An attitude indicator from an aircraft cockpit, with the horizon slightly tilted relative to the fixed orange aircraft symbol

We live in an era of apparent polarization: of winners or losers, of good guys or bad guys. You’re either with us or against us.

I don’t presume, nor above all do I have the tools, to say whether this is a bad thing or a good thing (or, here too, something in between). I think it’s all so complicated that dismissing it as “all shit” or “all awesome” is a gamble, or at the very least yet another act of cognitive outsourcing to the logic of polarization.

Don’t get me wrong: my fallible human brain, aware of what’s happening around us, is on constant alert, and it’s the first to feed me thoughts like: “Oh yeah, bro, it’s all shit, why the fuck are you wasting time thinking… RUN!”. But as someone interested, more or less directly, in how the human brain works and in metacognition, I know this isn’t enough. So, in a maximalist, post-realist era, I want to venture down the old, winding road of realism, with all its heaviness and its scant appeal to today’s tastes.

By realism I mean something simple, at least on paper: understanding what’s happening, studying its causes, and only then taking concrete action. Without deciding from the start that it’ll be a disaster, because it’s not a given that it’s all shit: some things might actually turn out well.

An example close to home is my work. I was convinced that artificial intelligence would completely screw me over. For now, though, it’s mostly taking away the part I don’t care about: writing tons of repetitive code. In exchange, it gives me back time to think about biology again, which is the reason I do this job in the first place.

This doesn’t mean it’s all awesome: that would just be the same generalization with the opposite sign. The line between what the machine does and what’s left to me will shift, and I don’t know where. That’s why I keep my seatbelt fastened, but I don’t just sit and watch. I actually read the code it suggests, instead of settling for the fact that it runs. And I invest in the part AI is still worst at: deciding what to measure, how to design a study, figuring out whether a result makes biological sense.

Maybe that’s all realism is: being afraid, checking whether the fear was justified, and then doing something.