3 minute read
The real pitch isn’t burning the app store down
Before we go all doom and gloom on the future of app stores let’s show some data. The “appless future” doesn’t mean Apple and Google cease to exist. In 2025 the App Store ecosystem moved over $1.4 trillion in developer billings and sales, and Apple plus Google together account for more than 90% of all app downloads. Nobody sets fire to a trillion-dollar machine because a chatbot says it can do better. Heck, how many times have you seen output from a chatbot that’s been frustrating and half-baked? So, if the prediction is that “all apps die” well that’s not happening.
The goal was a phone with zero icons, it was the outcome
You don’t love the grid of forty rectangles. You tolerate it. What you actually want is the seat booked, the bill paid, the amazon order, the form filed. Apps are the middlemen between you and the thing you wanted to get done. That’s the shift people are talking about, a move toward intent-based UIs instead of tap-and-swipe, where you say the goal and the system handles the booking and comparing for you. The app still runs the transaction. You just stop being the one clicking through it. Hmm, sounds like the need for an interaction interface or back-end for agentic agents. Just sayin’.
Agents do the riding, they don’t replace the train
The doomers don’t get this part. An agent may be the one booking your table when they tap into the restaurant’s system but there’s still the need for things to get paid. It still uses opentable for whatever system to get there. Its it easier for an Agent to tell the app to do something. Granted I’ve seen and read lately that some AI chatbots are actually “getting lazy”. This means not doing the hard lifting and making assumptions instead. Which is what people do and a great way to make mistakes. Reddits has developers complaining about this a lot lately. So, if all the agent needs to do is leverage the app then things still get done. So, apps survive as agent-readable backends while the agent gives you another interface. Apple and Google don’t lose that fight, they enable it. The store becomes plumbing plus a storefront, and they own both. Ninety percent of downloads don’t evaporate.
The money guys are building it as a feature, not a funeral
OpenAI is reportedly building an agent-first phone with Qualcomm and MediaTek, targeting mass production in 2028. Meanwhile Qualcomm’s CEO keeps saying agents become the next interface layer. But notice the smart version of this doesn’t happen on some custom gadget. It happens in the OS already in your pocket, the one wired straight into the store. The agent becomes a feature of the platform, and the platform is the thing that already won. Now if someone tries to release an “appless” phone good luck to them. I can smell the VC cash already burning on that hill.
The gadget graveyard proves the point, not the opposite
Humane and Rabbit failed because agents alone can’t do it all on some custom device, yet. Sure, maybe one day they can but for now agents do things because devs “hack” a way to do them. What’s simpler hacking every single possible interaction or making a common interface to chat and work with apps we already have? The phone is already in your pocket and not everyone wants to spend cash on yet another piece of tech to carry around. The winning move was always Apple and Google baking it in, which is precisely what’s happening.
Less clicking, same landlords
The realistic appless future isn’t anarchy. It’s the same app economy with a smarter front desk. You do fewer chores, the store keeps its cut, and the menus you never wanted quietly retire. That’s a good trade. Take it, but keep your expectations realistic.
Sources: Apple, Business of Apps, TNW, Qualcomm, OpenAI, Vocal Media, TechnoGen Solutions, Medium