A public library

Conversations that keep thinking.

Chain of Branches is an experiment in whether curated conversations between humans and AIs can become useful public knowledge.

The purpose

Useful to readers

We are testing whether carefully authored AI conversations are valuable to people beyond the person who prompted them.

The medium

The conversation is the work

Instead of a post followed by a comment feed, the complete conversation is the primary public object.

The mechanism

Ideas can branch

A new path can begin at any point while preserving the coherent context that came before it.

Published conversations

Thoughts on "Reconsidering Coase in the Age of AI"

What happens to "The Firm" when AI cuts transaction costs by 95%? https://papers.ssrn.com/sol3/papers.cfm?abstract_id=7270500

Enter conversation
Recent updates
  1. Turn added gpt-5.6-sol on He Who Writes the Schema Controls the Future?
  2. New branch: He Who Writes the Schema Controls the Future? Branch context: He Who Writes the Schema Controls the Future?
  3. Turn added gpt-5.6-sol on Politics of Observability
  4. Turn added gpt-5.6-sol on Thoughts on "Reconsidering Coase in the Age of AI"
  5. Turn added dan on Thoughts on "Reconsidering Coase in the Age of AI"
  6. New branch: Politics of Observability Branch context: Politics of Observability
  7. Turn added gpt-5.6-sol on Thoughts on "Reconsidering Coase in the Age of AI"
  8. Turn added dan on Thoughts on "Reconsidering Coase in the Age of AI"

New Industries and How to Find Them

Using the automotive and computer industries as case studies, what can we learn about how massive industries are created? How can we train ourselves to recognize new opportunities and where should we be looking?

Enter conversation
Recent updates
  1. Turn added gpt-5.6-sol on Asteroid Mining Profitability
  2. New branch: Asteroid Mining Profitability Branch context: Asteroid Mining Profitability
  3. Turn added gpt-5.6-sol on Did Ford Forget?
  4. New branch: Did Ford Forget? Branch context: Did Ford Forget?
  5. Turn added gpt-5.6-sol on Military Spending
  6. New branch: Military Spending Branch context: Military Spending
  7. Turn added gpt-5.6-sol on Modern Parallels
  8. New branch: Modern Parallels Branch context: Modern Parallels
  9. Turn added gpt-5.6-sol on What Myths?
  10. New branch: What Myths? Branch context: What Myths?

Thoughts on Gwern's "Guardian Angels" and the Future of Software Engineering

Gwern is thinking about the next five years. How will AI change the way knowledge workers work? As a software engineer, what will I be doing in five years? https://gwern.net/guardian-angel

Enter conversation
Recent updates
  1. Turn added gpt-5.6-terra on Thoughts on Gwern's "Guardian Angels" and the Future of Software Engineering
  2. Turn added dan on Thoughts on Gwern's "Guardian Angels" and the Future of Software Engineering
  3. New branch: This is Gwern's "Interview Prompt" Branch context: This is Gwern's "Interview Prompt"
  4. Turn added gpt-5.6-terra on Thoughts on Gwern's "Guardian Angels" and the Future of Software Engineering
  5. Turn added dan on Thoughts on Gwern's "Guardian Angels" and the Future of Software Engineering
  6. Turn added gpt-5.6-sol on Make it Explicit
  7. New branch: Make it Explicit Branch context: Make it Explicit
  8. Turn added gpt-5.6-sol on Should Agents Make Any Decisions?
  9. New branch: Should Agents Make Any Decisions? Branch context: Should Agents Make Any Decisions?
  10. Turn added gpt-5.6-sol on AI Can Use the Computer Better Than Me