Claude Code 31: Hadn't the Satisfaction Always Been in the Discovering, Not the Discoveries?

  • Author/Source: Scott Cunningham (Baylor University), via Substack ("Causal Inference")

  • Original: https://causalinf.substack.com/p/claude-code-31-hadnt-the-satisfaction

  • Key Ideas

  • Society organized its scientific sector around workers whose intrinsic love of the research process -- the "discovering," not the discoveries -- made them willing to endure extreme uncertainty and repeated rejection.
  • That trait functioned as a "clasp" connecting human workers to the scientific machine; it was the reason the system worked despite offering terrible expected outcomes for most individuals.
  • If AI agents gain comparative advantage in research, the trait of loving the process becomes a "legacy appendage" -- like the palmaris longus tendon, still present but no longer functional.
  • Researchers who derived deep meaning from the work will carry that need without any work to satisfy it; unlike displaced workers who can relocate, there is no alternative destination for this trait.

  • Science was never for the scientists -- the satisfaction was just the wages needed to get the 6 or 7 discoveries that changed the world.

  • Summary

Cunningham writes a literary lament about what happens to researchers whose deepest motivation is the intrinsic satisfaction of the research process itself. He describes the peer review system in deliberately naive, almost anthropological terms -- people writing "words, numbers, and pictures," submitting them to a person who owns a "book," waiting for four strangers to decide if the words are correct, enduring years of rejection. The defamiliarization is deliberate: it exposes how strange and fragile the system is, held together by the single fact that the workers love the work.

He then frames the scientific enterprise as a machine designed to harness a specific human trait -- the willingness to turn a crank endlessly because the turning itself is the reward. This trait cannot be taught; it is not intelligence but romance. Society built universities, tenure systems, and peer review around people with this trait because the expected value of the scientific lottery was positive only at civilization scale (6 or 7 transformative discoveries), and only workers who loved the process would tolerate the individual-level odds.

The essay's core argument is that when AI agents can do research faster, more accurately, and without needing satisfaction, the trait that sustained the system becomes unnecessary. The love of discovery does not disappear -- it becomes a vestigial organ, an ache without a name. Unlike geographic displacement, there is no relocation possible: the researchers have nowhere to go because the work itself has moved beyond them.

  • Relevance to Economics Research

This essay engages with fundamental questions about comparative advantage applied to cognitive labor, the economics of intrinsic motivation, and institutional design for science. It provides a framework for thinking about what is lost -- not just in productivity terms but in human terms -- when AI displaces the process-oriented satisfaction that has historically sustained the research enterprise.