Artificial intelligence – “Practical parenting for your AI engine”
23 August 2016
The world has become very excited about the idea of artificial intelligence. We dream of the panacea of the “machine that learns everything”… the “machine that never forgets”… the “machine we can always rely on”.
But how realistic is this?
I am often musing over the real world. It is full of ideals, great ideas… but sadly many fail to materialise. The practical realities of life always seem to have an impact - and usually not in a good way.
As I pondered the practicalities of artificial intelligence, I started to reflect on the realities of intelligence. I didn’t have to look far before I started to worry.
My children have grown up to become worldly-wise people, but as I look at their journey through life, I recognise that they haven’t always been fed the best information to make the right decisions.
I imagined my newly built artificial intelligence engine. It's beautifully written and eager to learn. I just need to press the “button” and very soon it can guide me through the complexities of my business world. The stress of complex decision making will soon be gone.
But now the dilemma begins. Where do I find the facts to feed it? What if it gets “in with” the wrong crowd? What if it decides that “evil” is a fun and exciting thing?
Suddenly I'm visualising my children again. Should I have stopped them from playing outside? Should I have selected their friends for them? Have I inadvertently skewed their views of the world and made them narrow minded?
I'm sure that I exercised a few of these controls over their early lives, but now they are free and roaming their world. They're now free agents, who must apply the essence of the controls I hopefully passed on to them and choose carefully what they learn and who they learn from.
Maybe this is a good end point, but I don’t have time for my new artificial intelligence engine to learn so dramatically from its mistakes, or should I call this “experience”.
So, where has my musing taken me?
Well, I'm seriously worried about how I educate my learning engine. I'm worried about how I control what I expose it to. And I'm wondering about how I check that what I feed it, is in fact right.
Suddenly my artificial intelligence engine has become my child and I need to be a good parent. I need to love and protect it from bad things. And as every parent knows, this will not be an easy task.
I feel the need to write a new book… “Practical parenting for your AI engine”.
I'm sure you'll agree… this is something for us all to think about.