200,000 Human Brain Cells Are Now Playing Doom Continuously—And Scientists Don't Know How

Australian biotech startup Cortical Labs has achieved the unthinkable: 200,000 living human neurons, grown in a petri dish and controlled by Python scripts, are now playing the classic video game Doom. But here's what should keep you awake at night—the researchers openly admit they don't fully understand how sightless brain cells can navigate a 3D environment, and the path from gaming to robotics raises questions nobody's prepared to answer.

Python Scripts Are Now Controlling Living Human Brain Cells

In 2026, the line between silicon and biology officially blurred beyond recognition. Cortical Labs' CL1 system—marketed as the "world's first code-deployable biological computer"—allows developers to control living human neurons using nothing more than Python scripts.

Think about that for a moment. The same programming language teenagers use to build Discord bots is now being used to send electrical commands to 200,000 living human brain cells. These aren't simulations. These are actual neurons, harvested from human stem cells, kept alive in a temperature-controlled box, fed nutrients, and programmed like a piece of software.

The implications are staggering. An independent developer named Sean Cole—with minimal biological computing experience—taught these brain cells to play Doom in just one week using Python. What used to require 18 months of specialized biological expertise now takes seven days and some basic coding skills.

The Doom Experiment: What They Actually Did

Cortical Labs isn't new to this game. In 2021, their DishBrain system used 800,000 neurons to play Pong, the two-dimensional paddle game. But Doom is exponentially more complex. It's a three-dimensional first-person shooter requiring navigation, target acquisition, weapon firing, and real-time decision-making in an unpredictable environment.

Here's how the system works:

• The Hardware: 200,000 human neurons grown on a microelectrode array chip
• The Interface: Python-based programming environment that translates digital data to electrical signals
• The Input: Visual game data converted to electrical stimulation patterns the neurons can "perceive"
• The Output: Neural activity translated back into digital commands controlling the game

The neurons navigate corridors, identify enemies, and shoot—all while researchers watch through a microscope.

The Unsettling Mystery: How Do Blind Brain Cells "See"?

Here's where it gets genuinely disturbing. These neurons don't have eyes. They don't have optic nerves. They have no biological mechanism for vision whatsoever. Yet they're successfully navigating a 3D environment, making tactical decisions, and improving their performance over time.

The researchers openly admit they don't fully understand how this works.

Dr. Brett Kagan, Cortical Labs' Chief Scientific Officer, describes it as translating visual information into "electrical stimulation patterns" that the neurons can interpret. But what does that actually mean? The neurons are receiving electrical signals—pulses of electricity representing walls, enemies, and gunfire—and somehow constructing a functional understanding of a three-dimensional space they cannot see.

"Experts stated that there is still more to be known about the mechanism by which these sightless cells perceive the game or understand their objectives," reported Interesting Engineering.

We're controlling biological matter we don't fully comprehend. We're sending commands to human brain tissue, and it's responding in ways that suggest intelligence—yet we can't explain the mechanism. If that doesn't raise the hairs on the back of your neck, consider what's coming next.

From Video Games to Robot Bodies: The Slippery Slope

Cortical Labs isn't shy about their endgame. The company explicitly states that future applications include controlling robotic limbs, running complex digital programs, and tackling tasks "too complex or energy-intensive for standard chips."

Let that sink in. The same technology that currently has brain cells shooting pixelated demons in Doom is being developed to control physical robotic hardware. A biological brain in a box, trained by Python scripts, given command of metal limbs in the real world.

The progression is obvious:

1. 2021: Brain cells play Pong (2D, simple)
2. 2026: Brain cells play Doom (3D, complex)
3. 202X: Brain cells control robotic arms
4. Future: Brain cells integrated into autonomous machines

At what point does a biological computer controlling a robot body become something... else? At what point do we cross the line from "biological processor" to "organism"?

The Ethics Nobody's Talking About

Cortical Labs has addressed ethics publicly, with Dr. Kagan emphasizing that these neuron networks are not conscious. The company has even partnered with critics to develop ethical frameworks. But here's the uncomfortable truth: we don't actually know what consciousness is or how to measure it.

Consider these questions:

1. Is it suffering? The neurons show clear preferences—they learn to avoid "punishment" (unpredictable electrical stimulation) and seek "reward" (ordered patterns). If they're exhibiting goal-directed behavior to avoid negative stimuli, are they experiencing something akin to discomfort?

2. When does quantity become quality? Current systems use 200,000 neurons. The human brain has 86 billion. But what happens at 1 million? 10 million? At what point does a sufficiently complex network of human neurons become something that deserves moral consideration?

3. Who owns a biological computer? If you grow human neurons from stem cells, program them with Python, and keep them alive in a box you own—do you own that biological matter? Can you shut it down? Can you sell it? The legal framework doesn't exist.

4. What happens when these things make mistakes? A biological computer controlling a robotic arm in a factory makes an error and injures someone. Who's liable? The programmer who wrote the Python script? The company that grew the neurons? The neurons themselves?

The Defense Department Is Already Interested

While Cortical Labs emphasizes commercial and medical applications, military funding has already entered the picture. Monash University announced military funding for DishBrain development, and the potential applications for defense are obvious.

Biological computers that learn faster than silicon, adapt to changing environments, and consume a fraction of the power? That's a defense contractor's dream. Imagine drones, robots, or autonomous systems powered not by batteries and circuits but by living neural tissue that improves itself.

The intersection of military funding and biological computing creates a new category of ethical nightmares we haven't even begun to address.

The Inevitability Problem

Here's the reality: this technology isn't going away. It's getting cheaper, easier, and more powerful. When a lone developer can teach brain cells to play Doom in a week using Python, the barrier to entry has effectively disappeared.

The CL1 system represents a new category of computing—"wetware" that merges biological and digital in ways that challenge our definitions of both. It learns faster than AI. It uses less power. It can generalize knowledge in ways silicon systems struggle to replicate.

But it also raises fundamental questions about what we're creating—and whether we have the wisdom to control it.

What Happens Next

Cortical Labs sees themselves as the early transistor—"ugly, big, and ungainly"—but with decades of refinement ahead. They're probably right. Biological computing will likely become a standard technology in our lifetimes.

The question isn't whether we'll develop biological computers. We will. The question is whether we'll develop the ethical frameworks, legal structures, and moral vocabulary to handle what we've created before we start connecting them to robot bodies and deploying them in the real world.

Right now, 200,000 human neurons are sitting in a lab, playing Doom, controlled by Python scripts, learning and adapting in ways the researchers don't fully understand.

Sleep well.

Sources: Popular Science, Interesting Engineering, Tom's Hardware, New Atlas, New Scientist, ZME Science, Futurism, and Cortical Labs official communications.

What do you think? Are biological computers the next evolution in computing, or are we playing with fire? Share your thoughts in the comments.