What If We Stop Seeing Apps as Apps?

I was planning a road trip with Claude recently. Not asking it to search for destinations — actually planning, working through real criteria: drives under three hours between stops, quiet campsites, a mix of places I’d been and places I hadn’t, with some room for the kind of detours that make a trip worth remembering.

At some point, knowing Claude had rolled out some visualization capabilities, I asked if it could help me see the trip laid out. I wasn’t sure what to expect.

What I got was a clean, purpose-built webpage: stops listed in order, distances between them, a simple visual flow from point to point. Modern enough to feel intentional, functional enough to actually be useful. I pushed Claude further — could I see it on a map? The page reassembled itself around a zoomable map with pins for each stop. As I refined the plan, the pins moved. A trip came together.

I never asked for an app. I didn’t install anything, subscribe to anything, or open anything from a home screen. Claude just built the right thing for the moment — and when the planning was done, that thing could have simply ceased to exist.

That’s not an app. That’s software doing exactly what the situation called for, nothing more, and disappearing when the job was done.

The paradigm we haven’t questioned

Apps have been the organizing principle of computing for as long as most of us can remember. You have a task; you go to the app for that task. It’s a useful mental model: clean, compartmentalized, predictable.

But it comes with a cost we rarely name: cognitive load. Not just the effort of using an app, but the effort of figuring out which app. Does this belong in Notion or in Notes? Is this a spreadsheet or a doc? Should I start a project in Asana or just manage it in Slack? We make these micro-decisions constantly, and they add up to a kind of low-grade friction that we have to tolerate.

The app paradigm also boxes us in. Tools like Notion and Obsidian have pushed against this — they’re flexible, they blur categories, they try to meet you where you are. But they’re still apps. They still have paradigms, still impose structures, still require you to work within their logic even when your actual need doesn’t map neatly onto it.

The boxes are just bigger now.

Software that reads the room

What my road trip experience pointed toward is something different: software that doesn’t ask you to fit into its model, because it builds its model around you.

I didn’t need to decide whether my trip planning belonged in a spreadsheet, a map app, a notes tool, or a travel planner. The right shape for the task just emerged. Tabular when I needed to compare stops. Spatial when I needed to see the route. Adjustable as my thinking changed. No imports, no exports, no switching between tools and losing context in the gaps.

This is where things get interesting, and where I think we’re headed faster than most people expect.

Imagine the icons on your phone’s home screen. Right now, each one represents a fixed application: a thing that exists permanently, waiting for you to need it. Now imagine those icons representing your current needs instead. Not apps, but contexts.

Road Trip 2027 is on your home screen. Tap it on a Tuesday night and it picks up where you left off — comparing campsites, checking availability, maybe flagging a weather window. Tap it the morning you leave and it shifts: your itinerary, turn-by-turn, the reservation confirmation for tonight’s stop. Tap it mid-drive and it knows you’re moving — it surfaces what’s relevant right now, not what was relevant last Tuesday.

The same icon. Completely different software, assembled for the moment you’re in.

And when the trip’s over? Maybe Road Trip 2027 becomes an interactive map — photos pinned to the places you stopped, a record of where you went and what you found. Or maybe the icon just disappears. A thing that exists precisely as long as it is useful, no longer, no less.

Why this matters now

This isn’t science fiction. The road trip experience I described happened. The pieces are assembling themselves in real time — not as a product someone is building toward, but as a side effect of AI getting genuinely good at understanding context and assembling things that fit the moment.

The shift, when it comes, won’t feel like a revolution. It’ll feel like relief. The friction we’ve normalized — the wrong tool, the missing feature, the app that almost does what you need — starts to dissolve. Software stops being a compromise you manage and becomes something that works for you, specifically, in the moment you need it — and knows when that moment is over.

The app had a good run. But the better version of this might not look like an app at all.

If you want to see what software-as-a-verb looks like in practice, I followed this up with a more grounded piece — a road trip that turned into an unexpected live demonstration of the idea. Software as Verb