Concept · part of Building bridges that last
Universities, scientific journals, common-law courts and open-source software communities have something in common: none was built in a product cycle. Each took generations to build, and has taken generations to drift.
The book closes its prescription on that timescale. The attention economy it describes built itself in two decades. Institutions able to hold against it will need at least that long, and will have to be funded, staffed and politically protected the whole way. The book calls this the work of generations: rebuilding what societies already had, plus what the new media now require.
Part of the reason is people. The versatile experts such institutions run on take longer to train than ordinary specialists, and training, once it lapses, is slow to regrow. The rest is design: money from outside the attention market and an aim of outlasting the attention economy.
The book is frank about the cost of saying so. Telling readers the job takes decades invites the reply that there is little point starting now, and its only answer is to start building anyway. It sees one possible accelerator: AI tools run under an institution’s own control, extending the reach of the few experts there are.
Read the full argument
- Chapter 12: Infrastructure for Integration, section “Where I land”
- Chapter 9: The Integration Problem, section “The five operational problems”
How it connects
- Examples: Common-law courts as a curation institution
- See also: Differential decay rates (preservation vs. training), Funding decoupling from attention markets, LLM as capability extender (inside substrate custody), Path dependence and lock-in, Survival, not victory (integration institutions), Versatile expertise
- Where the book discusses it: Chapter 12: Infrastructure for Integration
← Survival, not victory (integration institutions) · Building bridges that last (12 of 12)
Explore all the ideas · Read the book in order · Open the interactive map