Tyler Cowen muses about algorithmic curation in Twitter, and he’s wrong, but he’s wrong for an interesting reason. Even more interestingly, while I was musing on it, that interesting reason was recently stated curiously but quite nicely by James Pethokoukisregarding a totally different topic – climate change:
“…humans have a poor record of understanding risk in complex systems, full of interdependencies, feedback loops, and nonlinear responses. Perhaps humility and caution and consideration are warranted.”
What’s curious about Pethokoukis’ take is that he frames it in the headline of his post as “[]a] conservative way to think about climate change.” And depending on your definition of conservatism, this may be so. But it is decidedly not an economist‘s way of thinking about anything.
The economic way of thinking is deeply rooted in margins. This is, in many instances, very useful, and has a lot of explanatory power in dealing with a diverse set of market outcomes. Tyler Cowen is very much an economist rooted in this way of thinking (I mean, look at the title of his blog), and thus applies this general way of thinking to Twitter. How does this change impact behavior at the margins? How does that partial equilibrium translate to general equilibirum?
The problem, however, that this mode of thinking runs into is that it’s very poor at understanding ecosystems, and therefore poor at detecting the boundaries between changes that do and don’t disturb that ecosystem. Ecosystems, as Pethokoukis writes about, you know, The Ecosystem, is characterized by ‘complexity, interdependency, feedback loops, and nonlinear functional forms.’ But ecosystemic characteristics apply in a lot of intra-human activity as well.
Twitter is a good example of this. As it stands right now, there are many marginal changes one could make in Twitter’s operation – a few more promoted tweets here, a change to profile structure there. But some changes risk affecting the Twitter ecosystem, which risks destabilizing what makes Twitter work on a more fundamental level. Involuntarily excising tweets from feeds, even on a very small scale, risks destroying broad trust in some of the core principles and benefits – sponteneity, discovery, democracy, meritocracy, platform autonomy – that have made Twitter so useful and appealing to so many. A seemingly marginal change could actually collapse the ecoystem.
The lesson here is generalizeable – where economists and that mode of understanding and problem solving tend to go wrong are those places where ecosystemic factors and risks go undetected by a mindset not trained to see them. The result is one where marginal changes – or, more insidiously, a collection of marginal changes – causes an ecosystem to malfunction. That doesn’t mean chaos; a new ecosystem fills the vaccum. But that ecosystem may be substantially different than the previous one, and not always, or even often, for the better.
Advertisements