approaches to solving it, is essential to understanding where the strategic opportunity lies.

The Scaling Problem

A laser, at its core, is a device that converts electrical energy into coherent light — photons that are all the same wavelength, traveling in the same direction, in phase with one another. That coherence is precisely what makes laser light useful: it can be focused to a point, transmitted over long distances without spreading, and delivered with extraordinary precision. The problem is that as you push more power through a laser, several things start going wrong simultaneously.

Heat is the first enemy. High-power lasers, even modern fiber lasers, are not particularly efficient — a large fraction of the input electrical energy becomes heat in the gain medium, the material that actually produces the light. That heat distorts the gain medium, degrades beam quality, and if not managed, destroys the device entirely. This is a fundamental physical constraint that has limited laser systems for decades.

Beam quality is the second problem. A laser that produces a megawatt of power is not very useful if that power is spread across a diffuse, distorted beam rather than concentrated in a tight, controllable spot. Power and beam quality tend to trade off against each other in conventional laser architectures — pushing one up pushes the other down. For applications like missile defense or power beaming, where you need both high power and a tight, precisely aimed beam delivered over long distances, this tradeoff is untenable.

Size, weight, and cost round out the picture. The high-power laser systems that exist today — the ones deployed on naval vessels or tested in airborne platforms — are large, heavy, expensive, and require substantial supporting infrastructure. A megawatt laser weapon system that requires a ship to host it is useful, but shrink that system down so that if fits on a truck or a drone and you’ve transformed warfare forever. The gap between those two things is largely an engineering problem, if a difficult one.

Coherent Beam Combining: The Elegant Solution

This is where things start to get a bit speculative, so feel free to sound off in the comments if you disagree. Based on my research, the most promising approach to cracking all of these problems simultaneously is coherent beam combining, or CBC. The DoW seems to agree, as this is the approach they chose for the HELSI (High Energy Laser Scaling Initiative) program. The core idea is surprisingly simple: instead of pushing more power through a single laser and fighting the resulting heat and beam quality problems, you run many lower-power lasers in parallel and combine their beams into one.

The trick — and it is a seriously non-trivial trick — is that the combination only works if the individual beams are coherent with one another: same wavelength, same phase, precisely synchronized so that their light waves add together constructively rather than canceling each other out or producing a chaotic mess. Achieve that synchronization across an array of emitters, and you get a combined beam that behaves as if it came from a single, much more powerful laser with the heat load distributed across all the individual emitters rather than concentrated in one.

The elegance of this approach is that it sidesteps the fundamental physics tradeoff between beam strength and quality, and transforms it into an engineering problem. The constraints that limit single-aperture lasers — heat concentration, beam quality degradation at high power — become manageable when the power is distributed.

Why CBC Is Becoming Viable Now

Coherent beam combining is not a new idea, researchers have understood the principle for decades. What has changed is that several independent streams of technological progress have matured simultaneously, and their convergence is making practical CBC systems viable.

The most significant of these is the maturation of high-power fiber lasers. Over the past two decades, driven largely by commercial telecom and industrial cutting and welding demand, fiber lasers have improved dramatically in efficiency, output power per channel, and cost. This matters enormously for CBC because more power per individual channel means fewer channels are needed to reach a given combined power level, and the complexity of the phase control problem for CBC tends to scale exponentially with the number of channels.

Phase control electronics have kept pace. The core engineering challenge of CBC is maintaining coherence across all channels in real time: measuring and correcting the phase of each beam faster than vibration, temperature changes, and air turbulence can disrupt it. The control electronics required to do this at high channel counts have become faster, cheaper, and more capable, tracking the same trajectory as computing hardware generally. Emerging AI-driven wavefront sensing and correction algorithms are accelerating this further, enabling adaptive phase control that can respond to disturbances more intelligently than prior rule-based approaches while potentially scaling better with the number of channels.

Photonic integrated circuits (PICs) enter the picture here as a critical enabling technology for the next stage of scaling. A PIC is to light what a conventional integrated circuit is to electricity: a miniaturized platform that routes, manipulates, and controls photons on a chip rather than in discrete bench-top optical components. For CBC specifically, PICs offer a path to integrating the phase measurement and control functions directly at chip scale with the emitter arrays themselves, collapsing what was a complex and expensive assembly of discrete components into something manufacturable at semiconductor scale. This is highly analagous to electronic circuits: early computers were built from discrete components wired together on circuit boards. The Integrate Circuit collapsed that onto a chip and created the basis for the electric stack. PICs are that, but for photonic systems.

PICs are a big part of why CBC could scale from impressive defense programs into a broadly deployable, cost-competitive technology. The systems being demonstrated now rely primarily on mature fiber laser technology and advanced control electronics. PICs are a key piece of the puzzle that may determine whether the technology crosses from specialized military hardware into the kind of manufacturable, widely deployable platform that defines a true stack layer — which is precisely why they deserve strategic attention, and precisely why the manufacturing ecosystem around them matters so much.

Importantly, even if I’m wrong about CBC specifically, all these same technological advancements are likely to feed into whatever takes CBC’s place just as well. Regardless of the specifics, the result will be that lasers get much, much more powerful.

Where the Technology Stands Today

Many-channel CBC systems are already being developed, which is the primary reason why I think they’re likely to win out in the long run. Programs like the aforementioned HELSI are pushing fiber-based CBC toward the Megawatt (MW) scale for defense applications. Commercial silicon photonics are getting a much needed boost from data center interconnect demand, which is already helping mature the underlying PIC manufacturing base in ways that are directly transferable to directed energy applications.

Yet there is still much work to be done. What we do not have is a mature, manufacturable, cost-competitive MW scale CBC system ready for broad deployment across defense and commercial applications. There is still substantial research and engineering work to be done to solve problems with local phase sensing, control algorithms, thermal management, system integration, and more. Even with current technology, there is a gap between laboratory demonstration and the kind of scalable, manufacturable technology that can anchor a new layer of the electric stack. That gap is what the next decade of investment will be racing to close. That race is the strategic contest, and unlike the races over batteries and rare earth magnets, the United States has not lost it yet.

The Strategic Landscape: Where the US and China Stand

So what, exactly, is the state of play for this technology from a geopolitical standpoint? The United States has a long and distinguished history of inventing foundational technologies, failing to notice they were foundational, and then writing alarmed Senate reports about it a decade later. Will directed energy become another entry in that tradition, or is the US position actually defensible this time? The answer, unsurprisingly, is complicated.

What the US Has Going For It

Start with the good news, because there is good news.

The defense research ecosystem is significant. DARPA, the Air Force Research Laboratory, the Naval Research Laboratory, MIT Lincoln Laboratory, Livermore, Sandia, and various university labs provide an incredible concentration of directed energy expertise. The NIF fusion ignition milestone was achieved for precisely this reason: the United States has spent decades building serious institutional knowledge in the kinds of photonic power delivery problems that underlie CBC and related systems. That knowledge lives in people, in cleared facilities, in accumulated experimental data, and in the kind of hard-won understanding of failure modes that cannot be replicated quickly by throwing money at a problem. It is a major strategic asset.

The commercial silicon photonics industry is a second advantage. Data center demand for optical interconnects has driven Intel, Broadcom, and a growing ecosystem of startups to invest in PIC fabrication and design capability. This means the US is not starting from scratch on the manufacturing base critical to next-generation CBC systems. It is starting from a commercially validated foundation that needs scaling and direction, but not re-invention.

Defense procurement as a demand-side anchor is a third advantage, and arguably the most important for the near term. The US military’s willingness to pay for early-generation directed energy systems at prices no commercial market would sustain underpins the development ecosystem. It is the same mechanism that seeded commercial aviation, semiconductors, and the internet. China can replicate a lot of American advantages, but it will struggle to win a contest of spending money against the Pentagon.

Finally, and most importantly: the US enters this race with something it did not have (at least not to the same degree) in the battery and solar contests: a mature, competitive domestic industry in the most relevant existing technology. IPG Photonics, the world’s leading fiber laser manufacturer, is a Massachusetts-based company with its primary manufacturing and R&D campus in Oxford and a newly opened dedicated defense facility in Huntsville, Alabama. Coherent Inc., another major player, is headquartered in Pennsylvania with production facilities across more than a dozen states. NLight, which won the contract for the DoW’s HELSI, is another major player headquartered in Camas, Washington. The US is not starting from scratch on the manufacturing base. That, by itself, is a notably stronger starting position than the one it occupied when China began its battery industry build-out. However…

China still has big advantages

Tell me if this sounds familiar: the gain media in high-power fiber lasers — ytterbium-doped fiber, neodymium-based crystals — depend on rare earth elements that China dominates from mine to finished product. Once again China’s control of Rare Earth refinement gives it major leverage directly upstream of a critical supply chain, leverage which China knows how to use.

What’s more, while the US may currently lead in high-end fiber laser systems, China is not absent from this market. It already has a significant and scaling fiber laser industry through firms like Raycus and JPT, and leading Western manufacturers like IPG have facilities in China — though precisely what share of their production is Chinese is hard to know. China’s existing fiber laser industry means it has a running start on the manufacturing base, supply chain relationships, and cost discipline most relevant to scaling CBC technology.

China also has the advantage in scale: it is by far the largest market for fiber lasers, with approximately 30% of global sales compared to the US and Europe at about 15% each. The most common uses for high powered lasers are mostly in manufacturing: tube cutting, laser welding, materials processing, etc. If the US gets its demand bridge from the military, China gets its from a deep customer base of domestic manufacturers.

That manufacturing base also includes adjacent industries that could cross-polinate with high power lasers, like LiDAR. Chinese firms like Hesai, RoboSense, and others have scaled aggressively in automotive and consumer LiDAR, becoming dominant in what is now a high-volume precision photonics manufacturing market. LiDAR involves precision optical components, beam steering, phase-sensitive detection, and photonic component supply chains that are adjacent to, though not identical with, what CBC systems require. The manufacturing skills don’t transfer perfectly — coherent beam combining demands phase control precision that goes far beyond what automotive LiDAR requires — but the supply chains, the precision optics manufacturing base, and the engineering talent are meaningfully transferable.

As it does with every other layer of the Electric Stack, electricity generation also matters. Manufacturing and testing high-power laser systems is, unsurprisingly, electrically intensive. China’s combination of massive generation capacity, cheap industrial electricity pricing, and state-directed investment gives energy-intensive advanced manufacturing a structural cost advantage that is difficult to compete with. When you are trying to test a 500kW CBC system, electricity costs and availability become non-trivial.

The Most Important Asymmetry

Laid out side by side, the US and Chinese positions look something like this: the United States has a lead in research depth, institutional knowledge from decades of serious defense programs, a commercially validated PIC manufacturing base, and a defense procurement system that can function as a demand-side bridge to commercial scale. China has dominant upstream positions in many of the materials and components most critical to current laser technology, an existing and cost-competitive fiber laser industry, adjacent photonic manufacturing scale from industries like LiDAR, abundant electricity, and a national industrial strategy that treats stack-layer dominance as an explicit objective rather than a market outcome.

The US advantages are meaningful, but they are overly-concentrated in early-stage activities like research, design, and defense procurement. The Chinese advantages are concentrated in exactly the activities that determine whether a technology becomes a foundational layer or remains a specialized capability: materials, components, manufacturing scale, and industrial strategy. This positions it well to take advantage of any paradigm shift in the industry: even leading American manufacturers like IPG could be disrupted if a major technological shift occurs, akin to how Chinese auto makers used the transition to EV’s to leapfrog incumbents.

Thus we have an uncomfortably familiar pattern. The US invented the solar cell and foundational battery chemistry. What it did not do was treat the manufacturing layer as strategically important, at least until someone else already owned it. The question for directed energy is whether the lesson has actually been learned, and will we react before this industry is in crisis, or will we be reading about China’s dominant position in fiber laser and PIC manufacturing in a Senate report circa 2035.

The window is open, but the fiber laser market share data, the history of rare earth export controls, and adjacent industries like LiDAR are all telling the same story — and it is a story about a country that has studied the electric stack playbook carefully and is running it again. If we want to maintain a lead in this new layer of the electric stack, we need to approach it with the kind of deliberate, intentional strategy we lacked for the previous layers. Here’s what that strategy might look like:

Policy Recommendations: Claiming the Photon Layer

1. Increase and Coordinate R&D Investment

The US directed energy research ecosystem is genuinely world-class, but also fragmented. DARPA, DOE, NSF, NIST, and the service research laboratories are all running their own programs, producing research that is excellent, and isolated. More funding is key, but it risks being rendered ineffective without a unified national roadmap that aligns investment across agencies around shared milestones, prevents duplication, and accelerates the path from laboratory demonstration to manufacturable system.

The National Nanotechnology Initiative, launched in 2000, provides a reasonable template: a cross-agency coordination body with a shared strategic vision, common metrics, and enough institutional weight to actually influence budget priorities. It was imperfect, as all such bodies are, but by most accounts it accelerated the translation of nanotechnology research into commercial and defense applications. A National Photonics and Directed Energy Initiative modeled on that framework would give the ecosystem a coordination layer.

2. Secure the Supply Chain

The supply chain problem in directed energy has three distinct layers, each requiring a different intervention.

Upstream, the rare earth problem is urgent and tractable within existing policy frameworks. The gain media in high-power fiber lasers depend on ytterbium, erbium, thulium, and other rare earth elements that China dominates. The fix is not complicated: domestic mining and processing investment, allied sourcing agreements with Australia, Korea, Japan, and others, and integration of laser-critical rare earths into existing critical minerals frameworks. The goal is to ensure that no single export control decision in Beijing can switch off the American directed energy industry overnight. This will become all the more crucial as Directed Energy Weapons become a staple of defense technology. Nobody wants to go into battle against the worlds dominant drone manufacturing power and suddenly have their supply of drone-killing lasers disappear.

Midstream, the component supply chain — optical fiber, laser diodes, PIC fabrication capacity — is probably most amenable to a CHIPS Act model. Targeted incentives, R&D subsidies, and defense procurement can anchor domestic and allied manufacturing capacity in these components before Chinese manufacturers achieve the same kind of cost-driven dominance they have established in solar panels and battery cells. This is, critically, not about building a hermetically sealed domestic supply chain. It is about ensuring enough allied manufacturing depth that cost competition from Chinese suppliers does not hollow out the laser industrial base the way it hollowed out solar manufacturing.

Downstream is probably the most difficult, but nevertheless important: civilian proliferation of directed energy technology ultimately requires a domestic industrial base capable of absorbing and deploying it. Manufacturing equipment suppliers, medical device manufacturers, system integrators — these are the customers that will determine whether directed energy becomes a true stack layer or remains a specialized defense technology. Building that downstream base is a larger industrial policy challenge than any single initiative can address on its own, but policymakers should be clear-eyed that the upstream and midstream investments will only achieve their full strategic value if the downstream civilian customer base exists to create the demand that drives manufacturing scale. This is another area where allied-scale can help, taking advantages of the strong industrial bases in countries like Japan, Korea, and Germany.

3. Use Defense Procurement as a Demand-Side Bridge

The commercial market for high-powered directed energy systems is clear, but does not yet exist at the scale required to justify the manufacturing investment needed to drive costs down to the level at which the commercial market will exist. This is the classic chicken-and-egg problem of deep technology commercialization, and the US government has a well-tested tool for breaking it: structured defense procurement commitments that de-risk technological innovation and eat the high costs of developing that into viable new products.

The commercial space industry is probably the most recent and instructive example. NASA’s Commercial Crew and Commercial Resupply programs provided the early revenue stream that allowed SpaceX to invest in manufacturing infrastructure and innovation that has driven launch costs down by an order of magnitude, spawning a commercial space economy that was unthinkable fifteen years ago. The same logic applies here. Air Force directed energy programs, Missile Defense Agency investments, and Navy shipboard laser systems should be structured not merely as procurement contracts but as deliberate demand-side anchors — designed with an eye toward driving the cost curves and manufacturing scale that will eventually make civilian markets viable. This is precisely what the US defense procurement system has done, at its best, throughout history.

4. Invest in the Workforce Pipeline

Technology competition is ultimately talent competition, and the directed energy talent pipeline is undernourished relative to its strategic importance. Computer Science and AI attract extraordinary investment in fellowships, university programs, and corporate recruitment, investment that reflects the foundational importance of those fields. Photonics and directed energy engineering attract a fraction of that attention despite being, arguably, comparable in importance to the next generation of the electric stack.

The intervention here is straightforward: targeted graduate fellowships in photonics and directed energy engineering, expanded investment in national lab resources for researchers, and dedicated program funding at universities with existing photonics strength. The goal would be to redirect some of the talent that would otherwise graduate into the depressed CS labor market into becoming the engineers that staff the research programs, defense contractors, and commercial companies that a serious directed energy industry requires. Talent gaps are slow and expensive to close. The time to act is before the gap becomes acute, which is now.

5. Deliberately Diffuse Defense Innovations into the Civilian Economy

The history of successful US industrial policy is largely a history of defense-funded innovations that achieved their full strategic and economic potential only when they escaped into the civilian economy. The internet was ARPANET until it wasn’t. GPS was a military navigation system until it became the invisible infrastructure underlying a trillion dollars of civilian economic activity. Semiconductors were funded by defense contracts until they became the foundation of the modern economy. In each case, the transition from defense-sequestered capability to civilian stack layer was what created the manufacturing scale, the cost curves, and the ecosystem density that let the technology truly flourish.

High-power lasers need to make that transition. Defense classification, ITAR restrictions, and the structural incentives of defense procurement can easily push toward keeping the technology inside the bubble of cleared contractors and military programs where it is useful but limited in scale. Without deliberate policy, directed energy risks follows that path; it will be a capable, expensive defense technology, and nothing more.

Importantly, none of these recommendations require me to be right about CBC to be valuable. Investment in workforce, supply chains, and technological diffusion will benefit whatever technical solution wins the day. That is the essence of smart industrial strategy. China understands this, and has run it’s own strategic playbook over and over again to claim the infrastructure of the modern economy layer by layer. The United States should meet that strategy with one of its own, and finally own a piece of the Electric Stack.

Thanks for reading! If you enjoyed this post, consider subscribing. This is a personal blog that I write when I have something to say that doesn’t fit in my Twitter feed. I used to be a Tech and Trade Staffer in the US Senate so I tend to write about politics, semiconductors, technology, industrial policy, and any other random topic I find interesting. All posts are free, I don’t make money off this blog.

Subscribe now

Share

If I sound insane, it’s probably because I am. That doesn’t mean I’m wrong, though. Yes, by any reasonable metric we’re in the middle of the greatest escalation of gerrymandering in our nations history. Yes, even states that had passed nominal limits gerrymandering, like California and Ohio, are undoing those limits in an escalating tit-for-tat where states compete to draw more of the opposition out of their seats and yes, this is bad. Very bad. Like, potentially existentially bad for our democracy. I don’t want to live in a country where politics is reduced to partisan battles over who can draw the most absurd maps, and neither should you. If I wanted to live in a dysfunctional third world pseudo-democracy, I would move to the UK (Joking, joking. I would obviously move to France). I don’t want to do that though, and thankfully I may not have to, because as bad things are, this also represents the best conditions for reform that we’ve seen in a long time.

Between now and the 2026 elections, dozens of house members on both sides will be drawn out of their seats. Every one of those members will have a powerful incentive to fight back. For their careers, the stakes are quite literally existential. California alone will draw 5 or 6 Republicans out of their seats, and Texas has already drawn out the same number of Democrats. The House of Representatives currently has a governing margin of just 6 republicans, 5 if Representative Elect Adelita Grijalva ever gets sworn in (where are the Epstein Files, Donald?!). That means that a bipartisan coalition of desperate members could take control of the House, deny either side a majority, and demand passage of Gerrymandering reforms. They could refuse to move any legislation from the Senate until they passed such reforms as well. Democrats should push for this, as they are in the minority anyway and have more to lose in these gerrymandering fights. A coalition of Democrats plus blue state Republicans could force a bill through the House. If Democrats are smart, they could potentially get those same reforms through the senate and signed by the president.

Unfortunately, Democrats mostly are not smart when it comes to political strategy, at least in my experience as a Senate staffer (no shade to my many friends on their side, who actually tend to agree with me on this). As such, I will spell out in great detail how they can do this and why it must be done soon.

First, Democrats have some leverage left in the Senate. Though they stupidly gave up some of it by agreeing to pass a clean Continuing Resolution (CR) and end the shutdown, the CR only goes until January 30th, at which point they could force another shutdown. A second shutdown so soon would be embarrassing for Trump and for Republicans, whose poll numbers suffered significantly during the first one. In exchange for avoiding another shutdown, Democrats could demand passage of Bipartisan Gerrymandering Reform. This is a much more realistic ask than their previous demand for hundred of billions in healthcare spending, which was never going to happen (fight me, it’s true). In it’s simplest form it could just be a law stating that redistricting can only happen once a decade, and all the new maps since 2024 are moot. Ideally they would go further and ban partisan gerrymandering entirely, but they should be pragmatic here. If need be, Democrats could (and frankly should) sweeten the deal by offering up something Republicans (especially Trump) may want even more than a budget: nationwide voter ID.

Voter ID might seem out of left field, but I’m advocating Democrats make demands related to elections, so it’s reasonable that they should give something election-related to Republicans. Voter ID has the benefit of being both popular and good policy, it’s really something we should be doing anyway. By offering this long-sought concession to Republicans, Democrats could allow Trump to assume the mantle of Dealmaker, winning his signature and clearing the final hurdle to passage. I’m not saying it’s a sure thing, but it’s a far cry more likely than getting anything on healthcare.

There isn’t much time to get a deal like this, sadly. We have about a 3 month window in which Gerrymandering Reform seems feasible, because Democrats and blue state Republicans have both the incentive and (some) leverage to make a deal. If you or anyone you know works in politics, send them this article and tell them to make it happen. If you’re reading this, I probably don’t need to explain to you how high the stakes are.

Thanks for reading! If you enjoyed this post, consider subscribing. This is a personal blog that I write when I have something to say that doesn’t fit in my Twitter feed. I used to be a Tech and Trade Staffer in the US Senate so I tend to write about semiconductors, technology, industry, and any other random topic I find interesting.

Subscribe now

Share

By now most people nerdy enough to read this blog will know the vague arc of the semiconductor industry: Early transistors were invented and commercialized in Silicon Valley, with Intel being the main company to scale production of what we now know as the modern CPU. This created the Intel of yesteryear, one of the largest and most iconic technology companies in the world which was effectively a monopoly (part of the famous “Wintel” alongside Microsoft Windows). Eventually, though, Intel got lazy and stopped innovating at the rate needed to maintain that monopoly (in particular, it failed to invest in EUV technology despite playing a key role in inventing it) which allowed Taiwan Semiconductor Manufacturing Company (TSMC) to leapfrog it and become the new de-facto monopoly on high end chips. This has made TSMC arguably the most important company in the world, which is a major concern given its… um… volatile location. To understand how to rescue Intel, though, we need first to understand how a modern semiconductor manufacturing company operates. That means we need to look at TSMC.

TSMC operates under a “foundry” model, meaning it doesn’t design the chips it makes. This complements the business model driving most of the chip companies that the average person might know like Nvidia, AMD, Apple, etc: so called “fabless” semiconductor companies that design their chips in house and then outsource the actual physical construction of those chips. TSMC’s foundry model has proven highly successful; it builds nearly all of the worlds high end chips. This has provided it with the scale needed to continue to push the boundaries of semiconductor manufacturing even as the cost of building a next generation chip plant has ballooned to north of $30 Billion. Basically everyone now agrees that the old Intel model of vertical integration just doesn’t provide enough scale to maintain a leading edge on the manufacturing side. Even with margin stacking (vertical integration has its advantages!), Intel just can’t generate enough free cash flow to match TSMC’s investment without the market dominance it used to enjoy (not to mention the aggressive subsidies the Taiwanese government provides TSMC). The solution: if you can’t beat ‘em, join ‘em.

That’s exactly what Intel is doing, or at least trying: it is investing 10’s of Billions of dollars to create new manufacturing capacity in the hopes that it will entice some of TSMC’s best customers to switch, or at least dual source. Unfortunately, that hasn’t happened yet. Intel has yet to sign any significant customers, resulting in big hits to its bottom line. The response from investors has been harsh: the board of directors fired the CEO and installed a new one, who promptly pulled back on investment in fabs abroad and slowed investment in the US. The message is clear: Wall Street will not tolerate taking any more big risks that imperil their short term profits, long term consequences be damned. This has prompted calls for a bailout, or at least some kind of intervention by the US government: it is simply untenable for the global semiconductor supply chain to be so concentrated within a single company located on an island which could, at any time, be conquered by China.

The Trump Administrations response thus far has been to take a stake in Intel to the tune of about 10% of the company. This was done by converting previously allocated grants (aka money the company was already going to get from the CHIPS Act) into equity, effectively giving taxpayers free Intel shares. While not a bad deal for the public, this does not solve either of the immediate problems Intel faces: it needs money to make big investments, and it needs customers to sign up for Intel Foundry Services (IFS) to de-risk those investments. In theory the first problem could be solved by raising money through capital markets, but as the Intel Board has demonstrated investors currently view that as too risky. Fortunately, as I alluded to in the beginning of this post, the solution is simple: in addition to taking a stake in Intel, the government should be IFS’s first big customer.

“But Matt!”, you might exclaim, “The government doesn’t design chips!” Ok ok, you got me there. The government can’t be the *direct* customer of IFS. The best companies to be IFS customers are the same ones that currently buy TSMC foundry services. Those companies have, thus far, been highly skeptical of IFS. The reasons for this are many, and not all public, but boil down to a combination of cost and risk. IFS is unproven, while TSMC has a now decade+ long track record of delivering great service. Being IFS’s first customer means being their for all the initial growing pains as Intel figures out how to offer good customer service, how to respond to non-Intel design needs, and how to balance external vs internal capacity. In a sense, it means helping teach Intel how to be a foundry like TSMC. That’s an added burden nobody wants.

The US Government can solve this problem by making what’s called an Advance Market Commitment. Basically, some US department would promise that anyone who spends the time and energy to help Intel get IFS running will have a guaranteed customer paying an attractive price. This would be expensive, probably on the order of a few billion dollars, but it wouldn’t necessarily be a total loss. In exchange for making such a commitment, the US government would receive a bunch of CPUs and/or GPUs that it could then use for any of the compute intensive services the government already offers, such as building new supercomputers at DOE national labs. If it were really ambitious, it could build out a network of publicly owned datacenters and lease them to the big AI companies like OpenAI, turning the whole thing into a potentially profitable endeavor. After that initial commitment was exhausted, the new foundry capacity built would be pivoted to private sector customers, who could evaluate it’s now proven track record. In the worst case, if Intel and partners can’t deliver, then the government need not buy anything. (this is the beauty of Advance Market Commitments: you only pay for success).

This would shore up Intel’s manufacturing capacity for at least a few years, and give it a fighting chance to reclaim the lead in process technology. In the long run the US will still need to figure out a more systemic industrial policy to counter those of China, Taiwan, Japan, Germany, and … basically everywhere else, lest we be forced to repeat this process of bailing out our industries every few years. That, however, is a much bigger problem. For now, if we want to save Intel, we should keep it simple: they need customers. Let’s give them one.

Thanks for reading! If you enjoyed this post, which is my first in a long time, consider subscribing. This is a personal blog that I write when I have something to say. I used to be a Tech and Trade Staffer in the US Senate, and I lived on TSCM’s campus in Taiwan for a while, so I tend to write about semiconductors, technology, industry, and any other random topic I find interesting.

A Non-Traditional Path to Politics

My background doesn’t naturally lend itself to politics. I began my career at Texas A&M University as a cybersecurity analyst, and after graduating in 2021, I worked as a back-end software engineer for a startup called Evernym. I loved my work at Evernym, but after we were acquired (and then acquired again), I decided I wanted a change (mergers ruin everything). Evernym had spent a considerable amount of time working with governments on digital identity policy, though almost exclusively with European governments. I wanted to see if I could do something to help the U.S.

I’d long had an interest in public policy, but until then, it had mostly manifested as a minor Twitter addiction. It wasn’t until I happened upon a nonprofit called TechCongress that I realized I could do more than just watch from the sidelines. TechCongress places computer scientists, engineers, and other technologists as technology policy advisors to Members of Congress—an opportunity I couldn’t pass up. After writing a few essays and going through several interviews, I signed an offer and made plans to move to D.C. This was my chance to do what I love—solve problems—at a scale unlike any other.

Tackling Big Problems from Day One

And wow, did I get to tackle some big problems. I came in intending to focus on my area of expertise: digital identity. But that plan quickly fell apart when I joined a Senate office and had no fewer than six bills dropped on my desk on my first day. My new boss, Senator Bill Cassidy (R-LA), was not one to waste time. Over the next year, I led policy work in areas both familiar and unfamiliar—and I often found that the unfamiliar was the most enjoyable.

I had several opportunities to work on digital identity policy, most notably by writing the POST ID Act, but I ended up spending most of my time on two subjects: privacy and customs. All six of those initial bills were privacy-related, including the Children’s and Teens’ Online Privacy Protection Act, which I became the lead Republican staffer on. Senator Cassidy had also begun a long-term project to reform U.S. policy on international trade, particularly focusing on balancing our trade relationship with China and preventing abuses like trade-based money laundering. A major part of this effort involved modernizing the U.S. Customs system—especially its technical backbone, the Automated Commercial Environment (ACE). This technical component drew me onto the trade team, and it opened my eyes to a whole new area of policy—one I am now somewhat obsessed with.

A Year of Learning and Influence

In the last year, I have written letters to agency heads, conducted oversight on major federal programs, authored multiple pieces of legislation, and nearly started a fight with another Senate office over a health program in Africa. I even got to meet Bill Gates! Throughout it all, I saw both the best and worst of our political process. While it was often frustrating, it was also oddly inspiring.

I came in with plenty of opinions about what was wrong with the government. While some of those were correct, I discovered that Congress was just as broken as I had imagined—but in dramatically different ways than I expected. Here are some of the lessons I learned:

Lesson 1: Congress Is Not Too Old—It’s Too Young

This might sound absurd at first, but bear with me. Members of Congress are the oldest they’ve ever been—the average House member is now in their mid-to-late sixties, and senators are even older. And yet, you would be forgiven for failing to see the wisdom that should come with age. The reason? Members of Congress are not the ones primarily responsible for writing policy. That role is delegated (to varying degrees) to congressional staff, and this staff may be the youngest it has ever been.

Wouldn’t older members want to surround themselves with experienced advisors rather than a bunch of 20-somethings young enough to be their kids (or grandkids)? They might—but older, experienced advisors are expensive, and Congress no longer has the budget to afford them.

Starting in the 1990s, Congress slashed its own budget as a mostly symbolic gesture to prove they were “living their values.” Since then, congressional spending has continued to decline in real terms, leaving the average congressional office with fewer and lower-paid staffers than it had in the mid-20th century. If it seems like Congress doesn’t know what it’s doing, that’s because, to a large extent, you get what you pay for.

The staff Congress does hire tend to be very bright and, contrary to public perception, usually have genuinely good intentions. The problem is that D.C. is an expensive city, and the average congressional staffer is not paid much. Staff assistants make about $40,000 a year on average, with legislative correspondents earning only slightly more. These are the front-line staff who interact with constituents, and they’re paid less than a night manager at a Chick-fil-A. Legislative aides make about $60,000, and legislative assistants around $80,000—but even that is far below what similarly credentialed professionals make in the private sector, especially in D.C.

As a result, turnover rates are extremely high—even beyond what you might expect, given that every time a member loses re-election or retires, their entire staff loses their jobs. This lack of experience and institutional knowledge severely hinders congressional policymaking and has led to a substantial amount of policy being outsourced to federal agencies and lobbyists. It also means that policymaking is disproportionately influenced by a highly educated, mostly childless, and disproportionately affluent set of young elites—because those who come from wealth are often the only ones who can afford to take these jobs.

Lesson 2: Congress Actually Gets a Lot Done—You Just Never Hear About It

If you’re outside D.C., you’re probably thinking I’m full of it. After all, the prevailing narrative has long been that Congress is a do-nothing body full of dirtbags and lazy bureaucrats. There’s a grain of truth to that—some members genuinely are terrible—but the reality is more complicated.

In today’s polarized environment, getting things done often requires that the public believe nothing is happening. It sounds counterintuitive, but here’s why:

Politicians respond to electoral incentives. Most members of Congress genuinely want to make positive changes. However, our current political climate strongly discourages them from working with the opposing party. Many high-profile bills have failed simply because one side didn’t want to “give the other side a win.”

And yet, members still want to accomplish something. The solution? “Secret Congress.”

Congress actually does work together on many issues—but only on topics that don’t dominate public discourse. This allows progress without alienating partisan voters. As a result, many important but less politically charged policies get passed quietly, while highly visible, controversial issues remain gridlocked.

This is not a good system. But it does mean that Congress isn’t entirely dysfunctional—it’s just constrained by the political realities of our time.

Final Lesson: We Can Fix This

These problems are solvable. They aren’t inherent to our constitutional system; they stem from policy choices, and they can be fixed with new policy choices.

We need to reform partisan primaries, consider ranked-choice voting, and, perhaps most importantly, increase congressional staff budgets to attract experienced professionals.

Above all—don’t despair. Change is possible. Cynicism is easy, but there is hope.

What interesting times we live in.

If you liked this post, consider subscribing! I use this Substack to write about trade, technology, and public policy. If you have thoughts or questions, drop a comment—I’ll respond as soon as I can.