[KCLPASS Columns 2016-2017 Issue #3]
The immediate reaction the term ‘artificial intelligence’ conjures up in all of us depends on the type of films we grew up watching. On the one hand, we have the helpful, lovable, and endearing droids from Star Wars such as R2-D2, and on the other, we have the apocalyptic dystopian death machines of Terminator. A closer inspection of the two depictions reveals the two extremes of people’s projections about the future of artificial intelligence: our helpful servants, or a serious problem that will come to bite us in the ass a hundred years down the line.
There is no better time to seriously consider this future than the present. We are in a golden age of technological advancement, with no end or limit in sight. As it stands, computer chip performance has doubled approximately every two years since the beginning of the 21st century, with computers capable of hosting programs sorting through petabytes upon petabytes of information not far off. Development of the software in question has also been advancing in leaps and bounds. Thus, we must address the potential issues raised by artificial intelligence now, before it grows its own legs (a very possible outcome).
A major concern of the public is that robots will eventually take up jobs that could otherwise be taken up by human beings. In no doubt, this would catapult unemployment rates to unprecedented heights, churning out as many problems as its creators believed it might solve. Another even more chilling argument against artificial intelligence is that a truly sentient computer may eventually supersede humanity’s dominion over planet Earth. Professor Stephen Hawking warned against development of a program that would rival or even surpass human intelligence on the basis that while human beings are limited by the slow rate of biological evolution, AIs will be able to develop and improve at ever-increasing rates. Humans simply cannot compete.
Rules and regulations must therefore be determined as soon as possible, before any major breakthrough in technology brings us to the point where it is already too late to consider alternatives. One need only look to Microsoft’s disastrous release of an AI on Twitter in March to see the pressing need for regulation. Otherwise, AIs releasing racist, sexually charged messages on social media would be the least of our worries.
Author: Aidan Shum
KCLPASS Sub-committee member 2016-17
(The view of the author does not represent the position of the Society. KCLPASS is a politically neutral society.)