Skip to main content
 

Thee formal difficulty of modelling an unpredictable system on the fly.

2 min read

... talk of “evidence-based policy”, which we’re all supposed to be wildly enthusiastic about at the moment, makes my blood run cold, because at the back of it is the assumption that it is possible to gather reliable evidence from human society in real time. That’s a *really* big ask, and no amount of big data or surveillance is going to get around the formal difficulty of modelling an unpredictable system on the fly. In many cases, having more data makes the problem worse. So there’s always going to be this tension between politics, which is an historically contingent activity, and science, which as much as possible treats the world and everything in it as if it were a linear, symmetrical system. So, then, look at this the other way: can we ever actually *separate* science and politics? No. There’s no way citizens can be expected to simply hand money over to some scientific priesthood and expect *no* return for their investment. But the bitter truth is that any experimental activity is a hugely wasteful, and demanding rigid outcomes and takeaways is a sure-fire way to deaden and demoralise your science base.

Simon Ings