We are moving towards a shift from "hand" to "head".
From "hand", in leveraging machines to improve performance by using mechanical arms, to "head", in leveraging machines to improve performance by using artificial intelligence and making better decisions.
But expecting computers to help us make better decisions implies to be aware about what « better » exactly means in numbers, and what are the effects of our typical decisions - from conscious to unconscious ones - in the light of those numbers.
If it might be easier to do so for simple problems, when it comes to addressing complex issues like climate change, health, poverty and inequality, knowing about the full spectrum of the decisions we are making starts to become less obvious.
And usually somebody will say, oh let's use artificial intelligence!
And that’s where we fail.
Before coding anything for any artificial intelligence to come to life while expecting to use it to make the world a better place, we should first and foremost answer three simple questions: (1) What are the metrics that define « better »? (2) What are the decisions that we are currently making so far in this purpose? (3) What are the uncertainties we have to face?
Having to explain to a computer how to solve a problem is a good exercise for explaining to ourselves how to think about the problem. As Albert Einstein said: if I had only one hour to save the world, I would spend fifty-five minutes defining the problem, and only five minutes finding the solution.