Originally published in Svenska Dagbladet by Björn Jeffery, October 9, 2025
With competitors closing in fast, OpenAI is borrowing its new strategy from Steve Jobs. Now apps are coming to the chatbots. Could it be as big a success as the iPhone?
The video clip has become legendary. Steve Ballmer, then CEO of Microsoft, is asked what he thinks of Apple’s new product, the iPhone, in 2007. The blunt and self-confident Ballmer laughs — almost scoffs — and says it is the world’s most expensive phone, and not something that will attract business customers because it lacks a keyboard.
Ballmer’s prediction aged poorly. But it was neither the price nor the keyboard that proved decisive for the iPhone’s success. It was the apps.
When OpenAI invited developers to its developer day earlier this week, this insight was crystal clear. The company’s new strategy follows the iPhone — and could produce similar results.
If you analyse why the major tech companies ended up where they are today, you can see a couple of common denominators. One of them is trying to become a platform for other companies to develop products and services on.
Look at Ballmer’s Microsoft, whose operating system Windows underpinned everything in the PC era. Facebook’s games in the 2010s — remember Farmville? Today there is Nvidia and its development platform CUDA.
Perhaps the most well known of all is Apple’s App Store. But Steve Jobs’s phone did not launch with an app store — that came later. Jobs wanted to ensure the software experience on the phone was good enough, and so Apple made all the first apps themselves. But even Jobs was eventually bested by the platform strategy. When they opened the doors for developers to build their own apps for the iPhone in 2008, things really took off.
Sam Altman, CEO of OpenAI, knows this, of course. Not least because he himself was on Apple’s stage there and then — in 2008 — presenting his then-startup Loopt. At that point he was an app developer. Now he wants to become the platform itself.
Because OpenAI’s lead is diminishing rapidly. ChatGPT is still market leader, but depending on what type of AI use you need, other models are close behind — or sometimes better. Well-funded companies like Elon Musk’s xAI and Anthropic are impressive. The threat from China with DeepSeek and similar models persists. And the biggest of them all — Google — has made great strides with Gemini.
Competing with Google is both expensive and difficult. The search giant has billions from its ordinary business to plough into AI development. On top of that, it has several advantages when it comes to marketing and distribution. Think about how many people visit Google every day to search for something. Today, many of them are met by AI services as a complement to the ordinary search results.
OpenAI has none of these advantages. So they must do something different — compete in a different way. And now they are inviting other companies to appear and function inside ChatGPT.
It becomes like an app inside a chatbot. Want to book a flight? Then you can chat directly with the booking site to find something that suits you. Need a playlist for dinner, you can describe it in ChatGPT, and an app from Spotify creates one for you with the right feel. ChatGPT becomes the interface through which you reach other services.
For now, it is small-scale. But so were the iPhone’s apps at the beginning. There are probably many companies that would prefer to have users on their own services instead. But over time, it is users who ultimately decide how they want things. Today they want their own apps on their phones. Tomorrow it is possible they will want apps inside AI services instead.
The launch makes OpenAI step further ahead on an increasingly competitive AI market. Expect similar launches from competitors in the near future. Because no player can afford to miss out if the AI market goes the same way as much other technological development.
OpenAI is first out. But they will not be the last. Everyone wants to become the platform for AI — with the definite article.
