OpenAI DevDay 2025
Yesterday was the OpenAI DevDay 2025. The opening keynote was live streamed on YouTube.
At the 03:43 mark, I was able to confirm that Sam Altman was again wearing the Adidas Ultraboost DNA × LEGO Colors Shoes. I do wonder how many pairs of these shoes he has stashed away!
I also enjoyed the snake countdown animation that the live stream kicked off with. I’m a sucker for ASCII and Unicode art and appreciated the creative use of characters such as § (Unicode U+00A7).
The moment celebrating the developers in the room who had passed “tokens processed” thresholds was interesting. Given the obvious link to electricity consumption, I do wonder if some of them squirmed in their seats. Assuming a fixed energy efficiency, the goal is clearly to process fewer tokens in exchange for a similar amount of value.
There was a live Codex demo that gave off strong (and likely unintentional) surveillance vibes. As someone who’s dabbled in controlling DMX devices in the past, the light show was a little underwhelming, but they avoided the kind of demo disaster we were treated to by Zuck recently at Meta Connect 2025. The demo gods also delivered a Ghostty update prompt that popped up at the start of the same demo.
Key Takeaways
- Apps SDK - Third parties can now get their apps into ChatGPT. It’s launching with a limited set of pilot partners, but developers can start building with the Apps SDK preview.
- AgentKit - A set of tools to build, deploy, and optimise agents.
- New models released in the API - Sora 2 (still haven’t got my hands on an invite for this AI slop machine), GPT-5 Pro, gpt-realtime-mini, gpt-image-1-mini.
Apps SDK feels a bit like Apple’s App Store moment in 2008 to me. Not a surprising development given the hiring of a CEO of Applications earlier in the year. It’s built atop the Model Context Protocol (MCP), so it should already feel familiar to developers. Your app’s custom UI is served up to ChatGPT as an embedded resource and then displayed inline (in an iframe) in the chat interface. There are design guidelines, developer guidelines, and details about a review process will follow later in the year. I can imagine companies flapping around trying to figure out how to have their apps surfaced in ChatGPT through the discovery mechanism. I’ll be interested to ship something and step through the process just for the experience.
Central to AgentKit is Agent Builder, a node-based, visual workflow tool for building agents. Not quite a “feel the AGI” moment and doesn’t feel like the future, but it’ll be appreciated by builders who have less experience or interest in the code. The demo in the live stream made it look like a cinch to drop in tools, guardrails (e.g. a PII check), and vector search. You can deploy to the OpenAI platform, but also export the code and run the agent on your own infrastructure. ChatKit provides you with an embeddable widget that you can deploy in your own app and hook up the workflow you created with the Agent Builder. It tipped me off about the existence of their Guardrails framework, and I plan to give that a closer look.
Closing Thought
ChatGPT is on the trajectory to be the new Google with their platform push. I’m still in the opening chapters of Empire of AI, and it’s interesting to watch in real-time how far OpenAI appear to have strayed from their founding ethos (the democratisation of AI research). They have to raise money to reach AGI somehow though! In Altman’s words, it’ll be “interesting to see what you build with it”.