one of the most beautiful things about OpenAI is that every employee really has a voice. i wanted to capture what it feels like to work here, what our mission means to me, and why you should join us. so i made this video with Codex, shared it with the team, and they felt it was worth producing and sharing with the world. this is our mission. and it’s why i’m here.

codex tip: teach code review what your team keeps repeating. add a “Code Review Rules” section to AGENTS.md. codex will apply your repo-specific guidance and cite it on pull requests.

5.6 is the most usable model ever made not saying it's gonna magically solve all your problems, it's just very easy to get good results from with low effort at this point if you're still struggling to get good code out of the machine it might be a you thing

CHATGPT JUST BECAME THE EVERYTHING APP Codex just merged into ChatGPT, making it an AI "super app" for knowledge work. I broke down the newly released GPT 5.6 Sol, ChatGPT Work, and what Codex merging into ChatGPT means for builders (plus recommended settings and prompt guides) TIMESTAMPS 0:00 ChatGPT Just Became The Everything App 1:28 Codex vs ChatGPT Work vs Chat 5:45 GPT 5.6: Sol, Terra, Luna + when to use which 6:57 Model vs Effort vs Speed 8:29 GPT 5.6 Recommended Settings 9:12 How to Prompt GPT 5.6 12:16 Top New Features 12:24 ChatGPT Sites 14:23 Subagents in Codex 15:56 Appshots (Computer Use) 17:09 Built-in Browser (Browser Use) 18:27 Try GPT 5.6 for yourself

My dear UI developers, ML practitioners, and fans of programming language: Many months & billions of tokens later, I’m proud to present to you the first step in our long, long collective journey to turn vibe coding onto proof engineering, starting with: making user interfaces verifiable. Introducing: Freerange, a zero-API tool that automatically deduces your code’s numerical ranges. By doing so, Freerange is able to prove that e.g.: - your TS layouts obey your specified sizing - that they’re free of NaNs and Infinity - that your array indices stay within bounds All of that, done statically. No browser, no running code, droppable into any codebase, and for the ML folks: RL-friendly

I believe Freerange might be one of the biggest open-source projects written with Fable & Sol. And technically barely feasible with current AI At its peak it had 50k lines of code. Was a long journey iterating & reducing it down to 18k today. The pivots were brutal If I started it a few months earlier with worse models I'd have lost my mind. Thank you OpenAI for the internal model access!

I spent most of my weekend on ChatGPT Sites. Here is why I think it’s the biggest unlock yet in the mainstream consumer being able to build with AI: - Aesthetics. Almost everything I have tried to build on Claude has a similar format, similar font, similar overall (recognizable) feel. IMO, it’s great at putting things into a template - less good at creative choices. I got very different and more opinionated styles on Sites even with very little prompting. - Hosting. If you’re a non-technical person who has tried to deploy a vibe-coded website, it’s pretty much a nightmare. Figuring out what you need to do to buy a site, handling security, setting up auth if needed, etc are all big obstacles. This is even trickier with API keys that need to be stored properly. Publishing on Sites is literally one click. - Sophistication / interactivity. I was shocked by what 5.6 could one-shot within Sites. It built fairly complex real time game mechanics (ex swipe on cards, move around in Google Street View on a real map) with extremely simple text prompts. It would also make good choices towards making those interactive experiences work - ex. a “Bonus” page in a game should pop up with a different color and UI - Identity / Auth. Sites utilize “Sign in with ChatGPT” - this is the first step in making “bring your own token budget” or “bring your own context” a reality. And yet, as the host the deepest analytics I could see was aggregate number of users and what they engaged with (no identifying info on anyone). Right now, you can set a Site to private or public - I foresee more to come with org-level controls, password protected access, etc.