In 1968, Philip K. Dick wrote the story Do Androids Dream of Electric Sheep?, about a future where a bounty hunter is tracking a group of androids. The book dealt with themes of identity and artificial intelligence, and was later developed into the movie Blade Runner. The question this article asks is: can machines actually learn, and how closely does this resemble our own intelligence?

Artificial Intelligence, or A.I., has been a long-running stale of science fiction, and that is something of an understatement. The notion of computers that possess the capacity to learn and act on independent thought is a source of both wonder and fear – wonder at the notion of a new, truly intelligent artificial lifeform, and fear at the idea of such intelligence being alien to our own, or even hostile. Would such intelligences hold the same values as our own? Would such an intelligence harbour such qualities as empathy, concepts such as love?

Such points, of course, become moot if such a concept is not feasible, remaining the stuff of science fiction. Is there any veracity to such ideas? Is artificial intelligence on such a level possible?

Concepts such as true intelligence and reasoning are open to debate – when does the simulation of such processes truly qualify as such? To understand the question better, we must look at the way our brains learn and process information, and compare it to possible A.I. systems.

Of course, this becomes a circular reasoning process – we aim to create intelligence within the confines of our own intelligence, as we find it difficult to conceive of any way of learning or thinking but our own! This is part of the reason that distance learning is so popular – we enjoy and understand the way we like to learn, as opposed to learning using someone else’s methods, and many find it far more natural than trying to think within someone else’s framework.

As a result, the more successful A.I.s we create will likely approximate our own minds, and likely be structured in a similar fashion, telling us as much about the way our own minds work.

As to how close we are to achieving such lofty goals – we are not in danger of the machines taking over just yet, but early success has been made with specific tasks. We have robots that can react to a range of stimuli, navigate obstacles and even play ping-pong, to name but a few.

While there is still a wide variety of schools of thought in how to approach the problem and which model to adopt when producing such intelligence, it is hard to doubt that one day machines will be capable of independent, creative thought. As to the question of whether this will be true thought, or merely a simulation… by then it may be very difficult to tell the difference!