Podcast
Using AI to share knowledge with the world
Expertise has always been rationed by geography, language and cost. Jay Pandya on what changes when it stops being.
What if the world's best advice was just one question away, no matter where you lived or what language you spoke?
Expert knowledge has never been evenly distributed, and the reasons have very little to do with the knowledge itself. Advice is limited by who is physically near you, what language they work in, and what an hour of their time costs. Those are distribution constraints, and distribution constraints are exactly what technology tends to dissolve.
What AI actually removes
- Language, advice no longer has to be given in the language it was created in.
- Geography, proximity to an expert stops being the qualifying condition.
- Availability, knowledge is no longer rationed by an expert's calendar.
- Cost, the marginal cost of the second answer approaches zero.
Intent, not search
Jay makes a distinction that matters for anyone building in this space: search returns documents, while an intelligent system responds to intent. Someone asking a question rarely wants a reading list. They want the specific thing that applies to their situation, which is what an expert would have given them.
Where the humans stay
The episode is not an argument that expertise is obsolete. It is an argument that expertise should scale, that the person who spent thirty years learning something should be able to help more than the few hundred people who can physically reach them.
Key takeaways
- Expertise is rationed by distribution, not by supply of knowledge.
- Language, geography, availability and cost are the four barriers AI removes.
- Intent-driven systems answer the question; search returns documents.
- The goal is scaling expertise, not replacing the expert.