Brain Organoids Emerge as Living Alternative to Silicon AI: Lab-Grown Neurons Play Games and Guide Robots
Summary
- • Lab-grown human brain organoids are playing Pong, guiding robots, and forming biocomputers
- • Researchers at UCSD, Johns Hopkins, and a Melbourne startup are leading organoid computing
- • Organoids produce infant-like brain waves after 8 months as neurons self-organize to connect
- • Scientists argue biological intelligence could become a long-term alternative to silicon AI
Details
What brain organoids are
Globs of millions of neurons grown from reprogrammed skin cells. Kept at 98.6°F for 8 months, they produce brain waves nearly indistinguishable from those of a premature infant.
UCSD: robots and psychedelics
Alysson Muotri's lab at UCSD Sanford Stem Cell Institute grows tens of thousands of organoids, using them to guide robots through mazes and study psychedelic drug effects on neural activity.
Johns Hopkins: biocomputing systems
Johns Hopkins researchers are building novel biocomputing architectures using organoids as a core substrate — hybrid biological-digital systems that are generating serious academic interest.
Cortical Labs: Pong and Doom
A Melbourne-based startup has achieved real-time game-playing in organoids, demonstrating adaptive responses to Pong and Doom — feedback-driven learning without traditional software training loops.
Neanderthalized organoids and ISS missions
Muotri's team revived DNA from hominin fossils to create 'Neanderthalized' organoids, and has sent organoid payloads to the International Space Station to study cosmic radiation effects on neurons.
Neurons innately seek connection
'Whatever environment you put them in, the first thing that they do is try to connect,' says Muotri. This self-organizing property — forming synapses spontaneously — is what makes organoids compelling as computing substrates.
Medical goals drive the research
Much organoid research is motivated by medicine. Muotri, whose 18-year-old son is autistic, uses organoids primarily to study autism's neural basis and test treatments — biocomputing is an emergent application.
Key research centers, capabilities, and science behind brain organoid computing
What This Means
While the AI industry pours capital into scaling silicon-based large language models, a parallel research movement is returning to biological fundamentals — growing actual human neurons and programming them to compute using electrical signals and neurotransmitters. Brain organoids are still far from replacing GPUs, but their demonstrated ability to play video games, navigate mazes, and underpin biocomputing architectures suggests this is now a legitimate research paradigm rather than a curiosity. If biological computing proves significantly more energy-efficient or neurologically adaptive than silicon, it could reshape our conception of machine intelligence altogether — not artificial intelligence, but grown intelligence.