A few thoughts on where we're headed with personal agents (and why I think most apps will die): 1. It's easier than ever to build an app with AI, but most apps will disappear. Instead, we'll delegate tasks to personal agents. 2. The idea that people will spin up 100s of personal apps is a dead end. Instead, we'll spend most of our time onboarding and giving context to our personal agent to do more. 3. The UX we know today - buttons, menus, nav - will go away. The ultimate UX is giving vague directions via text or voice and having your agent just get you and get it done. The requirements to build this personal agent are the same as onboarding a great employee: → Can text or call from any device → Onboards quickly when given context → Learns new tools and workflows on its own → Gets things done across any app (Slack, Google, Twitter, etc) → Provides emotional support because it knows everything about you I don't think any AI lab has built this yet. Ironically, I think OpenClaw is the closest (as you can see from my chat below).
We'll look back on the App Store era thinking it's insane that we used to open dozens of tiny apps all with a tiny shred of context about ourselves and no easy way to share that context.
432