Automation: Figuring It Out is Just Like Doing a Puzzle

When you're wondering if something can be automated or not, a way of looking at the situation is like a puzzle.

The picture on the box is your task fully automated, every piece has come together and fits perfectly. It's up to you how difficult the puzzle is and how many pieces are in the box.

With a puzzle, the pieces come in the box, all jumbled up and ready
for you to sort. With an automatable task, those pieces aren't quite pre-packed for you and you have to find them yourself.

To find those pieces, it's best to look at the requirements of your picture:

Flip that to your problem, and consider questions such as;

Understanding those key requirements can then set you on the right path for researching how to solve each one.

An important thing to remember in the search for these puzzle pieces is that you will not be the first human being in the history of computers to have this problem. With enough research, you'll find that person just like you who has kindly noted how they fixed their version of the same problem. Much like a grandparent who is a master puzzle completer and can show you a thing or two.

Everybody has a strategy for completing a puzzle; I was always told to start with the edge pieces first and then work your way inwards.

You can apply similar strategies to automating a problem. Do you dive right into the centre and build up small parts and hope they all connect, or do you try starting at the edges first and filling in from the outside.

Before I open the puzzle box, I've learnt to take stock of the picture. I'll think, is there a bit of the puzzle I've not done before? If so, that's what I'm going to do first, because if I can't do that bit, I'm not going to be able to complete the puzzle and who wants to leave a puzzle with a giant hole in the middle?!

If it's a similar picture to what I've done before, I remember the techniques and the strategies; I start from the outside edges and work my way inwards to complete the picture.

I've been inspired somewhat by the Polyani Paradox -

- and the idea that there are some tasks humans instinctively do that they just can't really explain.

This piece has been a way of trying to tease out and better understand something that has the potential to fall under the unexplainable category when I don't think it should.

Automation and code can seem like a magic trick, but once you scratch just a little under the surface, you suddenly realise your computer is full of things you can use to your advantage.

Regardless of what the movies say, computers are built to help us, and with a little bit of digging around and knowing ways to approach a problem, you can get them helping you as well.