0
If machines with Artificial Intelligence have the potential to turn into monsters, as the philosophers argue, don’t you think it is perfectly valid in our context to wonder if human beings without basic intelligence do have the potential to turn into monsters? Perhaps not in as many words and not at the conscious level, but if you have any kind of regular interaction with Pakistani students at the university level, the thought would have surely crossed your mind.
Lest it be mistaken, universities alone are not to blame. It is the lack of momentum the students bring from the preceding dozen years that lies at the heart of the problem. The universities only add four more years to the dozen, which, come to think of it, adds to the individual’s chance of turning into a monster.
Things being what they are, it looks and sounds safer to talk about the mechanical monsters rather than their human counterparts. Over the last couple of weeks we have traversed a path that covered six decades; from the predictions of Alan Turing to the belief of Herbert Simon and right down to the progress reported in recent times by Ross King and Hod Lipson. Dotting the path were milestones like Aaron, the computer artist, Deep Blue, the computer that beat world chess champion Garry Kasparov, Asimo, the human-shaped robot, and the unnamed machine that independently and successfully conducted experiments on the genetic makeup of baker’s yeast. This, in short, has been a journey through the fascinating world of Artificial Intelligence.
There are two basic questions which have surfaced in the face of these developments. One, what should a machine should actually do in order to qualify as an ‘intelligent’ machine? Two, are scientists about to go out of vogue?
As for the former, philosophers argue that any ‘intelligent’ entity has to be in possession of high-level behaviours involving language understanding, planning and reasoning. The successes thus far, though spectacular, have been the result of one-off projects. What the philosophers are basically saying is that a computer can only be called ‘intelligent’ if it has the capacity to, say, create a computer on its own. This, naturally, is a pretty high benchmark they are setting. After all, it took man himself thousands of years to be that intelligent. It may take many more for machines to reach that stage.
The second question is far more interesting. There is a general consensus within the community that robo-scientists are not a threat to those who are trying to create and perfect them. King, for instance, hopes the application of intelligent robotic thinking to the process of sifting tens of thousands of compounds for potential new drugs will be particularly valuable in the hunt for treatments for neglected tropical diseases like malaria.
Lipson’s thinking is along similar lines. Machines, he says, could take over much of the routine work in research laboratories. At a later stage, they could be put to work on unravelling complex biological systems, designing new drugs and modelling the world’s climate or understanding the cosmos, but only on the basis of their programme input for which they will always need a human being.
This, in a way, answers the first question as well. Intelligence per se has to be viewed differently from Artificial Intelligence which is, and will always be, just that; artificial. The philosophers don’t have a problem with this. But they have two basic arguments against what they call fiddling with nature: one, that there are already too many human beings living in poverty and there is little or no reason to create mechanical labourers; two, with autonomous devices that mimic human form, thought and behaviour, will we entrust our children, educational institutions, businesses, and governments to reasoning machines as well? Their fear is that creating Artificial Intelligence when man does not fully understand the dynamics of intelligence will lead to more confusion and possibly disaster than good.
The two groups seem to be far apart, but there still is an area of consensus between the two. Everybody agrees that regardless of the level of designed intelligence, machines shall always be absolutely subservient to the humans. The consensus is fine, but therein lies the catch; what happens if the ‘intelligent’ machines go awry? After all, systems do collapse and misbehave at homes, offices and even in laboratories. What if a machine with too much ‘intelligence’ goes erratic on a grand scale? We have all heard of the famous, or rather the infamous Frankenstein. Haven’t we? It has a frightening resemblance to possible scenarios in an environment of robots moving alongside humans.
At another level, the philosophers wonder which ethical and moral values the scientists will instil in their robots. Looking at human civilisation with its diverse cultural, religious, ethical and moral values, what exactly are we trying to create and to what purpose? Do we need religiously-biased robots? Does human society need the perfect Catholic, Muslim or Buddhist mind? Or do we want a mind that is ruthlessly calculative; the perfect Capitalist or Efficiency Expert?
And what about the practical applications of these values? If one set of ethical or religious values dictates that we move in a certain direction, and another dictates that it is imperative to go the other way, isn’t it like duplicating current issues without any real answers?
Perhaps Artificial Intelligence will show the same diversity as humans. So what would be the point then of creating artificial humans? Don’t we have problems enough with the biological beings, they argue. And who can argue with that?
At our end, the task seems simpler … much simpler. All we have to do is to produce biological beings with basic intelligence. Failure is bound to leave us with a straight choice between mechanical monsters and human monsters. Destruction guaranteed in either case!
