The Original Computer Was Not Made of Silicon

3.8 Billion Years of Computing History

The first general purpose computer most of us learn about is the ENIAC, built in 1945, weighing 30 tons, consuming 150 kilowatts of power, capable of performing 5000 arithmetic operations per second.

But ENIAC was not the first computer. The first was invented roughly 3.8 billion years earlier, in an anoxic pool of prebiotic water. It weighed less than a nanogram. It ran on sugar. And it was so well designed that every single living thing on Earth is still running variations of its architecture today.

We call it a cell.

It Is Not an Analogy. It Is a Literal Description

Most people treat the "cell as a computer" as a cute, imperfect analogy. It is not an analogy. It is a literal description, if you use the actual definition of what a computer is. A computer is not a slab of silicon and copper. It is any programmable system that takes input, stores information, processes that information according to encoded rules, and produces output to achieve a goal. By that definition, a cell is not like a computer. It is the ur-example of the concept.

The Limitations of the Von Neumann Silicon Design

Our modern silicon computers almost all follow the Von Neumann design we standardized in the 1940s: discrete, separate components for memory, for processing, for power supply, for input and output. Data is constantly shuttled back and forth between the storage drive and the processor, a wasteful design flaw we call the Von Neumann bottleneck. All operations are synchronized to a central clock. If a single transistor fails, or a single bit flips uncorrected, an entire program can crash.

The Cell: The Computing Paradigm We Are Still Trying to Invent

The cell's architecture is so advanced we are only just beginning to replicate it in silicon. Most of the "revolutionary" new computing paradigms we research today -- in-memory computing, asynchronous processing, neuromorphic design, self repairing hardware -- are not new ideas at all. They are just features the cell has had since the beginning of life.

It has no central processing unit. It has no clock. It has no separate memory. Every part of the cell is simultaneously storage, processor, sensor, and actuator.

DNA is More Than a Hard Drive

We like to describe DNA as the cell's hard drive. This is a massive understatement. The sequence of the genome is not just stored data: it is executable code. The epigenetic marks written on top of the DNA are not just saved files: they are the configuration settings that determine which code runs, when, and for how long. The same strand of DNA can store the blueprints for every tissue in your body, and run the logic to decide if a cell should become a neuron, a liver cell, or a skin cell.

The Cytoplasm: A Trillion-Core Asynchronous Processor

The entire cytoplasm of the cell is a massively parallel asynchronous processor. For decades, biologists saw the random Brownian motion of proteins floating inside the cell as unwanted noise, a flaw biology had to work around. We now understand it is the system's bus. No wires, no copper traces, no energy spent routing signals. Proteins drift through the cell, and perform computation the moment they bump into a molecule that matches their shape. A single cell can run trillions of these operations per second, at room temperature, using less power than you would need to light a single pixel on your phone screen.

Does This Comparison Reduce the Magic of Life?

Critics of this framework will immediately object: Cells are alive, computers are not. The analogy breaks down here.

Some people recoil from this comparison, because they believe describing life as computation reduces it to something cold, mechanical, and unremarkable. The opposite is true. When you recognize a cell as a computer, you do not make life less magical. You realize just how unimaginably impressive the technology of life is. We have spent a century of human ingenuity to build silicon computers that can do a tiny fraction of what a single yeast cell can do, effortlessly, every day of its existence.

What If Life Is Computation?

This objection also forces us to ask a far more interesting question: what if there is no hard line between computation and life? What if being alive is simply the emergent property of a computer system complex enough to sustain itself, repair itself, and rewrite its own source code while it runs?

Unlike every silicon computer we have ever built, the cell does not need an external programmer to write its code. It writes its own code, through the process of evolution. It runs updates, patches bugs, and tests new features, continuously, for billions of years, without ever shutting down. The operating system running in your cells right now has been in active, uninterrupted development, passed down unbroken from cell to cell, since life first appeared on this planet. It is quite possibly the most well tested piece of software that will ever exist.

This Mental Model Rewrites Entire Fields of Science and Technology

Once you stop seeing cells as bags of messy chemicals and start seeing them as computers, almost every field of science and technology changes shape.

Medicine Becomes Software Engineering

Medicine stops being the practice of pouring chemicals into the body to fix malfunctions, and becomes the field of software engineering. We already demonstrated this with the mRNA COVID vaccines: we did not inject you with a drug that kills the virus. We sent your cells a small, simple program to run. Your cells executed the code, produced the spike protein, and your immune system learned to defend itself. We did not heal you. We gave your body's hardware new code to run.

Cancer is a Firmware Bug

Cancer, from this perspective, is not first and foremost a disease. It is a firmware bug. The code that regulates how and when a cell divides becomes corrupted, and the computer runs an infinite loop of self replication. The most durable cures for cancer will not be magic pills. They will be software patches that correct that broken code, or instruct the corrupted cells to shut themselves down.

We Have Been Optimizing the Wrong Metrics for Decades

Even our approach to building better computers is transformed. We are already hitting the hard physical limits of silicon transistor miniaturization. There is only so small we can carve silicon before quantum effects break our ability to reliably store bits. For decades we have chased faster clock speeds and more transistors, but the cell shows us we have been optimizing the wrong thing. The human body runs all of its functions, including a brain capable of writing poetry and inventing rocket ships, on roughly 100 watts of power. An array of GPUs with a similar level of raw computational capacity would consume megawatts of electricity.

The Future of Computing Already Exists

The future of computing is not more silicon. It is molecular computing, built out of the same materials cells are made of. We will one day build computers that grow, that heal their own damage, that can run for years on a single drop of glucose. We already have working prototypes of DNA storage systems that can hold more data than any hard drive, and logic gates built out of proteins.

For decades, popular science fiction has taught us to imagine that the most advanced computers we will ever see will be built by us, housed in server farms, or in the bodies of androids. We spend so much time wondering when we will build an artificial intelligence smarter than a human.

We rarely stop to notice that we are already surrounded by technology more advanced than anything we will ever be able to build from scratch. Every tree, every bee, every cell in your skin, is a computer so well engineered we can barely replicate the most basic of its functions.

Closing: The Age of Biological Programming

We are now at the very beginning of learning how to write code for these ancient machines. We can edit genomes, we can deliver mRNA programs to cells, we can build entirely new synthetic cells from scratch that run custom code no natural cell ever ran.

So the question is not if cells are computers.

The question is: what do you want to program them to do?