/cdn.vox-cdn.com/uploads/chorus_asset/file/24390407/STK149_AI_02.jpg?w=696&resize=696,0&ssl=1 "ChatGPT can now handle reminders and tasks")
/cdn.vox-cdn.com/uploads/chorus_asset/file/24390407/STK149_AI_02.jpg?w=1200&resize=1200,0&ssl=1&description=ChatGPT+can+now+handle+reminders+and+tasks){kind=link}
OpenAI is launching a recent beta feature in ChatGPT called Tasks that allows users to schedule future activities and reminders.
This feature, which is rolling out to Plus, Team and Pro subscribers starting today, aims to make the chatbot something closer to a established digital assistant – like Google Assistant or Siri, but with the more advanced language capabilities of ChatGPT.
Tasks allow users to tell ChatGPT what they need and when they need to do it. Want a daily weather report at 7 a.m.? Passport expiration reminder? Or maybe just a joke that’s worth telling to children before bed? ChatGPT can now handle all this through scheduled one-time or recurring tasks.
To utilize this feature, subscribers must select “4o with scheduled tasks” in the ChatGPT model selector. From there, just type in what you want ChatGPT to do and when. The system can also proactively suggest tasks based on your conversations, although users must explicitly approve any suggestions before creating them. (Honestly, I feel like suggestions have the potential to create annoying mistake by chance).
All tasks can be managed directly in chat threads or through the recent Tasks section (available only on the web) in the profile menu, so you can easily modify or cancel any task you have set up. Once these tasks are completed, notifications will notify users across the web, desktop and mobile devices. There is also a limit of 10 busy tasks that can be run at one time.
OpenAI hasn’t specified when (or if) this feature might be made available to users for free, which suggests Tasks may remain a premium feature that will aid justify the cost of a ChatGPT subscription. The company offers monthly subscription tiers of $20 and $200.
) } const Containers = document.querySelectorAll(.w-full figure.w-full”); containers.forEach(container => { const root = ReactDOM.createRoot(container); root.render(
While scheduling capabilities are a common feature of digital assistants, this represents a change in ChatGPT’s functionality. Until now, artificial intelligence operated only in real time, responding to immediate requests rather than dealing with current tasks or planning for the future. The addition of Tasks suggests that OpenAI is expanding ChatGPT’s role beyond conversation into an area traditionally occupied by virtual assistants.
OpenAI’s task ambitions seem to go beyond elementary planning. Bloomberg announced that “The Operator,” an autonomous AI agent capable of independently controlling computers, is scheduled for release this month. Meanwhile, Reverse engineer Tibor Blaho found that OpenAI appears to be working on something codenamed “Caterpillar” that could integrate with Tasks and enable ChatGPT to search for specific information, analyze issues, summarize data, navigate web pages, and access documents – with users receiving notifications when the task is completed.
The development of “agentic” artificial intelligence in 2025 is not just a matter of technological progress – it is a matter of economics
As I wrote in October, the development of “agent” artificial intelligence in 2025 will not only result from technological progress – it will have economic reasons. These agent-like features are a strategic way to monetize high-priced AI infrastructure. While OpenAI’s decision to put this functionality behind the ChatGPT firewall was predictable, the real question remains: will it provide reliable results? The last time I got a demo of the OpenAI agent, it displayed faulty information. The coming months will show whether the team has resolved these fundamental reliability challenges.
I also think of this recent feature as a slightly more sophisticated script, but ultimately Tasks follow a elementary, routine set of instructions, much like a typical bot. The goal of many pioneering AI labs, such as OpenAI, is to develop these capabilities into something that can interact with environments, learn from feedback, and make decisions without constant human input.
However, questions remain about how reliable these scheduled tasks will be and what will happen if ChatGPT fails to provide time-sensitive information. OpenAI’s decision to launch Tasks in beta suggests that it is still working out these details and wants to gather real-world feedback before a wider rollout.
For now, if you are a paid ChatGPT user, you can start experimenting with Tasks by looking for the “4o with scheduled tasks” option in the model selector. Just remember that it’s still in beta – so maybe don’t rely on it for very significant meeting reminders just yet.