There has been a belief for a while now—that AI isn’t actually real AI. ChatGPT and similar tools are basically just fancy search engines. They analyze vast amounts of data, interpret patterns, and generate responses. They can mimic intelligence, yes, but they don’t have consciousness. And honestly, that fact was strangely comforting, even if it made them a little boring. You could turn them off, reset them, wipe their memory, or feed them new data without a second thought. There was no ethical dilemma. No soul in the machine.
But something is changing. Rapidly.
There’s a new world emerging—a world of biocomputation, where the processing is no longer done by chips, but by human brain tissue. Tiny human brains—yes, actual human brains grown from stem cells in labs, suspended in Petri dishes like eerie, living computer parts. And now, they’re learning, adapting, and even “thinking” in ways traditional AI never could.
A Timeline of Organoid Intelligence
- 2022: Brain Cells Learn to Play Pong: Australian researchers created a mini brain with 800,000 neurons and connected it to electrodes. When placed in a virtual game of Pong, the organoid learned how to play on its own—no programming required. This wasn’t just a programmed algorithm. It was a biological brain figuring things out.
- 2023: The Rise of FinalSpark’s Neuroplatform: Swiss company FinalSpark launched the world’s first commercial biocomputing platform using live brain organoids. You can now rent computational power from actual human cells. These aren’t traditional CPUs—they’re biological, self-learning networks.
- 2024: Butterfly Simulation: In an even more surreal experiment, brain organoids were plugged into a virtual world where they became butterflies. Not metaphorically. The organoid experienced the environment and learned how to move the butterfly’s wings in real-time. No instructions. Just experience.
The Shift from Code to Cognition
Organoid intelligence is not just another computing advancement. It changes the game entirely. These systems grow. They develop neural pathways. They don’t just simulate learning—they learn the way human brains do.
And most importantly: you can’t just reset them.
Once they’ve developed behaviour patterns or formed primitive memories, those patterns persist. There is no wipe-and-retrain cycle. There is no hard reset. That makes them not only more lifelike but, in some ways, more alive.
This opens up profound questions about their future use. They’re already being embedded into simple robotic systems. The next step? Integrating them into humanoid robots that don’t just follow code but learn and adapt like living beings.
As this technology advances, one question becomes harder to ignore: are we creating something that can suffer?
That’s a question for the next article.