OpenAI launched an Apps SDK on Monday (see here). This effectively means that developers can now build apps inside ChatGPT:
This was an obvious play by OpenAI and it was only a matter of time until more dynamic experiences made it into chat. After all its very difficult to do some of the most profitable internet activities (ie ecommerece, gaming etc) with text only. The writing was on the wall here ever since MCP-UI came out and Shopify started incorporating it earlier this year:
There are some really cool technical details in how this all works. We heavily invested in MCP-UI, an extension of the MCP protocol to allow exposing widgets that a host can choose to display. The product cards you see in the video are exposed through MCP-UI and then securely… pic.twitter.com/FNsrMiB1Uq
— tobi lutke (@tobi) August 5, 2025There is a whole generation that uses ChatGPT as their de-facto browser. While there is no direct mention of it anywhere, the benefits to OpenAI here are massive. The flywheel is obvious and not unlike any other platform shifts we've seen in the past. However, there is one big difference here and that is memory. LLM memory is going to be the biggest battleground for LLM apps because it is what will keep users from leaving. As capabilities across all chat apps are converging more and more, it is essential to have a piece that makes it incredibly difficult for users to migrate from one app to the other. A memory layer that knows the hotel that you'd like to stay in, the type of food that you'd like to order and what books you like to read before bed makes for a killer user experience, and the more of those activities you do from ChatGPT the more valuable it becomes. This allows OpenAI to own a very crucial piece of the consumer internet. Up until recently, the only way to get access to this type of information was the browser and your device.
The timing here is perhaps notable. There is no real advantage in being a first mover here. As the MCP-UI experience shows, there are a lot of hard problems to figure out and there is a big chance that some of this blows up in OpenAI's face. Google can add this to Gemini within a quarter and the reality is that its a question of when not if. Launching early doesn't necessarily trigger any network effects either. Perhaps one of the most obvious reasons to get going here is that OpenAI's devices efforts are very real and the sooner they can stand up an ecosystem, the sooner they can have a standalone device that allows their users to do all the activities they do on a phone today.
There are more platform plays coming to ChatGPT, most notably Sam Altman brought up "Sign In With ChatGPT" in May:
another gem:
Sam mentions federated auth a bunch of times. "Sign in with ChatGPT" is going to be a big step change in defining OpenAI as a platform. Mainly because that is how OpenAI is going to get context on users from outside apps
See full video here:… pic.twitter.com/9uXjTPUYQv
All of this reinforces the same idea: AI chat apps are the new platform and the next land-grab will be over the user's activity in all other apps, especially ones that involve purchasing decisions.
.png)

