Quote of the Day

Researchers at Australian start-up Cortical Labs have taught human neurons grown on a chip to play the classic Doom game. In 2021, they had already used 800,000 neurons to play Pong. Now, with four times fewer brain cells, they can play a much more complicated game.

But the wildest thing about it is probably the “Cortical Cloud,” an interface that allows developers to program the neurons using Python. Using this API, independent researcher Sean Cole was able to teach the cells to play Doom in about a week.

…

The technology making this possible is the CL-1, a “biological computer” chip. The biological component of the CL-1 system consists of human neurons derived from induced pluripotent stem cells (iPSCs). These cells, typically repurposed from adult skin or blood samples, are differentiated in a lab into functional cortical neurons. Once matured, approximately 200,000 of these living cells are integrated onto a high-density microelectrode array (HD-MEA), essentially creating a specialized silicon chip featuring thousands of interface points.

Mihai Andrei
March 2, 2026
200,000 living human brain cells just learned to play Doom and this is just the start of it

With only a two hundred thousand neurons Copilot says this is comparable to the CNS of:

A small crustacean (e.g., a copepod or amphipod) is the best biological analogue to a 200,000‑neuron system.

• More complex than worms
• Less complex than insects
• Within the right order of magnitude
• Capable of simple learning and coordinated movement

I could support a claim the ethical issues are nonexistent at this level. What about when it is comparable to a small mammal? Or a primate? A human? An order of magnitude larger than any mammal?

What will be the capabilities of a Cortical Cloud requiring a warehouse the size of a city block?

We live in interesting times.