Thinking Toys #4 — Falsifiability

What’s something that you’re worried about or that is giving you anxiety? It may be a specific event that you fear may come to pass — losing your job! Alternatively, it can be a general sense of discontent or fear about some state of the world — Trump is President! This feeling is uncomfortable but it may also be useful. Some part of you is trying to raise a concern — something is wrong here and we need to pay attention!


But what? how do we figure it out, and how do we resolve it rather than just suppress it? One way is to seek falsifiability — what would it take to prove yourself wrong? Try to construct an easily testable claim related to the worry or anxiety that feels predictive of the thing you’re worried about. Construct different statements and ask your gut: if I see evidence X, Y, and Z, will that persuade? The mere discovery of such a falsifiable claim can cause anxiety to dissolve. Failing that, now you have something testable for which you can seek evidence. Of course, constructing these statements will often leave some part of your anxiety still there. We can’t always quite capture the whole thing with a statement or two. This is likely unavoidable but often a trade off worth making, anyway.


I had some anxiety over starting this newsletter. How would I know if it was good or useful? It’s a vague, hard thing to measure and I expect mixed signals. With some effort, I was able to construct a falsifiable claim that my gut was comfortable with. If I could organically get at least 100 subscribers within a month, then it would feel “good enough”. By finding that statement, I was able to dissolve most of the related anxiety.


I find two other good uses for seeking falsifiability: exposing crony beliefs and improving our models of ourselves.


Kevin Simler wrote a great essay on the concept of “crony beliefs”. We hold these for their social value — sustaining relationships and giving us status. This is in contrast to “merit beliefs” which seek correspondence with events in shared reality — “objective” predictive power. Canonical examples of crony beliefs are eternalistic in nature: religion, politics, and ideas like the limitless power of science. Crony beliefs are not “bad” — they are often useful — but it can be handy to know where ours are hiding. One of the most glaring signs is a resistance to seek falsifiability in some domain. If you are defensive about making claims that could disprove something, it may be a crony belief. Consider: how high is your bar for changing your mind about your current political stances?


Improving our models of ourselves is another way seeking falsifiability can help us. This improves our ability to calibrate expectations. In Building a Second BrainTiago suggests defining projects by a set of associated SMART goals — even (especially!) for open-ended creative work. We often don’t have a clear idea of where an interest is heading and especially fear to constrain it. However, by making testable predictions about where you may end up, you give yourself way points to orient by. These way points can help you calibrate your ability to predict yourself! You get concrete feedback on your ability to know what you’re capable of. Consider a newfound love for oil painting. If you predicted that you would paint 3 canvases within 3 months and end up with only 1, you learn a little something about yourself free of charge.


Our minds are so drawn to falsifiability — seeing whether something is true or false — because it works. We survive by projecting and reifying boundaries onto a nebulous world. With repeated prediction and feedback, we are able to improve our models of how our actions relate to our ability to survive and thrive. A gnawing sense of anxiety is often a sign that we have failed to make something falsifiable enough for our lizard brain.


I’m writing one of these every week: sign-up for the Thinking Toys newsletter

Also published on Medium.

Leave a Reply