Why Advanced ETs Use Classical Computing but Not ASML-Style Lithography
By H Y Nahm | 21 Jul, 2026

Advanced computing will include quantum acceleration for some purposes but will be based on classical processors fabricated using molecular self-assembly rather than etched silicon.

Speculating about alien technology may seem a fool's errand. 

Of course, we can't know what a civilization a million years older than ours has built, and most guesses say more about the guesser than about the aliens.  But there is one domain where physics and mathematics let us reason with unusual confidence: computation. The laws that govern what can be computed, and at what cost, aren't human conventions but immutable features of the universe.  Any civilization that computes must live within them.

That lets us make a surprisingly strong claim: any advanced civilization almost certainly performs most of its computation classically. And a second, weaker but still well-grounded claim follows: they almost certainly do not fabricate those classical processors the way we do, with room-sized machines projecting extreme ultraviolet light onto silicon wafers.

The Universality of Classical Logic

Classical computation—the manipulation of definite bits through deterministic logic—isn't a human invention so much as a human discovery. Alan Turing did not design the limits of computation; he uncovered them. Any physical system that can reliably distinguish two states and combine them through logical operations can, in principle, compute anything computable. This universality is substrate-independent. It works in silicon, in vacuum tubes, in DNA, in the protein-signaling networks of a single cell, and presumably in whatever exotic matter an ancient civilization prefers.

More importantly, the workloads of any technological civilization are overwhelmingly classical in character. Arithmetic, bookkeeping, control systems, communication protocols, databases, simulations of macroscopic physics, the coordination of machines and infrastructure—these are tasks defined by definite inputs, sequential logic, branching decisions, and stored state. There is no reason to think alien civilizations escape these needs. A Dyson swarm requires station-keeping calculations. An interstellar communication network requires error-correcting codes and routing logic. Whatever passes for their economy requires ledgers. These are classical problems, and the most efficient way to solve classical problems is with classical machines.

Why Quantum Computing Stays in Its Lane

It is tempting to imagine that a sufficiently advanced civilization would have transcended classical computing entirely, running everything on vast quantum computers. But this misunderstands what quantum computation is for. Quantum computers do not simply run all algorithms faster. They exploit superposition, entanglement, and interference to achieve speedups on problems with very particular mathematical structure: factoring integers, searching unstructured spaces, simulating quantum systems, and certain linear algebra and optimization tasks.

For everything else—which is to say, for the overwhelming majority of computation—quantum processors offer no advantage while imposing enormous costs. Quantum information cannot be copied, thanks to the no-cloning theorem, which eliminates the cheap fanout and memory hierarchies on which classical architectures depend. Measuring a qubit destroys its superposition, making the routine read-and-branch operations of ordinary software awkward at best. Quantum error correction demands large overheads of physical qubits per logical qubit, and quantum gate operations are intrinsically slower than classical switching in any implementation we can foresee. These are not engineering immaturities awaiting better fabrication. They are consequences of quantum mechanics itself, and an alien civilization operating under the same physics faces the same constraints.

There is also a structural dependency that runs in only one direction. Every quantum computer requires a substantial classical computer to operate it: generating control pulses, scheduling operations, and decoding error-correction syndromes in real time. Classical computation needs nothing quantum to function. The natural architecture, for us and presumably for anyone, is therefore a classical foundation with quantum acceleration attached where it pays: a quantum coprocessor invoked for chemistry, materials simulation, cryptanalysis, and certain optimization problems, in the same way we bolt GPUs onto CPUs today. Advanced alien computing is very likely hybrid in exactly this sense—mostly classical, quantum where quantum wins.

The Three Layers of a Computer

Here is where the argument must be handled carefully, because it is easy to slide from "classical computing is permanent" to "our way of building classical computers is permanent." Those are radically different claims. It helps to separate three layers that our current technology happens to fuse together.

The first layer is the computational paradigm: classical logic operating on bits. This layer, as argued above, is protected by mathematics and by the nature of the workloads themselves. It is plausibly universal across all technological civilizations.

The second layer is the physical substrate: for us, silicon CMOS transistors. Nothing about classical logic requires silicon. Bits can live in superconducting circuits, photonic switches, magnetic domains, molecular conformations, or mechanisms we have not imagined. Silicon won on Earth because of a contingent chain of events—the properties of silicon dioxide, the momentum of an industry, trillions of dollars of accumulated process knowledge—not because physics anointed it.

The third layer is the manufacturing method: photolithography, culminating in ASML's extreme ultraviolet machines. This layer is the most contingent of all. It is a way of making one particular substrate, which is itself one way of realizing one paradigm. Only the paradigm is protected by deep principles. The substrate and the manufacturing method are protected by nothing more than engineering economics.

EUV Lithography as a Local Peak

Consider what an EUV lithography machine actually is. A 180-ton apparatus fires a high-power laser at droplets of molten tin, fifty thousand times per second, vaporizing each droplet into a plasma that emits light at a wavelength of 13.5 nanometers. Because no material transmits that light, it must be bounced through a cathedral of the most precise mirrors ever manufactured, finally projecting a circuit pattern onto a photosensitive wafer. It is one of the most astonishing machines humans have ever built. It is also, viewed from a distance, magnificently baroque—a Rube Goldberg triumph that exists because of the specific path our technology took.

That path matters. We committed early to planar silicon and to top-down patterning: start with a uniform substrate and impose structure on it from outside, using projected light as the stencil. Every subsequent decade of progress consisted of heroically extending that single idea to smaller wavelengths and tighter tolerances. EUV is not the inevitable endpoint of nanofabrication. It is the summit of one particular ridge in the technology landscape, the ridge we happened to start climbing in the 1960s. A civilization that began from different starting conditions—different materials at hand, different founding insights, a different sequence of discoveries—would climb a different ridge and reach a different summit.

Biology's Existence Proof

There is, moreover, a standing demonstration that a fundamentally different fabrication paradigm works: life. Every cell in your body is a molecular factory that builds nanoscale machines—ribosomes, motor proteins, signaling networks—not by carving structure into a substrate from the top down, but by assembling it from the bottom up. Molecules encode their own assembly instructions and organize themselves into functional structures, in parallel, at ambient temperature, with error-correction mechanisms built in, at a scale of quadrillions of units per organism. Biology fabricates at the nanometer scale routinely, cheaply, and without a single mirror.

Human efforts toward bottom-up fabrication—DNA origami, molecular self-assembly, proposals for atomically precise manufacturing—are today primitive. We cannot yet coax self-assembling systems into producing arbitrary, complex, defect-free patterns at the billions-of-transistors scale. But the barrier is engineering maturity, not physical law, and that is precisely the kind of barrier that falls to civilizations with thousand-year head starts. Bottom-up assembly is in principle more elegant and more scalable than top-down etching: rather than imposing a pattern with an external stencil, you design components that find their own places. A mature version of that capability would make photolithography look the way vacuum tubes look to us—an ingenious solution to a problem that no longer needs solving that way.

The Asymmetry That Answers the Question

So we arrive at an asymmetry worth stating plainly. Classical computation is protected by the deepest kind of moat: the mathematical structure of computation itself and the classical character of most real-world workloads. Quantum computing cannot breach it, because quantum advantage is narrow by nature; it can only complement it. Any advanced civilization, anywhere, is very likely running mostly classical logic with quantum accelerators at the margins.

Photolithography enjoys no such protection. Its moat is ASML's decades of accumulated capability and an irreplaceable supply chain—formidable against terrestrial competitors on human timescales, meaningless against physics and deep time. The universe does not care whether classical bits are patterned by ultraviolet light or grown by molecular machinery, and biology strongly hints that the growing approach is what maturity looks like.

The aliens, in short, are almost certainly computing the way we do. They are almost certainly not manufacturing the way we do. Classical logic is universal; tin-vaporizing light cathedrals are probably just us.