Google’s secretive artificial intelligence researchers have revealed a computer that they hope will one day be able to program itself.
Developers at Google’s secretive DeepMind start-up, which it bought for $400 million earlier this year, are attempting to mimic some of the properties of the human brain’s short-term working memory.
By combining the way ordinary computers work with the way the human brain works, the researchers hope the machine will learn to program itself.
Described as a ‘Neural Turing Machine’, it can learn as it stores memories, and later retrieve them to perform logical tasks beyond those it has been trained to do.
‘We have introduced the Neural Turing Machine, a neural network architecture that takes inspiration from both models of biological working memory and the design of digital computers,’ the researchers wrote.
‘Our experiments demonstrate that it is capable of learning simple algorithms from example data and of using these algorithms to generalise well outside its training regime.’
The new computer is effectively a type of neural network that has been adapted to work with an external memory.
The result is a computer that learns as it stores memories and can later retrieve them to perform logical tasks beyond those it has been trained to do.
Click here to read the full article