The real problem of Artificial Intelligence

We are told that Artificial Intelligence cannot be creative, at least not yet. That depends on what you mean by creative. The best definition I know – “Creativity is the ability to perceive relationships” – leaves AI romping home with any relationships that have already been published. AI cannot perceive a relationship that has not been published. But be realistic, all but the very rarest of relationships are already known. The work that most creative people do is to quickly associate things that are not normally – or in other people’s minds – related.

Granted, there are relationships yet to be discovered but the act of sparking them is not all that difficult. We teach people creativity in three simple lessons. Such creativity doesn’t instantly produce a panacea for all mankind but great solutions are increasingly the outcome of lesser solutions linked together. Most creativity is a combination of memory and being quick-witted. Such work is already being overtaken by AI.

This will put people out of jobs and, worse, will leave several without a purpose in life. That will take some addressing. Idle hands make mischief unless they are directed to something useful. It’s a challenge, but one we can probably meet. We often do not know our needs until we have solutions to them. The Internet is a good example.

It is AI’s destructivity that we need to worry about more. War has always been a contest of strategy and quick wits. This means it has been mostly limited by the speed at which humans think. AI can think almost infinitely faster than we can. Simulated battles between fighter jet aircraft show the incredible speed with which AI can detect and destroy another plane, or any object, mobile or stationary, however responsive it is to threats. Add pilotless drones to the list.

In a world where country supremacy seems to be the goal, such aggression is inevitable just as the present ramping up of economic hostility between East and West is self-perpetuating. It is the ‘ramping-up’ that is disturbing. The world has maintained nuclear peace in spite of many local wars since the 1945 atomic bombings of Japan. Apart from ‘non-nuclear agreements’ the reason for this has been the scale of immediate destruction further use of more powerful weapons would inflict. Big nuclear weapons today would eliminate much of the planet’s population and certainly severely limit its habitable areas.

However, AI allows for much faster tactical thinking which creates quicker engagement – maybe too quick for human intervention. In a battle over, say, one single territory the pace will be such that human intervention will be unable to contain it. Even if this didn’t lead to a world war it would certainly create enough local wars to have the same collective effect. 

Ageing brings with it at least one thinking advantage, the realisation that to ponder before responding is often well worthwhile. The process of pondering allows our emotions to take their part in our deliberation, enables the ‘hunch’ aspect of our minds to kick in and empowers the atmosphere of our exchanges to exert an influence. It is what makes us human.

AI has no emotions, no hunch, no atmosphere of the sort we have. The problem is that it has infinitely greater speed.

Too fast for us? Certainly.

Too fast for peace? 

I really don’t know.

Good morning

John Bittleston

15 August 2023