Science has created an improved mouse, controlled by an artificial intelligence (AI) brain.
This virtual mouse, developed by researchers at Harvard and Google's DeepMind AI Lab, is capable of accurately mimicking the movements of real mice, a breakthrough that will advance our understanding of Promises how the brain controls complex, coordinated movements.
It's a feat that even today's most advanced robots struggle to replicate, and the team believes their findings could greatly improve the agility of future robots.
Professor Bence Olwiczki from Harvard's Department of Biology and Evolutionary Biology led the effort, using high-resolution data recorded from real mice to train artificial neural networks. His lab is devoted to mechanistic research into how the brain moves limbs.
“We found that neural activity… was better predicted by the network activity of the virtual rat than any feature of the real rat's movement, implementing the inverse dynamics of the two regions,” the researchers wrote. consistent with,” the researchers wrote.
According to research published in the journal The nature, the virtual rat simulation was developed using the MuJoCo physics simulator, which added realistic forces such as gravity to simulate real-world conditions. The artificial neural network that drives the movements of virtual mice was trained on inverse kinematics models, enabling it to predict neural activity in real mice with high accuracy.
The findings could help scientists interpret neural activity in different behaviors and link it to the principles of motor control, the report said. Principles derived from the virtual mouse study have implications for the development of better robotic control systems.
“These results show how biomechanically realistic virtual animal physiological simulations can help interpret the structure of neural activity in behavior and relate it to theoretical principles of motor control,” the research reads.
Image: Google DeepMind
“We've learned a lot from the challenge of building embodied agents: AI systems that not only have to think intelligently, but also translate that thought into physical action in a complex environment,” said Matthew Botwink of Google DeepMind. Is”. Harvard Gazette. “It seemed plausible that adopting the same approach in a neuroscience context might be useful for providing insights into both behavior and brain function.”
The team used advanced techniques such as AI, deep reinforcement learning, and 3D motion tracking to enable the virtual mouse to mimic a wide range of natural behaviors, including those for which it was explicitly trained. was not given This approach could establish a new field of “virtual neuroscience,” the paper noted, providing an accessible platform for studying the neural bases of natural behavior in order to better understand it. Learn how the brain controls movement.
This area of study is important for developing advanced prosthetics and brain-machine interfaces such as NeuralLink or technology developed by Precision Neuroscience.
Insights from this work could lead to new treatments for movement disorders by rewiring neural circuits. Additionally, the research suggests that the virtual rat offers a transparent model for studying neural circuits and the effects of diseases on those circuits.
Going forward, the researchers plan to give virtual mice the autonomy to solve tasks faced by real mice, furthering the understanding of brain algorithms for skill acquisition.
Ölveczky was unable to provide additional comment. Decrypt.
Edited by Ryan Ozawa.
Generally intelligent Newsletter
A weekly AI journey narrated by General, a creative AI model.
