CeleryKills

I was doing what I always do, scanning my feed across three of my favorite topics that unfortunately don’t usually line up cleanly. Abiogenesis. Nanotech. Materials science. Normally they behave like separate conversations, overlapping at the edges but rarely colliding. At best they lend a different way to look at each topic. That morning they stacked. Same headline, same idea, just refracted through three different lenses. That is usually when something interesting is happening, when systems stop respecting the boundaries we drew for them.

The SciTechDaily piece reads almost casually, as if materials that crawl, walk, and dig is just the next line item in engineering. But buried under the tone is a structural shift. These are not passive materials anymore. They do not wait for forces. They generate them. Internally. Continuously. Active matter systems draw on internal energy to respond dynamically rather than just deform under external inputs (SciTechDaily, Living Materials That Crawl, 2026). That alone would be enough to nudge the field. It does more than that. It blurs categories we thought were stable.

The PNAS work “Nonreciprocal buckling makes active filaments polyfunctional“ makes this concrete in a way that feels less like theory and more like demonstration. Chains of rods with embedded motors, interacting in asymmetric ways. Push from one side, the response is not mirrored. That breaks reciprocity, which is one of those quiet assumptions you do not notice until it refuses to hold. The structures do not simply buckle once like paper. They oscillate, crawl, dig, sustain motion without a controller (Binysh et al., Nonreciprocal Buckling, PNAS 2026). The material is not being used to do something. The material is doing something.

And then the Physical Review X (PRX) paper “More is Less in Unpercolated Active Solids“ complicates the story further. Increase the microscopic activity and the macroscopic behavior can vanish. More energy, less visible motion, because the activity localizes instead of propagating through the system (Binysh et al., More is Less in Active Solids, PRX 2026). That is not intuitive. It feels like a mistake until it repeats. What you get instead is something closer to biological systems, where local processes do not always scale cleanly. Connectivity matters as much as activity. Structure constrains expression.

I have seen this pattern from another direction. Abiogenesis work is full of systems that hover at the edge of becoming something else. Chemical networks that persist because they consume energy rather than dissipate it. Self‑organizing structures that resist equilibrium. Life, stripped of its narrative weight, is matter maintaining itself far from equilibrium (Kumar, Active Soft Matter, 2026). These engineered materials are doing something uncomfortably similar. Not life. Not yet. But not passive either.

That is where the connection tightens.

If you scale these materials down, or push abiogenesis upward into engineered systems, they begin aiming at each other. Materials that move, adapt, and respond using internal energy. Chemical systems that cycle, replicate, organize. Nanotechnology has been trying for decades to build systems with those properties, usually through top‑down design. This moves in the opposite direction. Local rules. Distributed energy. Behavior that emerges because it cannot collapse into simplicity (Active Matter Overview, 2025).

I am not sure people are fully registering what that means.

We have spent a long time treating materials as objects and life as process. That distinction is starting to look like convenience rather than law. Once materials can act, once they can generate motion and respond to environments internally, the question shifts. Not what something is made of, but how it behaves across time.

We have spent a long time treating materials as objects and life as process. That distinction is starting to look like convenience rather than law. Once materials can act, once they can generate motion and respond to environments internally, the question shifts. Not what something is made of, but how it behaves across time.

And once you follow that shift honestly, something else starts to surface, something less comfortable. The possibility that chemistry is not wandering toward life by accident, but being cornered into it. Given energy, gradients, and enough time, matter does not sit still. It organizes, reorganizes, builds feedback loops because there are statistically more ways for complexity to stabilize than there are for it to remain inert under constant flux. Not purpose. Pressure.

I remember the first time this crossed from abstract to irritatingly plausible for me. It wasn’t in a lab. It was staring at a simple reaction network model that refused to decay when every intuition said it should. It kept cycling. Not efficiently, not elegantly, but persistently. And I kept thinking, if you ran this long enough, scaled it up, gave it structure, at what point does “this is just chemistry” become “this is behavior”?

Because that line, like the one between material and life, doesn’t actually exist as a boundary. It exists as a threshold we decide to name.

Probability does the rest. Not in the small. In the overwhelming aggregate. Individual reactions fail. Systems collapse. Most configurations go nowhere. But given enough volume, enough time, enough retries, the improbable becomes statistically inevitable. Structures that store energy, replicate patterns, maintain themselves across perturbations. Not because they are designed to, but because those are the configurations that survive the filtering effect of reality itself.

It starts to look less like emergence and more like selection before there was anything we would call life.

And that reframes abiogenesis in a way I find hard to shake. Not as a rare miracle, but as a convergence point. A basin in the landscape of chemistry where systems are pulled simply because there are more stable pathways leading there than away from it. Once you allow active materials, energy-consuming matter that maintains itself out of equilibrium, the gap narrows even further. The “dead” material is already doing most of the work.

So what are we actually waiting for? A name? A threshold event? A moment where we agree to stop calling it chemistry and start calling it biology?

Or is that transition something we imposed after the fact because we prefer clean categories over gradients?

If the governing laws plus probability continue to push matter into states that move, adapt, and persist, then the inevitability question stops being philosophical and starts being statistical.

Not “could life arise.”

But “how often could it fail not to.”

Science fiction noticed this early. You can trace it if you care to. Early stories gave us robots with circuits and instructions. Later ones gave us swarms. Then programmable matter. Then environments that move, adapt, consume energy. Not machines you control directly, but systems you configure and release. This research sits squarely in that lineage, except the engineering is catching up.

You can actually map the shift almost cleanly across a few canonical science fiction works if you’re willing to read them less as stories and more as physics anxieties in disguise.

Take Lem’s Solaris. The planet is not a machine and not quite an organism, more unsettling than both. It responds, adapts, generates structures out of itself in direct interaction with observers. The scientists don’t study it so much as provoke it, and every experiment feeds back into the system in ways they cannot predict. That is active matter before we had a lab version. The environment itself acts, consumes energy, and reorganizes under observation (Lem, Solaris, 1961). You don’t control it. You participate in it, often unwillingly.

Herbert’s Dune moves the scale but keeps the logic. The sandworms, the spice cycle, the entire desert ecology operate as a distributed system with no central authority. Energy flows create behavior. Individual parts matter less than the network. The system adapts. It resists simplification. Try to intervene directly and the response propagates unpredictably. That is much closer to what the PNAS results suggest. Local rules producing global outcomes that cannot be scripted (Herbert, Dune, 1965).

Then you get Stephenson’s The Diamond Age, where matter itself becomes programmable at fine scales. Not just shaped, but instructed. Nanotech systems assembling, adapting, moving at the boundary between material and machine. It reads differently now. Less speculative, more like an early description of what happens when you give matter internal energy and rule sets. Not quite active matter in the strict physics sense, but pointing directly at it (Stephenson, The Diamond Age, 1995).

And the Borg from Star Trek sit somewhere in the middle, whether people realize it or not. Not the individual drones, those are just interface. The interesting part is the collective behavior. Distributed control. Emergent coordination. Adaptation to external pressure in real time. The system learns by being perturbed. That is the same feedback loop active matter experiments are now showing at the material level. Except here the “material” also decides you are part of it now (Star Trek: The Next Generation, 1989).

All of these share one quiet assumption that lines up uncomfortably well with the PRX results. More activity does not always mean more visible control. Sometimes it fragments. Sometimes it localizes. Sometimes the system becomes harder, not easier, to influence as you push energy into it. The stories showed that as chaos or resistance. The physics papers show it as emergent constraint.

And that leaves a strange symmetry.

Science fiction assumed materials would start behaving like systems.
The lab is now building systems that behave like life.

The difference is getting harder to point to without gesturing vaguely and changing the subject.

The SciTech framing leans toward application. Soft robotics. Adaptive systems. Exploration. That is fine. Necessary, even. But the physics underneath is less reassuring. Systems far from equilibrium do not settle into predictability. You can tune parameters. You cannot script outcomes cleanly. That has always been the bargain with life. It is now becoming the bargain with materials.

So the questions start stacking up in ways that don’t resolve cleanly.

If materials can generate behavior, how much control do we actually have?

If active systems can localize and disappear at scale, how much are we missing in our measurements?

At what point does calling something nonliving stop carrying meaning?

I keep coming back to that morning feed. Three fields aligning long enough to show the same thing from different angles. Not convergence exactly. More like recognition.

We are not building life.

We are building systems that make the distinction harder to defend.

References

• Binysh, Jack et al. Nonreciprocal Buckling Makes Active Filaments Polyfunctional. PNAS, 2026.
• Binysh, Jack et al. More Is Less in Active Solids. Physical Review X, 2026.
• Emergent Mind. Active Matter System Overview. 2025.
• Herbert, Frank. Dune. 1965.
• Kumar, Nitin. An Active Soft Condensed Matter Approach to the Physics of Living Systems. 2026.
• Lem, Stanisław. Solaris. 1961.
• SciTechDaily. Scientists Create Living Materials That Crawl, Walk, and Dig on Their Own. 2026.
• Star Trek: The Next Generation. 1989.
• Stephenson, Neal. The Diamond Age. 1995.