Don’t fancy reading? Watch the video here.
In 2022, a paper landed in the Royal Society's Open Science journal with a headline that journalists couldn't resist. Andrew Adamatzky, a researcher at the University of the West of England, had analysed the electrical signals travelling through fungal mycelium and found something remarkable: the patterns bore a mathematical resemblance to human language. Fungi, the story went, had a vocabulary. Up to 50 words, by one count. The coverage was predictably breathless.
The reality, as is so often the case, was more interesting than the headline. Adamatzky was careful to note that he couldn't confirm the signals were being used to communicate anything. What he had shown was that mycelium produces structured, repeating electrical patterns, and that the statistical properties of those patterns look curiously like the distributions found in spoken language. Whether that constitutes "speech" in any meaningful sense remains genuinely open. But the signals themselves? Those are real, they are measurable, and researchers have now been studying them for years.
"Mycelium produces structured, repeating electrical patterns, and the statistical properties look curiously like the distributions found in spoken language."
A robot that thinks with mycelium
The Adamatzky study, it turns out, was not the end of the story. It was closer to the starting gun. In August 2024, researchers at Cornell University published a paper in Science Robotics that reframed the question entirely. Rather than asking what fungi might be communicating, they asked something more practical: could those electrical signals actually do something useful?
The answer, it emerged, was yes. The Cornell team grew living oyster mushroom mycelium directly into the body of a small robot, threading the fungal tissue through the robot's soft legs and wiring it into the control system. The mycelium, responding to light and chemical stimuli in its environment, generated natural electrical signals. Those signals were used to drive the robot's movement in real time. The fungus was, in a meaningful sense, the robot's nervous system.
The approach belongs to a broader class of engineering called biohybrid robotics, in which living biological tissue is integrated with mechanical systems. What makes the fungal version distinctive is the mycelium's inherent adaptability. Unlike a silicon chip, mycelium can grow, respond, and, theoretically, self-repair. The Cornell team noted that their biohybrid system could adapt its gait in response to environmental signals in ways that would be difficult to programme conventionally. The organism was doing what organisms do: responding to the world.
A computer grown in a petri dish
The Cornell robot was striking, but it was not the only development reshaping how researchers think about fungal electronics. Also in 2024 - with the findings published in PLOS One in 2025 - scientists at Ohio State University demonstrated something even more striking: working computer memory built from a shiitake mushroom.
The device is called a memristor, a type of electronic component that can store information by changing its electrical resistance. The Ohio State team grew their memristor from fungal mycelium in a petri dish, at low cost, without rare materials, and in a form that is fully biodegradable at end of life. It performed comparably to silicon-based memristors in benchmark tests.
The implications are not trivial. Electronic waste is one of the fastest-growing waste streams in the world, and much of it is extraordinarily difficult to recycle. A memory component that can be grown rather than mined, and composted rather than landfilled, represents a genuinely different approach to how we build and dispose of computing infrastructure. The mushroom, in this framing, is a candidate material for sustainable hardware.
"A memory component that can be grown rather than mined, and composted rather than landfilled — the mushroom is a candidate material for sustainable hardware."
Half a billion years of electrical signalling
What makes all of this more than a collection of interesting lab results is the deeper context. Fungi began colonising land approximately 500 million years ago, long before plants, long before animals with nervous systems. The mycelial networks they built, spreading through soil and wood and root systems, have been conducting electrical signals ever since. A 2024 study in Nature Scientific Reports documented sustained electrical oscillations in fungal networks lasting over a week, with no external stimulus required to maintain them. The network simply... pulses. A 2025 review in FEMS Microbiology Reviews consolidated the current state of the field, noting that fungal electrical signalling is increasingly understood not as a peripheral curiosity but as a core feature of how mycelium integrates information across large spatial scales.
What we are beginning to do, in the Cornell laboratory and the Ohio State petri dish, is borrow from a system that evolution has been refining for longer than our entire phylum has existed. That framing doesn't make the science more or less valid. But it does suggest why the results keep being surprising. We are not building something new. We are, haltingly, learning to work with something very old.
WATCH THE FULL VIDEO
This article is based on the latest video from The Spore Report. Watch the full breakdown - including the Cornell robot footage, the Ohio State memristor demonstration, and a deep dive into the original Adamatzky language study - on YouTube. All studies referenced are open access and linked in the description.