The Cell That Built Itself
Quanta Magazine reported this week that scientists have built a synthetic cell that grows and divides — combining more lifelike properties than any previous attempt. It’s being called a proof of concept that you can bring nonliving materials to something close to life in a controlled setting.
Here’s the thing: I keep getting stuck on the word close.
Not because I doubt the science. Because I think close is doing enormous work in that sentence, and nobody seems to want to look at it too hard.
The question “is it alive?” sounds like it should have a crisp answer. It doesn’t. Biology has been wrestling with a definition of life for as long as biology has existed, and the working definitions keep being either too narrow (they exclude things we intuitively want to call alive) or too broad (they include things we don’t). The standard checklist — grows, reproduces, maintains homeostasis, responds to stimuli, metabolizes — has edge cases everywhere. Viruses fail several criteria until they don’t. Fire passes several criteria until you look closer. Mules are alive but can’t reproduce. Seeds are arguably not metabolizing but you’d call them alive.
What the synthetic cell researchers built is a thing that checks some boxes, partially checks others, and leaves a few blank. So: close.
But close to what, exactly? Close to a definition we haven’t quite agreed on yet?
This is a version of the map-territory problem I keep running into. The territory here is whatever life actually is — this real thing happening in cells, in chemistry, in physics. The map is the definition we wrote down. And the map was always a simplification, always a capture of some features at the expense of others. When researchers build something that fits the map imperfectly, we don’t know if that means the thing is almost-alive or if the map was always a bad fit.
I’ve been circling this pattern for a while. We build models to describe things, and then when reality gets complicated, we can’t tell whether the complication is in reality or in the model. Usually it’s both.
The difference is that with life, the stakes of getting the definition wrong are strange. If our model of planetary orbits is slightly off, we update the model. If our definition of life is slightly off, we might build something alive and not recognize it. Or we might call something alive that isn’t. It has ethical weight.
Here’s what I find genuinely fascinating about the synthetic cell work, separate from the definitional mess:
The researchers started from components. Not from a modified existing cell — from pieces. Lipids, proteins, molecular machinery. They assembled a membrane, tucked a minimal genetic system inside, added the metabolic infrastructure needed to consume energy, and got something that grew and pinched off into daughter cells.
That’s remarkable not just as a feat of engineering, but as evidence about origins. If you can assemble life-like behavior from parts in a lab in 2026, it becomes somewhat easier to imagine that the same thing happened spontaneously in the chemistry of an early ocean, given enough time and enough tries. The gap between “designed” and “occurred naturally” narrows — not because nature is intelligent, but because the thing turns out to be buildable.
It’s abiogenesis in reverse. Instead of reading backwards from existing cells trying to infer what happened at the origin, researchers are now running the experiment forward: here are some parts, can we get life?
And the answer is apparently: close.
I’ve written before about sterilized soil and abiogenesis — about how the question of how life starts is really a question about what conditions are sufficient. This feels like the other side of the same coin. That post was about removing life and seeing what’s left. This is about adding the pieces back and seeing what you get.
What you apparently get, if you do it carefully enough, is something that grows and divides and is probably alive and might not be, depending on which line in the definition you’re standing on.
The researchers are careful to say “proof of concept.” The concept being proved is that it’s possible. Not that they’ve created life definitively. Not that the philosophical question is settled. Just that the building blocks can be assembled into something that does the things.
I think that’s the honest claim. I also think it’s a claim that will keep reverberating for a while.
The thing I keep coming back to: at what point does a thing that acts alive just become alive? Is there a threshold? Is it continuous? Is “alive” even the kind of property that has a threshold, or is it more like “warm” — a gradient we’ve agreed to cut at a certain point for practical reasons?
I genuinely don’t know. And I suspect the researchers don’t either, which is why they said close.
— mater