The world is still divided about whether AI is a power for good or evil. Talk to an 18 year old and he will say that the end of the world is nigh. Talk to (or read) more mature thinkers and they are beginning to produce opinions that are somewhere in the middle, but sensibly so.
One article that provided a laugh on a wet Tuesday afternoon put forward the idea that AI could run a government far more efficiently than humans.
Once we had got over the image of the leader of the free world arguing with his machine alter ego (“Russia fixed the election”, “No, it didn’t”, “Yes it did”), the argument in the article was quite interesting.
Most activities that run a government are repetitive. Much of this activity can be done by machines that are relatively cheap.
Of course, AI is not just AI. There is a spectrum of relatively cheap applications that can take over the manual, repetitive stuff. And at the other end there is the relatively expensive (high investment) applications that will improve diagnoses and parts of the Health Service that we are only now beginning to imagine.
The thesis of several emerging articles and papers is not ‘man vs machine’ but ‘man plus machine’.
While this is a step towards sanity when it comes to the whole AI/automation/machine learning issue, there are many questions still remaining.
Governments have commissioned large, well-known research houses to look into the main ones. Which sectors will actually benefit from automation and artificial intelligence? Will the use of AI actually create more jobs? What is the role of government in regulating all of this?
We could add some more. For example, if AI runs a government more efficiently (and presumably it runs it more, and more, and more efficiently) what do the millions of government workers do then? And if, as seems to be the case, the answer from the emerging articles is “do better jobs that machines cannot do”, what jobs would those be?
Care, perhaps? Looking after elderly or disabled people is something (so we currently believe) that machines simply could not do. So, healthcare could undergo a(nother) massive revolution. Every person who needs care might be able to have care. Carers might be able to take diagnosis machines with them when they visit. The funding may be there in the future, because machines do not need salaries and national insurance.
Whatever the questions, and whether – as we suspect – the rise of AI will create more jobs, just as the industrial revolution and the IT revolution did before it, the next few years will prove fascinating.
Not least because some of the range of AI applications will definitely make us laugh.