top of page
Writer's pictureSofia Ng

Teaching Jelly to Play Pong?

Imagine a blob of jelly—yes, jelly—playing a game of Pong and getting better at it with practice. Sounds like science fiction, right? Well, it’s not. Researchers from the University of Reading have managed to do exactly that, and the implications for future tech, especially robotics, are fascinating.



How Does It Work?

This isn’t your typical store-bought jelly. The team used a special polymer gel infused with ions, making it responsive to electric currents. When electricity is passed through, the ions shift, causing the gel to expand or contract. By strategically placing a grid of electrodes around the jelly, the scientists could interpret the movement of these ions to represent the paddle’s position in a game of Pong. Think of it as jelly with an attitude—learning, adapting, and improving at the game as it goes along.


The Power of Memory (in Jelly)

What makes this jelly so unique is its ability to remember. Unlike traditional neural networks that require complex algorithms and mountains of data, this blob relies on a very basic form of memory. As it plays, the swelling and shrinking patterns of the gel change based on previous movements. In essence, it "remembers" how the ball moves and reacts accordingly, showing improvements in gameplay over time. In fact, with practice, it boosted its accuracy by up to 10%.


Potential Uses: Jello-Brained Robots?

Now, this might sound like a quirky experiment, but the implications are huge. If jelly can learn and adapt like this, it opens the door to new ways of creating robots or computing systems that are incredibly energy-efficient. Imagine robots with brains that aren't silicon-based, but powered by materials like this jelly, which learn and adapt over time. It's not just about building better Pong players—this kind of tech could revolutionize how robots respond to their environment.


What’s Next?

While it’s still early days, the researchers believe that their jelly could be trained to handle more complex tasks. So, while we're not expecting jelly-controlled robots anytime soon, this breakthrough shows just how creative—and weird—science can get when exploring the future of computation.


Link to the New Scientist Article


Link to the research article

Recent Posts

See All
bottom of page