A good analogy is to imagine filling a large lake. Every year you put in some water.

The first year you put in just 1 drop of water.

For every subsequent year you double the amount that you put in last time…

e.g.

year 2 will be 2 drops

year 3 will be 4 drops

year 4 will be 8 drops

…etc…

It will take a long time to fill the lake like that…but eventually you will fill it full of water! (Ignore evaporation etc…)

Now imagine that you finally reach the year that you fill the lake.

Last year it would have been only half full. You probably wouldn’t think of it being much of a lake then, but now it is full and the lake can now have fish, ducks and water sports!

Just two years ago, only one quarter of the lake was full. Definitely not a lake, more like a big puddle…

The bulk of the work came in just that last year. Approximately half to be precise.

The amount of water that is added each year is approximately as much water as has been added in all the years that came before it.

Imagine the first 5 years:

1, 2, 4, 8, 16 drops each year

1, 3, 7, 15, 31 cumulative drops in the lake

So each year, you are making as much progress towards filling the lake as you did from adding up all the years since you started.

An observation has been made, which is called “Moore’s Law” that states that roughly computing power doubles every 2 years. i.e. Exponential growth that we discussed.

This means the same as before, the progress in computing speed in the next two years, will be the same as all the progress that has come before, since computers began…

This principle is (the major reason) why things like smart phones have grown in use so dramatically in recent years. Computing power has got fast enough, small enough, and cheap enough that suddenly a huge number of people on the planet can own something that is very useful for everyday tasks.

Imagine a future (some way off) where computers are so quick to learn, together with being cheap enough, that Artificial Intelligence is EVERYWHERE. i.e. most people have self driving cars, robot cleaners and autonomous kitchens.

This future will happen extremely quickly in the sense of history. There will be a point where it is possible for these things to be available to everyone. Two years before this point, the tech will only be half as good, and 4 years before just a quarter fast enough and probably not good enough for the general public to adopt it.

The future of tech is of course significantly more complicated and intricate than this; however the principles of exponential growth are still valid, making this quite a scary concept.

Watch out, robots will take over quickly!