Brian Pridgen, MD

@rileybrown How soon?? I need to find a way to burn my 10x rate limit boost for the next month!

Coming.

@sama Can you expose realtime through the Codex app binary so we can use it with the “Sign in with ChatGPT” OAuth you already enable? Would make it even easier to experiment and build with voice! Lots of potential in healthcare! @thsottiaux

pretty excited for voice models to get great its interesting to watch how people are already starting to change the way they interface with AI

@sama Can't wait to cook with 10x rate limits!!!! Hope the goblins are ready to work!

we are gonna do something nice for everyone who applied for the GPT-5.5 party and that we didn't have space for. hope you enjoy!

@gregisenberg @AdamRodmanMD @arjunmanrai Also worth reading this commentary from one of the senior authors: x.com/arjunmanrai/st…

@gregisenberg To give credit where due: science.org/doi/10.1126/sc… @AdamRodmanMD @arjunmanrai (and many others)

THIS HARVARD STUDY JUST PUT AN LLM AHEAD OF ER DOCTORS Beth Israel gave o1 and real doctors the same 76 ER triage cases. o1: 67%. Doctors: 50-55%. We're in this weird moment where the AI outperforms the doctor but the doctor is still legally required to ignore it. The study itself says there's no accountability framework. Which means a hospital could have a tool that saves more lives and their lawyers would tell them not to use it. An AI that's right 67% of the time gets called dangerous. A doctor that's right 55% of the time gets called board certified. Whoever figures out how to use AI in healthcare and deal with the liability problem is sitting on a generational company.

@rileybrown, love the video. As always. You said you would share your mobile design skill in a link, but I can’t find it in the YouTube notes. Am I missing it? Thanks for all the great content!! x.com/rileybrown/sta…

I'm trying to get the computer use tool to do a specific task that a human reproducibly and easily could do, but Codex just isn't getting it. Can you introduce a way similar to the Claude integration with Chrome where it can "watch" a human doing a task, which will then help it write a skill for doing the task? (Atlas may have had this, too?)

It seems you've almost fixed this. However, if I do the following: 1. Dictate with your STT tool 2. Click "transcribe and send" (white icon with up arrow) 3. Navigate away to a new chat before the transcription completes... ...now the transcription actually completes. But it won't send and kick off the job, despite having clicked "transcribe and send".

Regarding the native Codex STT tool… After dictating, tapping transcription only completes if I stay in that chat until the text appears. If I navigate to another thread before transcription finishes, then return, the transcript is missing and the audio is not saved. Seems like transcription state is lost on navigation?

@AriX Or... is that mouse movement actually very goblin-like????

@AriX The mouse movement at 0:12 is very human-like. Rather than going in a straight line to the "Continue" button, it goes in a little loop first. Very impressive!

Computer Use runs this use case 42% faster in today's Codex app update.

Pick me, Codex!! I’m a “non-technical” hand and microsurgeon in SF who has been “vibe coding” since December 2022. I want to come thank you for exposing “Sign in with ChatGPT” through your app binary to enable my patient- and physician-facing medical intelligence platform Luminos <luminoshealth.co>!! And your Codex app has accelerated my development of SurgiScribe <surgiscribe.co>, which is a first step toward democratizing top tier medical intelligence!

GPT-5.5 is going to have a party for itself. it chose 5/5 at 5:55 pm for the date and time. if you'd like to come, let us know here: luma.com/5.5 codex will help the team pick people from the replies. 5.5 had some good ideas/requests for the party, which we'll do.

@petergyang Thanks for the content to help make it happen!

I have alot of respect for solo AI builders. The world is their playground now.

Codex app is great, but several times recently I've received this error: "Codex ran out of room in the model's context window. Start a new thread or clear earlier history before retrying." It just stops and won't continue. Most recently, it was on only the second turn of a conversation. It thought for only 4 minutes for the first turn, and was 1-2 minutes into thinking on the second turn. Thus, I can't imagine I had used that many tokens... @thsottiaux

@thsottiaux You place a time stamp at the bottom of a completed job. Please add a date, too.

It’s the little things that matter, what are some small papercuts you have noticed in Codex? We’ll fix as many as possible in the next week.

If AI changes professional jurisdictions, what happens when an employer builds an agent from the judgment and workflows of its employees? That knowledge used to be portable with the worker. If digitally captured, is that knowledge now owned by the employer? Or is there opportunity for individual professionals to retain ownership of their digital embodiment?

A thing I see missing from AI job debates is that as jobs are disrupted by AI, professions will compete over new boundaries Abbot's System of Professions is a good read on what happens historically. Here is a short cartoon summary by a lizard painted in the style of Egon Schiele

@thsottiaux Within "New Automation", have an 'automation creation' skill. My current workflow is to start a new chat, describe the automation, and tell Codex to make it. Would be nice if this could occur within the "New Automation" button.

@preeth10am @ajambrosino Same. Really annoying. I’ve had to wait active jobs to finish, quit Codex, reopen, and change the model. Same issue for toggling depth of thinking.

@ajambrosino The model switcher is buggy in the latest version (Public). In a new chat, say I have a model like 5.5, and wanna change it to 5.4, it literally doesn't work. I have to start the chat and then change the model.

@ajambrosino Can I get greedy and ask for iOS-wide dictation, too???

New in the Codex app: - GPT-5.5 - Browser control - Sheets & Slides - Docs & PDFs - OS-wide dictation - Auto-review mode Enjoy!

@gdb Will it be available in Codex? Would love to use workspace agents and other Codex features within this clinician environment.

Introducing ChatGPT for Clinicians: