Biohybrid computers made of organoids—tiny human brain models using real cells—have reached an important new benchmark by reading braille for the first time, even under challenging conditions.
A group of UK scientists working at Bristol University reported their biohybrid computing achievement in Frontiers in Artificial Intelligence, with a new paper describing the results.
The team’s achievement marks a major step toward the long-sought goal of harnessing organic brain cells for computing. The human brain is far more energy efficient than traditional computers, making the technology highly attractive as large-scale computing operations, like data centers, come under heavy criticism for their massive resource use.
Biohybrid Computer Accuracy
Although biohybrid computers could eventually offer advantages in energy efficiency, demonstrating that they can reliably process and transmit meaningful information remains a work in progress. The Bristol team sought to advance this research by developing organoids modeled on the human forebrain and testing their ability to process information encoded from Braille characters.
The researchers ultimately achieved an 83% classification accuracy rate by combining signals from three brain organoids to identify Braille letters.
“We can take information from a real tactile sensor, turn it into a form that a living neural network can receive, and then recover meaningful information from the organoid response,” said first author Tianyi Liu, PhD candidate at Bristol University. “That gives us a practical way to connect artificial sensors with biological neural systems, which are a very interesting possible route towards more energy-efficient computing.”
Creating a Biohybrid Computer
The researchers developed tiny but complex models of the human brain containing approximately 10,000 neurons and supporting cells, although they lacked many of the structures found in a fully developed human brain.
The organoids were connected to electrodes that allowed researchers to deliver electrical signals and monitor their neural activity. They were also maintained in a controlled environment designed to keep the living cells functioning throughout the experiments.
The next step was determining whether this biohybrid computing system could reliably process recognizable information.
Testing began with a tactile sensor detecting physical Braille characters. The team then converted the sensor readings into electrical patterns, which were delivered to the organoids through the electrodes.
By experimenting with different input patterns and measuring the resulting neural activity, the researchers investigated whether specific Braille characters produced distinguishable responses.
Biohybrid Tests
The researchers then used computational methods, including a large language model (LLM), to classify the organoids’ neural responses and determine whether those signals corresponded to specific Braille letters.
Combining the responses of three organoids produced substantially better results than using a single organoid, increasing classification accuracy from 61% to 83%.
“After processing by the organoids, the tactile data still contains enough information to classify Braille letters with good accuracy,” said senior author Dr. Benjamin Ward-Cherrier of Bristol University. “Combining outputs from several organoids also showed an improvement in performance.”
Exactly why combining signals from three organoids improved accuracy remains uncertain, although the team suspects that differences in their individual neural networks may allow them to provide complementary information.
“Different organoids may have slightly different network connections, developmental states, and ongoing activity, so the same input can produce similar but not identical responses,” said Liu. “Combining those responses may give us complementary information that we would not get from a single organoid. At the same time, if every organoid behaves differently, it obviously makes it harder to build a predictable and standardized system.”
A Computer that May Learn like a Brain
After achieving promising results with clean input data, the researchers introduced an additional challenge: noise, or interference that can make information more difficult to interpret.
Interestingly, when noise was introduced, the Braille characters could be classified more accurately from the organoids’ responses than from the raw sensor data. This suggests that the biological neural networks may help preserve or transform useful information under challenging conditions.
However, the findings do not demonstrate that the organoids learned to recognize Braille in the same way a human brain might. Whether these systems can adapt through experience remains an important question for future research.
“We showed that tactile information could be encoded, processed through the organoids, and decoded, but we did not demonstrate learning or long-term changes in the organoids,” Liu said.
“We also tested one tactile task using a fixed encoding scheme and a relatively simple measure of neural activity. The next step is definitely to move beyond Braille; we are interested in things such as textures, edges, and different kinds of touch, and eventually in closed-loop tasks where the organoid receives feedback from what it does.
“That would allow us to start asking whether the organoid can actually adapt through experience,” Lie said.
The paper, “Encoding Tactile Stimuli for Braille Recognition with Organoids,” appeared in Frontiers in Artificial Intelligence on October 15, 2026.
Ryan Whalen covers science and technology for The Debrief. He holds an MA in History and a Master of Library and Information Science with a certificate in Data Science. He can be contacted at ryan@thedebrief.org, and follow him on Twitter @mdntwvlf.