Semtex in Every Pot

“Semtex in every pot, a 3D printer in every garage.”

It lands as a joke because it sounds like a mutation of an older promise—a chicken in every pot—but something in it feels less like satire and more like recognition. Not a policy. Not even a warning. A condition.

For most of the industrial era, dangerous capability was difficult to produce and easy to locate. It required factories, supply chains, capital, and coordination. The state did not merely wield force; it organized the conditions under which force could exist. Production was legible. Control followed from that legibility.

That model has not disappeared. It is still how tanks are built, how missiles are designed, how companies like Rheinmetall understand the world. Innovation, in that frame, is large-scale, proprietary, and slow enough to be governed. A thing is invented, then produced, then deployed.

It is also easy to forget what that system actually contains: decades of materials science compressed into a battery cell; fabrication processes measured in nanometers; supply chains that span continents and concentrate risk in a handful of foundries. The visible factory is only the surface of a much deeper structure.

What has changed is not the existence of that system, but the emergence of another one alongside it.

Small teams assemble strike-capable systems from components they did not invent, using processes that did not originate with them, guided by feedback loops measured in days rather than years. The work is not invention in the classical sense. It is adaptation, iteration, recombination. It happens in workshops, garages, improvised labs. It looks informal. It often is.

This is where the argument tends to split.

On one side: this is dismissed as non-innovation—mere assembly, derivative work, a shadow of “real” industrial production.

On the other: it is elevated into a revolution—proof that production has escaped the factory, that capability has been democratized, that the old system has been disrupted.

Both readings are incomplete. What is happening is simpler, and harder to contain. Production has become layered.

At the component level, nothing is democratized. Motors, chips, batteries, sensors—these remain products of highly centralized, capital-intensive, global supply chains. The system depends on them absolutely.

At the assembly level, however, capability has spread. The barrier to turning components into working systems has dropped. Knowledge circulates. Designs propagate. Iteration accelerates.

Distributed at the top. Centralized at the base. This is not a replacement of the old model. It is an interleaving. And because it is layered, it behaves differently.

The systems built this way are inconsistent, fragile, often inefficient. They fail more frequently. They lack standardization. They depend on coordination that is easy to romanticize and difficult to sustain. Left alone, they do not scale cleanly.

They do not need to.

They operate in the gaps—where cost asymmetries matter, where speed outruns procurement, where adaptation matters more than durability. They are not general-purpose replacements for industrial systems. They are tactical distortions that force everything around them to adjust.

Improvisation, under pressure. But improvisation has a habit of persisting.

The IED did not remain an improvised weapon. It altered vehicle design, doctrine, procurement, training—less by its permanence than by the permanence of the response it forced. The system that encountered it did not return to its prior assumptions.

This is how adaptations become structure. A layer is added. It does not disappear. It is built over.

The same pattern is visible here. The specific conditions that produce distributed drone systems—war, urgency, constraint—may not generalize cleanly. The systems themselves may be temporary in their current form.

The lesson is not.

Once demonstrated, capability does not return to a prior state of impossibility. The threshold has moved.

This is the part that resists containment.

States can still build larger, more capable systems. They can invest in countermeasures. They can dominate the electromagnetic spectrum, disrupt signals, degrade performance. Electronic warfare, autonomy, and software will continue to concentrate advantage in ways that look familiar—expensive, centralized, difficult to replicate.

New ceilings will be built. But the floor does not rise with them.

It remains possible, now, for small teams with access to global supply chains and shared knowledge to produce systems that were once the exclusive domain of organized militaries. Not at scale. Not with the same reliability. But enough.

Enough for what is rarely agreed upon in advance. Enough to force a response is sufficient.

Enough is sufficient to change assumptions.

Enough is sufficient to persist.

This is where the older model begins to strain—not because it is obsolete, but because it was built on a different assumption about where capability resides.

It used to be that the state provided security by monopolizing force. Now it manages security in an environment where force is ambient.

Control, in that environment, is no longer a matter of restricting access to production. It becomes a matter of shaping the ecosystem in which production occurs—supply chains, standards, interfaces, signals, narratives. What is allowed to scale. What is suppressed. What is rendered legible.

Not everything that can be done will be normalized. That, too, will be contested. But the underlying condition does not reverse.

“Semtex in every pot, a 3D printer in every garage” is not a forecast. It is an exaggeration that points at something real.

Not a promise of provision.

A condition of possibility.