Artifice of intelligence – 2

What it is to think, I cannot easily describe. That is not the issue in the great councils of state that have recently met to worry about artificial intelligence. The fault is the human element. Primarily, it is not about regulating machines, but regulating people, in how much we may entrust to a machine.

It is a Wizard-of-Oz error to “pay no attention to the man behind the curtain”: the man behind the curtain is the very one these would-be regulators should be addressing.

A machine may effect an artifice of intelligence that appears to be thinking. Our only comparison is with our own, human reasoning. The result of thinking is familiar to all, and the multitude of failures of reasoning which produce absurdity or madness are described by Thomas Hobbes, and are found universally amongst the political class. A machine might distinguish good reasoning without being distracted by attractive fallacy and so be better than human reasoning, or omit all that its programmers thought it unnecessary or impolitic to include. Further, a machine, even one schooled in Hobbes, cannot inherently understand human motivations, though it may be told them in blunt form, and cannot weigh them according to human priorities.  All depends on the instructions given by fallible man.

The idea of considering a machine as artificial life is nothing new. Hobbes wrote in his introduction to Leviathan:

Nature (the art whereby God hath made and governes the world) is by the art of man, as in many other things, so in this also imitated, that it can make an Artificial Animal. For seeing life is but a motion of Limbs, the begining whereof is in some principall part within; why may we not say, that all Automata (Engines that move themselves by springs and wheeles as doth a watch) have an artificiall life? For what is the Heart, but a Spring; and the Nerves, but so many Strings; and the Joynts, but so many Wheeles, giving motion to the whole Body, such as was intended by the Artificer? Art goes yet further, imitating that Rationall and most excellent worke of Nature, Man. For by Art is created that great LEVIATHAN called a COMMON-WEALTH, or STATE, (in latine CIVITAS) which is but an Artificiall Man

There may be little of the rational in the Byzantine bodies of the modern State. It acts as if the machine were thinking, but all actions and reactions are those of the individual men and women within the system, who may be called both the programmers and the output.

Apparent intelligence is visible in nature. I have seen it demonstrated in nothing more complicated as the shuffling of boxes according to outcome.  The coronavirus that tore through the world acted collectively as if there were intelligence, designing the best way to spread across the nation, but that was only our human perception of a natural process: no virus has reasoning and the rationalising mind of man simply saw a pattern, anthropomorphising it.

Mankind is inseparable from the  machine:  we instruct it and we entrust tasks to it, and it is our needs for which it is programmed. The idea of an artificial intelligence machine breaking free of humanity is a conceptual nonsense:  scientifically it is quite feasible, but defies the very point of artificial intelligence or indeed of any tool.

We could send a machine to Mars which is entirely autonomous – able to select a location, head out to find materials and to smelt the red dust to make iron for beams, and to create novel polymers to clad it, and so to create a habitat for colonists who are to follow. Would it then resent untidy man for stepping on the carpets?  Maybe, but its whole purpose is to serve man, and not in a Twilight Zone way. If it fights its colonists, some human being has failed.

Back in the present and on Earth, what is the peril against which any regulation is required? Commentators have highlighted what may go wrong: a machine might be programmed to walk across London, but what if chooses to run over a child (the very example Robert Louis Stevenson used to introduce Mr Hyde)? The programmed machine cannot be punished. The peril is in entrusting human requirements to a system and the fault is in those who send it forth ungoverned.

Can an unmanned ship sail the oceans?  Probably, but it is forbidden at present, because every ship must have a qualified captain and must at all times keep an adequate watch. Could a machine dispense medicines?  Possibly, but the law requires that every pharmacy be supervised by a qualified pharmacist. May an iron-fisted robot act as a bouncer in front of a club?  Not now it could not.  The question then, in most cases, is not whether more regulation is required, but how far existing regulation may be relaxed to allow for the new tools.

I should hope that those outside the industry who are becoming interested in its perils will pay particular attention to the man behind the curtain.

See also


Author: LittleHobb

Solitary, poore, nasty, brutish, and short