Worm's mind put into a Lego robot.

I saw an article about some people who recently put a worm's brain into a Lego robot and watched as the robot acted like the worm without programming. Turns out it wasn’t quite that cool. I guess I fell for some click-bait, but it was still pretty cool, and the idea does bring up some deep questions.

What actually was done, is all the neural networks within the one millimeter long nematode worm were replicated virtually. They chose this particular worm for its simplicity, with only 302 neural connections comparing to a human brain’s many trillion. Its neurons were replicated via an object oriented program, using UDP to send messages in a way that very closely mimicked that of the communication within the brain. This program was then dumped into a lego robot, and connected properly to sensors and motors so it could react accordingly to proper stimuli.

It turns out it worked. What does that mean exactly? When this simulated neural network was put into the robot and turned on, the robot started acting like the worm. It responded like the worm was observed to react to the equivalent stimuli. It would detect walls and turn away from them, and if it detected food it would move toward it. These functions seem rather trivial, and wouldn’t be hard to program into a machine, but the point of it was it wasn’t programmed. No additional behavioral programming was added on top of the mimicked neural network. The behavior came solely from the ‘brain’.

This virtual neural network, known as the connectome, is being transferred to a Raspberry Pi and a self-contained Pi robot is being built. The idea is this kind of application would have use as a mobile sensor, to explore an environment and report back results. If a far more complicated neural network could be replicated, and an appropriate robot built, an incredibly effective mobile sensor, like suggested, could likely be built, without the need for any AI programming.

As the article (linked below) expresses, this brings up some philosophical questions. Is the robot a nematode worm in a different body, or something entirely different? Is it truly alive? These questions are trivial in this particular case, but if this connectome can be expanded to replicate much larger and more complex animal’s neural networks, these questions will carry a lot more weight.

Do you think it would be possible to get to the point to replicate a human’s neural network? It sounds like something out of science fiction. In fact it indeed is something out of science fiction. Creative AI created by mimicking the neural network of the human brain isn’t a new idea. If this is the beginning of that reality though, what do you think will be found? Do you think a person is just the sum of his or her neural networks, or something more? The answer may be different for humans vs. other animals, or it may not be. But the mere thought of finding out is rather haunting, to be honest.

source:

http://www.i-programmer.info/news/105-artificial-intelligence/7985-a-worms-mind-in-a-lego-body.html