The Collapse
The Chicxulub Impact was the moment the asteroid that wiped out the dinosaurs struck Earth.
Obviously, a lot changed about the world that day. It was a mass extinction event for vast swathes of animal species. Few survived - from Google AI overviews:
"birds (the only surviving dinosaurs), crocodiles, small mammals, amphibians, lizards, snakes, and some marine life. Their survival was aided by traits like small size, the ability to live in burrows or water, and varied diets"
I think we're about to go through an equivalent to the Chicxulub Impact for the mobile app ecosystem. It'll be what the iPhone did to hardware, and maybe worse.
The iPhone saw an entire kaleidoscope of hardware objects get collapsed inside a single device: cameras, MP3 players, video cameras, calculators, torches, compasses, and so on. All of these individual devices became largely obsolete, outside a few high-end and/or nostalgic niches. The iPhone was the all-encompassing Swiss army knife for the hardware era - it wiped out many species.
AI will see the same play out in the mobile app ecosystem. Many categories that once managed to survive as individual vertical apps will be consumed within wider platforms. The centralising impact of shared user context combined with lower costs to build new products and features makes it almost inevitable.
The most straightforward version of this reality is the major consumer AI platforms of today: ChatGPT, Gemini, Claude, Grok. All of these are looking to own the consumer, and that means eating more and more of the consumer's resources: time, attention and money. A vast array of dinosaur apps - from recipes to workout plans to restaurant recommendations - are about to meet their proverbial maker. If informative content is your core business, things are about to get tough.
The second way this could play out is the rise of mini-app ecosystems. We've seen the emergence of some of these - with Wabi, and even with Open AI's own ChatGPT Apps. In this scenario, the remaining use cases that are sufficiently verticalised to require a standalone product environment will survive - but will be published on platforms that enable a shared context layer that understands consumer context and applies it broadly across use cases. The winners of this era will not be the operating systems of the mobile app era. New winners - at both the platform and the app level - will need to be built.
Either way, the harsh reality is that to build anything standalone and valuable will require extraordinary craft and differentiation. The 'good enough' version now doesn't even need to be actively built by anyone at the BigTech platforms - the AI will do it automatically. As ever, Please Be Different and focus with maniacal obsession on how to build great things.