I'm mourning peak Twitter era, where the network was a useful way to keep in pulse with ideas, loose thoughts about where I wanted to go, and keep people updated about what I was thinking about.

I've been working towards a critical pragmatics over the past several years, and I've landed on several constructive principles, rather than polemics, forged by fire, that work for my idiosyncratic toolset:

  • Causal inference is a contested lingua franca for how people use data to make, or justify, decisions

  • Sometimes it's enough to have decision validity, but unfortunately we have a much better grasp on predictive validity instead.

  • Touching grass is good. A good model can touch grass without getting wet from the morning dew

  • Program and policy design is one bureaucratic-shaped lever for improving social welfare. There are others. 

There has been much progress lately in the past several years, despite institutional and environmental headwinds that threaten to snuff out the light of a candle. I think these headwinds clarify what matters most vs. what can be easily co-opted for regulatory capture and dumped when it starts being inconvenient. I believe that we will win.

About me - by listing some formative encounters:

  • Working with data from a nonprofit, then going to volunteer with some local nonprofits doing similar things occasionally

  • Backwards induction

  • Training as a legal observer 

  • Influence functions

  • Bonnans and Shapiro, perturbation analysis of optimization programs

  • Learning about partial identification from Francesca on Zoom before everything was Zoom

  • Printing on riso

  • going to fatml 2017 on a lark

  • media studies

  • some causal inference reading group hosted by buzzfeed in nyc summer

  • arcades project

  • managing a student design agency with designers far more skilled than I. Ju-jitsu with client needs

  • the good old days doing grad convex opt psets in a shared studio in 185 nassau st

  • fieldwork for bookmaking about place with danielle aubert