There are many ingredients for creativity. One is to have a lot of knowledge saved up, so it can combined in novel ways. One is to actually think about things for five minutes. One is to not flinch away from problems when you find them.

But the most important is to have a problem to solve. Or if you're Feynman:

You have to keep a dozen of your favourite problems constantly present in your mind, although by and large they will lay in a dormant state.


To have a problem in the back of your mind constantly is somewhat taxing. You need to find a problem which isn't emotionally draining. For this reason the problems need to be important. They don't have to be the most important thing in the world, but they do have to be important enough to you to be somewhat self-motivating. If your day-to-day life involves so many interactions with doorbells that you can think about doorbell design all day, then that's enough.

In some ways, I have it easy on this front, which makes it hard for me to give advice. My domain is biological chemistry, so I can do a good job just by pointing myself at the closest ageing-related issue. Ageing is very salient to me, and as I get older I will either succeed or it will become painfully more salient.


The problems need to be of the correct size. I don't hold the problem "cure ageing" in my head all day. That wouldn't work, because anything could in theory be related to "cure ageing" but there's no clear way for my brain to make the right connection. Instead you have to find a correct balance. The problem "replace lost stem cells in human tissue" is about the right level. There are lots of possible answers, but also the line between ideas I come across and that is not too many steps long.

I suspect that the right level is different for different people. This makes my advice somewhat unhelpful. Still, just knowing there is a correct level might be helpful. Perhaps a good idea is to try and keep a few different problems of different levels around, at least until you find the right one.

How to actually find the problem

Hopefully the two issues above have already clued a few readers into my method.

First off, think of something you want, or want to know. Just wanting to "discover things" is not enough. You have to work one level down to achieve that goal. It can even be silly or useless. As long as your emotional brain will let you focus on it.

If you're lucky, that problem will be really big, too big, even. This is good. It means you can just spend a bit more time splitting it into sub-problems. Try and be honest. Imagine that a given sub-problem was solved. Would it really contribute to the bigger issue? If you could solve one sub-problem magically, which would it be?

Once you have picked some then you have to actually do the thing, which is the hard part. As my current research involves a lot of (relatively) boring steps of making and purifying protein, it's easy for me to spend time mulling over the ideas. This might not be the case if you have a different day job or a different brain.

This is all you can really do. You cannot force yourself to have important ideas, only put your brain in the right place to have them. This is a technique which can be practised and improved on. So go and do it.


New Comment

New to LessWrong?