Transcript
Today's modern computers are based on mathematical concepts first
laid out in the 1930s by British mathematician Alan Turing. Now, Hava
Siegelmann of the University of Massachusetts Amherst, an expert in
neural networks, has taken Turing's work to its next logical step.
She is implementing her 1993 discovery of what she has dubbed
"Super-Turing" computation into an adaptable computational system that
learns and evolves, using input from the environment in a way much more
like our brains do than classic Turing-type computers. This
breakthrough was described recently in the journal Neural Computation.
The Super-Turing model is a mathematical formulation of the brain's
neural networks with their adaptive abilities.
When the model is
installed in an environment offering constant sensory stimuli like the
real world, and when all stimulus-response pairs are considered over the
machine's lifetime, the Super-Turing model yields an exponentially
greater repertoire of behaviors than a classical computer.
Therefore
the Super-Turing model is superior for human-like tasks and learning
applications.
Each time a Super-Turing machine gets input, it literally becomes a
different machine. You don't want this characteristic in a conventional
PC, which is essentially a super-fast calculator. However, 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.
In any application where a machine needs to interact successfully
with a human partner, you want one that can adapt to idiosyncratic
speech, recognize facial patterns, and allow interactions between the
human and machine partners to evolve just as it would between two
humans. That's what a Super-Turing machine offers.
Classical computers work sequentially and can only operate in the
very orchestrated, specific environments for which they were
programmed. They can look intelligent if they've been told what to
expect and how to respond. But they can't take in new information or
use it to improve problem-solving, provide richer alternatives, or
perform other higher-intelligence tasks.
Back in 1993, Siegelmann discovered via mathematical analysis that
the neural network models had some capabilities that surpassed the
Turing model. Each step in Siegelmann's model starts with a new Turing
machine that computes once and then adapts based on that computation.
The new Super-Turing machine will not only be flexible and adaptable,
but economical. This means that when presented with a visual problem,
for example, it will act more like our human brains and choose salient
features in the environment on which to focus, rather than using its raw
power to visually sample the entire scene using a camera. This economy
of effort, using only as much attention as needed, is another hallmark
of high artificial intelligence.
Siegelmann and her colleagues are now in the process of designing and building the first operational Super-Turing computer.
|
