The HBS hosts chat with Justin Joque about how we might get Thomas Bayes’ robot boot off our necks. 

Why does Netflix ask you to pick what movies you like when you first sign on in order to recommend other movies and shows to you? How does Google know what search results are most relevant? Why does it seem as if every tech company wants to collect as much data as they can get from you? It turns out that all of this is because of a shift in the theoretical and mathematical approach to probability. 

Bayesian statistics, the primary model used by machine learning systems, currently dominates almost everything about our lives: investing, sales at stores, political predictions, and, increasingly, what we think we know about the world. How did the “Bayesian revolution” come about? And how did come to dominate? And, perhaps more importantly, is this the best mathematical/statistical model available to us? Or is there another, more “revolutionary,” mathematics out there?

This week we are joined by Justin Joque, visualization librarian at University of Michigan who writes at the intersection of philosophy and technology. He is the author Deconstruction Machines: Writing in the Age of Cyberwar and, most recently, Revolutionary Mathematics: Artificial Intelligence, Statistics and the Logic of Capitalism.

In this episode, we discuss the following thinkers/texts/ideas/etc:

• Frequentist statistics
• Bayesian statistics
• Dan Rosiak
• Probability theory
• Bayesian statistics and machine learning
• Lorraine Daston, “How Probabilities Came to Be Objective and Subjective” (Historica Mathematica 21, 1994)
• Sir Ronald Aylmar Fisher
• Egon Pearson
• Jersey Neyman
•  Inductive behavior (or “inductive discipline”)
• Leonard Savage and “Dutch book” arguments
• Pascal’s wager
• Taylorism and Fordism (i.e., the automation of the production of things)
• on the Square (payment technology) scandal
• Karl Marx, “Estranged Labor” (from the Economic and Philosophic Manuscripts of 1844)
• technocapitalism
• Karl Marx on “commodities speaking amongst themselves” (in Capital 1, Chapter 1: “Commodities”)
• the “Sidney Poitier knotted tie”
• Social media’s personalized ads and how they work
• “efficiency” vs. “robustness” 
• Problems with predictive models like PredPol (predictive polcing)
• Matteo Pasquinelli, “On the origins of Marx’s general intellect” (Radical Philosophy 2.06, Winter 2019)

If you enjoy Hotel Bar Sessions podcast, please be sure to subscribe and submit a rating/review, follow us on Twitter @hotelbarpodcast, on Facebook, and subscribe to our YouTube channel!

You can also help keep this podcast going by supporting us financially at patreon.com/hotelbarsessions.