surprises and how to spot them

a lawlike generalisation of sorts

I’ve always wanted to have a law named after myself, like Murphy and Sodd and those guys.  I tried to coin one in the fraud book (“Anything that’s growing too fast needs to be checked out in a way it hasn’t been checked out before”), but the editors told me to knock it off.  They’ve also had a few tries in various drafts of the current book (“Decisions Nobody Made”, coming next year), but I keep plugging on nonetheless.

At present, I’m most keen on this little aphorism, which expresses my opposition to those virtuous but boring souls who caution against trying to make forecasts about our complex and surprising world:

Davies’ Law: “If you don’t make predictions, you’ll never know what to be surprised by”.

(I even have a corollary; if you don’t make recommendations, you won’t know what to be disappointed by).

As with many such things, the law’s most useful if you reverse it.  If something seems surprising, ask yourself whether, if it had occurred to you to do so, you might have predicted it.  The multiple layers of counterfactuals pile up a bit there, but I think the concept is coherent.  As I’ve argued before, the majority of Black Swan events are not intrinsically unpredictable ones, they’re things which people could have predicted but didn’t.

In fact, the strategy of reversing the question can be applied multiple times.  When you see a business or organisation making an announcement that they are going to use Artificial Intelligence, the question to ask yourself is “ahead of time, would I have predicted that this would be one of the only entities in the economy that didn’t make use of something that looks like an incredibly generally applicable technology?”