Machines that go beyond human logic could help us understand how we think – if we can only figure out what to ask them
WHAT is the answer to the ultimate question of life, the universe and everything? In The Hitchhiker's Guide to the Galaxy, aliens build a city-sized computer to find out. They are dismayed when it turns out, bafflingly, to be 42. The problem, the unrepentant computer suggests, is that they never really knew the question... and to find it they will need an even bigger computer.
Like all the best comedy, Douglas Adams's absurdity has an element of truth to it – perhaps more than you might expect. We already have machines that answer our questions in ways we can't fully appreciate: from quantum computers, whose physics remains opaque, to data-crunching black boxes that translate languages and recognise faces despite knowing nothing of grammar or physiology (New Scientist, 10 August 2013, p 32)
But despite their complexity, these computers are all of the type conceived by Alan Turing in 1936 – and they all have the same limitations. Turing showed that any computer predicated on human logic alone will struggle with the same questions that we do. They will always find some questions undecidable: not so much "computer says 'no'" as "computer says 'can never know'".
But Turing also conceived of an "oracle" that might transcend those limitations. Most computer scientists don't think we can ever build one. But a few people are trying, using neural networks (see "Turing's oracle: The computer that goes beyond logic"). If they succeed, we will gain new insights into thought itself – and perhaps into the human brain, whose staggering computational prowess remains deeply mysterious (see "Defending the grand vision of the Human Brain Project").
Conventional computers give us the answers to questions that we can articulate, but don't have the time to calculate. Turing's oracle could address issues we can't even articulate: an echo of Adams's "even bigger computer". It might not provide answers to life, the universe and everything, but even futile attempts to make one could help explain how we think about them – and figure out the right questions to ask.
This article appeared in print under the headline "Deep thoughts"
- New Scientist
- Not just a website!
- Subscribe to New Scientist and get:
- New Scientist magazine delivered every week
- Unlimited online access to articles from over 500 back issues
- Subscribe Now and Save
If you would like to reuse any content from New Scientist, either in print or online, please contact the syndication department first for permission. New Scientist does not own rights to photos, but there are a variety of licensing options available for use of articles and graphics we own the copyright to.
