The narratives described in the article are metanarratives in the critical theory sense. Lyotard defined postmodernism as “incredulity towards metanarratives.” So welcome to 1979!

Investors literally can’t stop talking about “pattern matching,” because it’s comforting to think that the past can predict the future if you’re clever enough. But the best investors are pattern breakers, not matchers.

>For example, “You can’t beat the market, dollar-cost average into index funds!” is the kind of viral story that will earn you nods of approval from millennials and upvotes on Reddit.

>Pattern matching, on the other hand, can drive you to make moves like an advanced chess engine. Moves that others find alien and incomprehensible.

Sure. Index funds is a viral story which has a lot of rigorous data behind it. As for the advanced chess engine of intuition... ever heard of Fooled By Randomness?

> Why does legibility bias persist across society, even when we are aware of it? One answer is: it soothes people.

This is false on so many levels. First of all illegibility also soothes. Many people prefer the “illegibility” of natural landscapes in the great outdoors over cold corporate offices with straight lines.

Second illegible systems are less predictable and harder to analyze. For example why do we have to build airliners in orderly ways using clean surfaces and simple shapes? Why not messy geometry and chaotic inconsistency? Because the math then becomes to hard to model and predict. Non linear chaotic systems are hard to predict using our quantitative modelling techniques.

In fact our modern engineering is built on top of quantitative modeling using applied mathematics and the only things we can really model are ordered and simple systems. Engineering is equivalent to prediction and it is our best form of prediction in the sense that you can get on a metal tube called an airplane and get hurled across the ocean and feel relatively safe because our predictive powers are so high that we know you’re incredibly safe.

Order is preferred and it is not bias. It’s because order has intrinsic advantages with analysis.

Performance is a different story. It may be the case that chaotic systems can lead to better results but we simply don’t have the tools to predict or understand how these systems work.

We are however hitting a new era Of engineering where we use blackbox neural nets to analyze illegible systems. It has downsides in the sense we can never truly fully understand the analysis nor can we fully trust it, but such an analysis is now possible to an extent. It also suffers from the fact that we need massive amounts of trial and error attempts to even get the data that enables building these models.