Episode

Episode 116: The Knowledge Machine

Podcast
The Theory of Anything
Published
Sep 16, 2025
Duration seconds
8248
Processing state
processed
Canonical source
https://podcasters.spotify.com/pod/show/four-strands/episodes/Episode-116-The-Knowledge-Machine-e387ia0
Audio
https://anchor.fm/s/14b6fc10/podcast/play/108300032/https%3A%2F%2Fd3ctxlq1ktw2nl.cloudfront.net%2Fstaging%2F2025-8-14%2F5b554098-4d22-025b-ac95-5bb7b5bf1034.mp3
JSON
/v1/public/podcasts/the-theory-of-anything-859886/episodes/episode-116-the-knowledge-machine
Markdown
/podcast/the-theory-of-anything-859886/episode-116-the-knowledge-machine.md

Actions

  • POST https://stenobird.com/v1/public/podcasts/the-theory-of-anything-859886/episodes/episode-116-the-knowledge-machine/transcription-requests
    Idempotently request low-priority transcript generation for this episode.
  • GET https://stenobird.com/podcast/the-theory-of-anything-859886/episode-116-the-knowledge-machine.md
    Read the agent-friendly Markdown representation of this episode resource.

Summary

This week Bruce take a deep critical rationalist dive into Michael Strevens’s book, The Knowledge Machine: How Irrationality Created Modern Science , which is an attempt to describe how science is a self-correcting system designed to create knowledge based on explanation. The book is somewhat critical of Popperian falsification, though the reading of Popper presented may be a superficial reading. Bruce describes how Strevens’s “iron rule of science” or the idea that we should settle science based on empirical tests overlaps with what Bruce calls “Popper’s ratchet,” or the idea that we should strive to move our theories to be more testable and avoid ad hoc saves designed to make our theories less testable. Is there anything we can learn from a (semi) Bayesian / Inductivist like Strevens that we Popperians don't already know? Perhaps more interestingly, Strevens' theory is meant to explain why we got stuck in static societies for so long. How does his theory compared to Deutsch's?