Discussion about this post

User's avatar
Michael Smith's avatar

This was really good and clear. I recently encountered Jay Azhang's article on reflexivity:

https://ethylacetate.substack.com/p/reflexivity

…and noticed that the concept finished letting something click into place about what I've been calling "subjective science":

https://morphenius.substack.com/p/subjective-science

Your description is incredibly clear. Really emphasizes what I think is the key practical element: you can't remove yourself from reflexive systems you're part of in order to understand them. That graphic is incredibly clear. (Did you make it?)

Minor nitpick: yes, "predictions" given within reflexive systems are usually either spells or ignorant. And politicians are professional spellcasters. But that doesn't mean they're trying to make WHAT THEY SAY come about. Often they're trying to create an effect based on THE FACT THAT THEY MADE SUCH A PREDICTION. Reflexivity isn't simple enough to be a battle of stated goals.

I'm reminded of how supposedly the CIA hyped up the impression of a government cover-up around aliens so that there'd be plenty of cover for military experiments. I don't know if that's true, but it's PLAUSIBLE, and it's the kind of reflexive spellcasting I'd expect to happen in advanced info warfare. "There are no aliens, these UFO nutjobs are just crazy" isn't trying to MAKE there be no alien abductions. It's much weirder than that. It's (maybe) actually trying to create an impression that they ARE happening and that there's some government cover-up. So you get something like two layers of illusion.

Likewise, when Elon says he expects SpaceX to be profitable or whatever, I don't think he's necessarily trying to make SpaceX profitable. He's trying to change how space travel works. Him guiding people to expect profits can maybe cause enough support to enact the change he actually wants to see. Even if it's never profitable.

Weirdly, the stuff about "Look at how bad Elon's prediction rate is" getting publicized can mess with his methods if it spreads enough. This is a case where people who aren't used to thinking reflexively can use objective methods to break reflexive things and then look around going "Aha! I just exposed that the emperor has no clothes!" Never mind that the emperor's new clothes WERE REAL, just not physical. In the story, the child broke the spell. Saying the clothes were never real begs the question: what then did the emperor think he was wearing? What did everyone else think he was wearing?

Expand full comment
Mi’sen's avatar

Very nice. I'm familiar with reflexitvity from some sociology theory I've mostly forgotten, and Soros' application to markets. I feel like this way of talking / thinking about it generalises it in a useful way, good clear examples, good explanatory power for social phenomena.

Funny also how this fits with something like convexity, where the downside of being 'wrong' when [making predictions] is limited, with huge upside potential if the social arbitrage works to re-allocate resources (information flows, attention, capital, etc). That said, I suppose there is a limit to how many social options you can write before you burn through all your social capital; at least in Elon's case he has enough power and enough huge wins where this seems less of a risk...he can afford to take even more narrative risk in that sense.

Good stuff, look forward to reading more of your writing on the topic!

Expand full comment
8 more comments...

No posts