Think Thrice

If you think you can't do something, prove it

I've been reflecting on the future recently. It seems so open to me right now and the more I reflect, the more ambitious I feel. The problem is, after the initial high, I'm pulled down to earth by the concerns of actually taking action towards that future.

Things like:

For me and my circumstances, I feel like my "real" answer to these questions is yes. "Real" as in: I'm not looking forward to handling these things, but at the end of the day, I could deal with it.

But - I only actually feel this way on my best days, and that means the chance I'll naturally decide to take action is slim. After thinking about how to get around this, I've found a question that raises my confidence, even on my worst days:

"Would my past self find me astounding?"

What I mean by astounding is a person who is physically possible for you to become (i.e. it doesn't violate the laws of physics), but who you didn't think you could actually become.

An example of this might be becoming the President of your country - it's physically possible, but (probably) doesn't feel actually possible. When the probability of something is low enough, we tend to just round it to zero. Then we're astounded when it happens because we forgot that it was even possible.

I think this question helps my confidence because it reminds me that my past self would find the fact that I'm a software developer and runner pretty astounding.1 My past self wouldn't have believed it was possible to become who I am today.

Why? Part of it was a misunderstanding of the skills that programming requires (it isn't math).2 But I think the main reason was a lack of prior evidence - I had zero background in programming and had always found endurance exercise to feel terrible. My current self didn't fit into the narrative of my past self, so they couldn't see a path from A to B.

If belief requires evidence but you don't have any, what can you do? You do experiments and prove it.3 If you think you can't program, learn Python for a month and show me that you can't implement Tic-Tac-Toe, no matter how hard you try. If you think you can't run a marathon, sign up, train for a few months, and show me that you actually collapse in the street on race day.4 Unless you do that, you don't have evidence, and neither I nor you should believe you.

Asking whether my past self would find me astounding helps me tap into the evidence I've gathered running those experiments myself. Similarly, you've probably proven a lot to yourself, but just don't remember it well. If you find it difficult to be confident about an uncertain future, ask yourself whether your past self would you astounding.

If you find your answer is yes, isn't it fair to be confident about taking actions now that seem unreasonable, that don't fit your narrative, but that might lead to an astounding future?

After all, if you've been astounded once, why can't you be astounded again?

Get the next post?


  1. If you know me, I wonder if you'd feel the same.

  2. I now think there's lots of skills in research which translate well to programming, and that most researchers would make excellent developers. More on this in a separate article, but the main factor is that much of programming is debugging, and if you know how to set up a good experiment, you basically know how to debug programs well.

  3. Of course, by prove it, I mean you take the null hypothesis that you can't program to be true and ask yourself after learning Python and implementing Tic-Tac-Toe in a month whether that evidence is compatible with the null hypothesis, and reject it if it isn't.

  4. They need to be fair experiments. If you try and pull a couch-to-marathon tomorrow, yes, you will actually collapse in the street. But would you trust that as conclusive evidence? I'm sure the majority of people (myself included) who have run marathons couldn't have achieved that either. Doesn't seem like a fair experiment to me.