Robots become self-aware, but think that they're rats, so we're okay for now

Scientists at the University of Reading have got a step closer to creating Cybermen – human brains inside a robot – after putting together a small robot that uses rat neurones to control itself.

The scientists have managed to grow around 300,000 rat neurones artificially in the lab by starting off with the brain of a rat foetus. These neurones have gone on to make connections with each other and work in much the same way a regular rat brain does, using electrical impulses to make the brain “do stuff”. The neurons are connected to a regular microchip, where they can be stimulated and the results analysed to see what happens. For example, they’ve built a robot on wheels with an ultrasound sensor, to spot when it is approaching a wall. I guess it’s like giving a rat the Bat-power of echo-location. Maybe.

Another interesting bit of trivia is that it seems that the artificial brain can get bored, just like a “real” brain – apparently if it doesn’t receive any information from the electrodes it’s hooked up to, the neurones will start breaking connections, and similarly, if the neurones are stimulated, then more connections will be formed.

Whilst this all sounds like great fun – who wouldn’t enjoy explaining that their job is to make robots and Frankenstein’s monsters – it does serve a serious purpose. They reckon that if they can make the neurones respond to information sent by sensors on the robot, then the technology could be adapted to help people with Alzeheimer’s, Parkinson’s and epilepsy. Just… don’t ask me to explain how though (I have an arts degree).

Expect to see rat-brained robot pets on the shelves in time for Christmas.

(via New Scientist)

Related Posts: Wall painting robot | Heart robot (not a rat’s heart, unfortunately)