ATProto, by design, refuses to be the place where things happen. Decisions live in apps. Value derives from execution, and execution will demand focus. This is application primacy — and it's why I don't expect an ATmospheric "everything app": numergent.com/2026-05/ATProto-Application-P…

what is our “moat” on atproto even? what keeps people using a specific atproto app vs a fully compatible other app?

some thoughts on atproto as a post-app paradigm

also, I kinda feel like calling the things that consume atproto data "apps" does a disservice to understanding what they are / how they work / what their potential is. But that is a partially-formed hot take I can't fully defend yet.

There are two theories-of-change emerging: displace vs. integrate They are by no means mutually exclusive, but much more effort has gone into spinning up brand new atproto apps with the goal to displace existing apps than has gone into embedding the protocol into every nook and cranny of the web.

the point of atproto is to make locked-open apps. apps which are constrained such that they cannot be e.g. bought by billionaires and ruined. yes, doing non-atproto things "enables app developers to do more". yes, non-atproto is "more intuitive" and "it sucks" when you hit atproto's limits.

"The fundamental asymmetry is that apps can be general-purpose or purpose-specific, but the *thing you are doing* is never general, it is always a concrete thing."

This is partly why I think atproto needs a more convincing vision about being a post-app protocol. A future where atproto is a bunch of apps and each app governs their own lexicon diminishes a lot of its potential in the long-term.