Knowledge Skyscraper

The knowledge skyscraper is CF’s image for how learning accumulates: each floor holds ideas, and higher floors are built on lower ones. The central claim is structural. Your building can rise at most about ten floors above its first floor with a major error, because that error cannot bear the weight of much built on top of it. Worse, the damage from a major error grows with distance: a flaw on floor five barely dents floor six but guarantees floor fifteen is broken or never stands at all.

This drives CF’s emphasis on prerequisites and quality standards. The more you intend to build on an idea, the higher its quality must be; an idea that merely “seems right” enough to use once may still collapse under ten layers. Because of error propagation, getting stuck rarely means the current floor is wrong — the cause is usually a subtle major error several floors below. The remedy is foundational review: drop down many floors at once to find the earliest error fast, rather than descending one floor at a time.

A distinctive CF point is that major and minor errors are nearly bimodal — “medium” errors that cap you at, say, 100 floors are rare, echoing the jump to universality. So there is a clean cutoff: any large error limits you to a handful more layers, while only-small errors let you build thousands. Minor errors, CF says, are not really errors at all, since they do not block the goal. Against Critical Rationalist objections (foundations don’t matter; knowledge is a web), CF replies that there are no ultimate foundations, yet which sub-parts you build from still matters greatly — especially through integration when learning.