80% of Apps Are Going to Die

Think about the last time you opened an app just to find one piece of information. You scrolled a list page. Maybe applied some filters. Clicked into a record. Read a number. Closed the app.

That whole flow exists because of screens. Humans need a visual way to sort, filter, and browse data. So someone built a UI for it.

But an agent doesn't need any of that.

Peter Steinberger put it well on the Lex Fridman podcast. He talked about how your personal agent already knows where you are, how you slept, what your day looks like. "Why do you need MyFitnessPal when the agent already knows where I am?" He said it might kill off 80% of apps. I think he's right.

A list page is a solution to a human limitation. You can't hold a thousand rows in your head, so you get pagination, sorting, search bars. But if you can just ask "which invoices are overdue?" and get the answer in two seconds, the list page becomes dead weight.

Same goes for dashboards. Most dashboards exist because someone needed a way to glance at metrics without running queries. An agent can just tell you what changed and why. No login, no navigation, no clicking through tabs.

The businesses that figure this out first are going to move faster than everyone else. Not because they built a prettier dashboard. Because they skipped the dashboard entirely and automated the workflow end to end.

I'm building this at FELTLAB right now. Email triage that classifies and routes hundreds of messages without a human touching an inbox. Process automation that replaces five different tools with one agent that just handles it. Reporting that shows up in your chat instead of a BI tool you forgot to check.

The future isn't better apps. It's fewer apps. The ones that survive will be the ones that do something an agent genuinely can't. The rest become slow APIs that agents call on your behalf until they're not needed at all.

If you're still building features for screens, it might be time to think about what happens when the screen goes away.