@gabrielchua Oh. So Codex is going to totally replace the browser chat bot. Codex is just now synonymous with Chat GPT in the very literal sense.
Wasnt obvious until this screenshot. Why use regular Chat GPT anymore?
Daily run that pulls context from email, slack, linear, fireflies, Claude code sessions, git commits, calendar, etc and provides:
Novel insights that I might be missing
Prioritized tasks
Mental framing on work streams and their progression (or lack thereof)
Commits to a daily memory markdown file and references prior day outputs to stay relevant
There are suddenly a lot of options for running AI on a schedule: ChatGPT started with scheduled tasks, which then got hastily wrapped into Pulse (?), and the Codex desktop app makes an attempt too. Anthropic recently launched scheduled tasks in Claude Code Desktop alongside the Claude Code /loop feature, while OpenClaw added fuel to the movement with a built-in cron scheduler.
At the root of all of these features is the cron job, a basic concept for engineers but unfamiliar to many others. Cron jobs let code run on a schedule, and they can be set up locally on your machine or orchestrated in a cloud environment. Both options require technical know-how that is likely too high a hurdle for mass adoption. That is what makes the labs’ focus on scheduled tasks so novel: they are consumerizing the cron job, while simplifying what it takes to build an AI Agent in the process.
With these new tools, users can schedule tasks (crons) in a user interface they already know, sidestepping some of the less navigable parts of agentic coding (deployment). Even bigger, these are no longer just scripts. In this context, scheduled tasks are powered by the same agent harness that powers Claude Code and Codex. Simply executing a task run can fit the definition of an AI agent.
While exciting, a big reason I wrote this post is that I think much of this is still messy and too complex for most users. That said, the capability is arriving quickly, and starting to think through which workflows become unlocked by a scheduled agent is likely to be a worthwhile investment.
Something I noticed alongside the launch of GPT 5.4: OpenAI has seemingly moved their computer use tool to general availability.
Big improvements over preview and hopefully broader access across usage tiers. Excited to see this capability keep maturing.
https://t.co/atgV4XiMn5
@trq212 I’m sure you guys are tracking this but fast mode doesn’t appear to work in the IDE extension.
Typing /fast kicks you into the IDE terminal vs. the extension Claude Code window
@trq212 Built a context engineering app for working with LLMs
App grabs a screenshot any time the screen changes --> converts images into markdown for documenting workflows
Also super copy/pastes all screenshots into a chat if needed
@thsottiaux Awesome work, loved reading through the doc. Do you have a sense for when Codex will be able to handle more complex, MCP-laden setups?
You noted this as a source of performance degradation and wondering if we simply just need to wait for gpt-6-codex for example
Current state of the moves to become an operating system by the likes of @OpenAI and @AnthropicAI - the game is just context management.
Your product needs to get context in the hands of the major models. That’s it. Otherwise you will no longer have a product
@michelelwang Hey Michele! Sent you a DM related to the role - 8 years of finance experience here across multiple disciplines including 3yrs on the buyside. Currently leading AI & Product at a startup. Would love to chat.
Can confirm Codex CLI is actually pretty powerful; was skeptical given it feels like Open AI is trying to catch up to Claude Code but there is something clean about GPT-5 High.
The overall interaction is cleaner and more focused. Highly recommend for challenging workflows
Working on a project this weekend? Try Codex CLI with GPT-5, included in your ChatGPT plan!
❯ npm i -g @openai/codex
❯ codex
Pro tip! Use GPT-5 with high reasoning via this shortcut:
❯ codex -m gpt-5 -c model_reasoning_effort="high"
Or type /model to pick interactively:
@sama A “screen record” feature - model can observe my screen and make live recommendations, similar to video chat as it exists today except viewing on screen activities. Will greatly improve rate of information transfer between user and model
I listened to the six-hour podcast titled "#Luck" so you don't have to.
The podcast chronicles the football career of #Colts quarterback #AndrewLuck, who suddenly retired at age 29, leading to the greatest what-ifs in #NFL & #sports history.
Here are the 10 lessons I learned👇
'How do I transition into climate tech'? For newcomers from traditional tech, finding your corner of the climate universe can feel overwhelming. If you’re beginning your journey, a few of us made a "climate tech talent+capital mind map" to help — 🧵 1/10