thought
Search will be dead by 2030
1 min read
aisearchproduct
Classic search assumes users can translate intent into keywords
The next interface assumes the system already understands intent from context and history
What changes first:
- query boxes become fallback UX
- ranking systems become orchestration systems
- links become evidence, not destination
The winners are not "better search engines"
The winners are systems that combine:
- persistent user context
- retrieval quality under uncertainty
- trustworthy synthesis
Search is not dead as infrastructure
Search is dead as a primary interaction model