A new breakthrough in neural networking might just lead to truly intelligent computers. Dubbed a ‘super-Turing’ network, the new approach makes the neural networks so common to artificial intelligence research work very much like how our brains do.
Hava Siegelmann of the University of Massachusetts Amherst is responsible for the breakthrough. Speaking of her model, she said:
“This model is inspired by the brain. It is a mathematical formulation of the brain’s neural networks with their adaptive abilities.”
The model is based on a previous theoretical model that she posited in 1993. Her super-Turing neural networks are capable of learning and morphing, completely rearranging their design every time a new fact is learned. Which is huge. It means that her neural network learns an order of magnitude more effectively and faster than more traditional neural networks. And, unlike traditional neural networks, Siegelmann’s model thrives when exposed to constant stimulation. Says Siegelmann,
“Each time a Super-Turing machine gets input it literally becomes a different machine. You don’t want this for your PC. They are fine and fast calculators and we need them to do that. But if you want a robot to accompany a blind person to the grocery store, you’d like one that can navigate in a dynamic environment. If you want a machine to interact successfully with a human partner, you’d like one that can adapt to idiosyncratic speech, recognize facial patterns and allow interactions between partners to evolve just like we do. That’s what this model can offer.”
Siegelmann’s model is superior not just in ability but also in cost. Unlike traditional neural networks that sample an entire scene when processing an image, Siegelmann’s model is able to quickly figure out which information is relevant and look only at that. That means that significantly less computational power is needed for each network. Explaining herself, Sigelmann said
“I was young enough to be curious, wanting to understand why the Turing model looked really strong. I tried to prove the conjecture that neural networks are very weak and instead found that some of the early work was faulty. I was surprised to find out via mathematical analysis that the neural models had some capabilities that surpass the Turing model. So I re-read Turing and found that he believed there would be an adaptive model that was stronger based on continuous calculations.”
Later she added,
“If a Turing machine is like a train on a fixed track, a Super-Turing machine is like an airplane. It can haul a heavy load, but also move in endless directions and vary its destination as needed. The Super-Turing framework allows a stimulus to actually change the computer at each computational step, behaving in a way much closer to that of the constantly adapting and evolving brain.”