causal

KV Caches Do Not Scale as Learning Memory

AIs cannot learn continually by building up a growing KV cache across users, because that is not scalable and is not how humans learn; human learning compresses information into weights rather than storing parameters and activations separately.

causalpending

Speaker

Dwarkesh Patel

Evidence Quote

“AIs can't just keep building up a KV cache that grows in size as you keep learning from more and more users. That's just not scalable, and it's also not how humans learn.”

Source

Dwarkesh Patel Podcast - The next big breakthrough will be AIs learning on the job— Dwarkesh Patel Podcast
Created: 11/17/2026, 6:25:50 AM

Your notes

Loading notes...