Biological Computing with Cortical Labs

🧠 🎮 Brain Cells In a Petri Dish Play Video Game with Cortical Labs

Find out more below…

1. Teaching “Mini Brains” to Play a Video Game 

Scientists at Singaporean deep technology startup Cortical Labs have embedded a collection of nearly 1 million living human brain cells onto the surface of a silicon chip. The chip reads the neural activity and also stimulates the neurons to perform goal-orientated tasks in the video game pong.

2. Organoids 

Organoids are miniaturized and simplified versions of an organ product in vitro. Cortical Labs produced these mini human brain cells by growing them from human-induced stem cells. These cells are integrated into computer chips to respond to pulses of electricity in the video game. Remarkably, human-derived organoids outperformed mouse-derived organoids in certain aspects of the game.

3. Biological Computing

The neurons are embedded onto a computer chip with thousands of mini electrodes for noisy signals to be sent to the neurons for incorrect behaviour such as when missing the pong ball. A simple and predictable signal is sent when the paddle hits the ball. The mini brains adapted and learned how to better play the pong game demonstrating the ability to learn.

4. Free Energy Principle 101

Cortical Labs apply a theory called the Free Energy Principle, developed by Professor Karl Friston at UCL, which states the brain always seeks to minimize the error between its predictions and observations. This is achieved by either changes its predictions or by changing its behaviour in the environment to change its observations.

5. Neurons Learn

The Free Energy Principle gives an explanation of how the neurons learn to hit the ball in the pong game. During training, the neurons modify their behaviour in order to avoid receiving the noisy signals, which are sent to the neurons when the ball is missed and move the paddle to hit the ball with a predictable signal. Over time, the mini brains adapt to move the paddle to deflect the balls in order to avoid the noisy signal.

Click this link to see the Spike Visualiser showing real-time neural network activations when playing Pong: https://lnkd.in/ewMeKAwD

Related Tags

Biological Computing




What SDG is this related to?


No items found.
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.