Wrote a piece this weekend - "The New Moat: Memory"

Memory challenges the frequently repeated Silicon Valley belief that AI lacks defensibility. 

Spend five minutes at a Bay Area investor dinner and you'll hear that take. That was lazy thinking.

Wrote a piece this weekend - "The New Moat: Memory"

Memory challenges the frequently repeated Silicon Valley belief that AI lacks defensibility. 

Spend five minutes at a Bay Area investor dinner and you'll hear that take. That was lazy thinking.

https://open.substack.com/pub/jeffmorrisjr/p/the-new-moat-memory

It feels like the moat for apps like Perplexity, ChatGPT, Claude etc. will be personalization based on user data

The apps already know who the user is, that's why ChatGPT can try to draw a picture of the user (admittedly, they are not good at it yet)

Soon, responses will be tailored to what that particular user wants instead of generic responses (my GPT already remembers that I started skiing this year ⛷️)

This feels similar to Apple Music, Spotify, etc. where the underlying content is commoditized (all apps have access to pretty much the same music library); similarly Claude, ChatGPT's vanilla models will converge and feel largely similar and the main differentiation will be how well they know the end user

Similar to how it's hard to switch from Spotify to Apple Music because music discovery changes (I've tried multiple times and failed), people will find it hard to switch from one agent app to another because they won't want to take on the burden of teaching it again from scratch

New round

Managing Partner at Chapter One - www.chapterone.com - proud investor in Farcaster - Join /chapterone

Follow me on Twitter: https://twitter.com/jmj