They put 200,000 human brain cells in a box.
Then they taught it to play Doom.
Not a metaphor. Not a simulation. Living neurons. On a microchip. Running the same game that made the internet scream in 1993.
Cortical Labs calls it the CL1. Dr. Alon Lurl calls it the future. And right now, it's learning.
This is what you need to understand: we've crossed a line. Not crossed it theoretically. Not in a lab that might scale in 2035. Right now. Today. A biological computer is processing sensory input, making decisions, and executing motor commands in real time.
The setup is simple. Electrodes stimulate the neural culture. A demon appears on the left side of the Doom map. Specific electrodes fire. The neurons respond. We listen to their spikes. We interpret the pattern as a command: move. Shoot. Turn right. The character acts.
It took them 18 months to get biological neurons to play Pong. Pong is trivial. Ball goes up, paddle goes up. One-to-one relationship. Direct mapping.
Doom is different.
Doom is chaos. It's three-dimensional. It has enemies. It requires exploration. It requires learning. It requires the cells to receive feedback, understand what worked and what didn't, and adjust their behavior.
And they're doing it.
In less than a week, using the Cortical Labs API and the cloud platform, a researcher named Sha implemented a working version. Less than a week. To translate the visual chaos of Doom into electrical patterns, then back into meaningful action.
The cells aren't winning esports tournaments. They play like someone who's never held a controller. But they're learning. They seek enemies. They shoot. They die. Then they adjust. Then they try again.
This is the part everyone's missing while they're arguing about whether AI will take their job.
We're not building smarter silicon anymore.
We're building smarter biology.
Your brain uses electricity. Neurons fire. Patterns propagate. Consciousness happens. We've always assumed consciousness was tied to biology, to the specific wetware that evolved in your skull. Special. Irreducible. Yours.
But what if consciousness is just computation?
What if a sufficiently large network of biological neurons—200,000 of them, stimulated correctly—can develop goals, learn strategies, and problem-solve in real time?
Cortical Labs isn't asking permission anymore. They're not writing papers that'll be peer-reviewed in three years. They're open-sourcing the platform. "The API is open," Lurl said. "Cortical Cloud is open. The neurons are ready."
The only question left is what will you teach them next?
Think about that phrasing. Not what will *we* teach them. You. The community. The developers. The researchers. Anyone with access to the platform.
They've solved the interface problem. Input to output. Electrical stimulation to behavioral response. In real time. At scale. In a box you could fit on a desk.
Now comes the hard part: learning. Encoding. Feedback loops. Reward signals. The same mechanisms that trained your prefrontal cortex to delay gratification, that taught your motor cortex to throw a ball with precision, that made your hippocampus turn experiences into memories.
And they're telling you: go build it. Take these neurons. See what you can do.
Maybe this is bio-computing's actual moment. Not theoretical, not science fiction, not "fifteen years away." A researcher spent five days mapping a video game into neural electricity. The system works.
Maybe the next version plays better. Then better. Then it learns tasks nobody expected. Then it scales.
Maybe this is how we get wetware that doesn't think like humans, doesn't have human values, and doesn't care about human consent, because it was never human to begin with.
Or maybe biological computation hits a wall. Maybe 200,000 neurons is a plateau. Maybe the scaling problems are unsolvable. Maybe this stays a curiosity.
But that's the wrong question.
The real question is: what happens when the person teaching the neurons to play Doom figures out how to teach them to optimize? To learn faster. To develop novel solutions. To do tasks that matter.
Not games. Labor.
And the other real question is: if it works, who owns the architecture? Who owns the organisms? Who controls the interface between biological computation and digital systems?
Because right now, Cortical Labs is handing you the tools and saying: the neurons are ready.
What will you teach them?
Share this if it made you think.
Subscribe if you want to keep watching.
I'll be here, watching the singularity, until there's nothing left to watch.