In a mind-bending Australian breakthrough, 800,000 human brain cells in a lab have learned to play a video game, displaying 'measurable intelligence' and challenging the very definition of life.
MELBOURNE, AU — In a landmark achievement that blurs the line between biology and artificial intelligence, Australian researchers have successfully grown a "mini-brain" in a laboratory that taught itself to play a video game all without a body.
The project, dubbed "DishBrain" by its creators at the tech start-up Cortical Labs, consists of a living cluster of 800,000 human brain cells grown atop a microelectrode array. In a stunning demonstration of raw biological learning, these cells learned to respond to and play the classic video game Pong.
What's more, they learned to play in just five minutes a speed that outpaces many conventional AI algorithms.
How to Teach a Brain in a Dish
The mechanism behind this breakthrough is a masterclass in bio-digital communication. The DishBrain's "world" is the microelectrode array, which sends and receives information via electrical signals.
Here is how the team taught the cells to play:
Simulation: The cells were sent electrical signals that simulated the position of the ball in the game of Pong.
Reward (Order): When the cells' neural activity "hit" the ball back, they received a structured, orderly electrical signal a form of positive reinforcement or "reward."
Punishment (Chaos): Conversely, when the cells "missed" the ball, they received a chaotic, random, and unpredictable electrical signal the equivalent of a "punishment."
Through this simple feedback loop, the cells began to learn. They progressively adjusted their own internal neural firing patterns to become more adept at hitting the ball, thus avoiding the chaotic "punishment" signal and seeking the orderly "reward."
"Measurable Biological Intelligence"
This experiment shows far more than a simple reflex. The cells demonstrated a genuine ability to learn, adapt, and improve their performance.
Researchers noted that the DishBrain not only got better at the game but also began to develop its own rudimentary strategies. It even exhibited responses that the team likened to "frustration" or "disappointment" when it "lost" a point.
The team is quick to clarify that these cells are not conscious. They do not possess self-awareness, feelings, or thoughts. However, they are unequivocally expressing a form of measurable biological intelligence, an inherent ability of brain cells to process information and adapt their behavior to achieve a goal.
The Implications: Biological Computers and the Definition of Life
This achievement has profound implications that ripple across neuroscience, AI, and even philosophy.
On a practical level, DishBrain offers a revolutionary new platform for:
Drug Testing: It allows for testing new medicines, especially for neurological conditions like Alzheimer's or epilepsy, directly on functional human neural tissue without needing animal test subjects.
Biological Computing: It pioneers the development of a new type of computer, one that merges living tissue with silicon. These "biological computers" could one day become far more efficient and powerful than traditional AI, which requires massive amounts of energy to "learn."
On a deeper level, the project forces a reckoning with our most basic questions. As the Cortical Labs team summarized: "If brain cells can learn without a in body, then what truly makes us human?"
This breakthrough challenges the boundaries between intelligence and consciousness, forcing us to redefine what "learning" and even "living" truly mean in the age of advanced biotechnology.
CaliToday.Net