An "oracle" computer proposed by celebrated mathematician Alan Turing in 1938 that goes beyond conventional human logic is the focus of a Missouri State University project exploring the possibilities of a super-Turing computer.
Turing himself showed that a universal computer based solely on logic would inevitably encounter undecidable problems where a straight answer is impossible. Later research by Hava Siegelmann focused on neural networks, demonstrating these brain-mimicking circuits can learn as they go along by using their outputs to alter their input weightings until the assigned task is performed optimally. Missouri State researchers Emmett Redd and Steven Younger subsequently have collaborated with Siegelmann on developing a chaotic neural system that responds very sensitively to small changes in its initial conditions. Such a system is driven by a random, infinitely variable noise.
The researchers currently are working on two chaotic machine prototypes — a neural network based on standard electronic elements, and a network that encodes its data in light. Redd and Younger hope an oracle computer such as Turing envisioned could yield insights into quantum theory and the universal limits of computation, as well as the human brain.
From New Scientist
View Full Article
Abstracts Copyright © 2014 Information Inc., Bethesda, Maryland, USA
No entries found