OpenAI’s ChatGPT provides scheduled duties characteristic in beta
OpenAI is launching a brand new beta characteristic in ChatGPT known as Duties that shall we customers time table long run movements and reminders.
The characteristic, which is rolling out to Plus, Crew, and Professional subscribers beginning lately, is an try to make the chatbot into one thing nearer to a conventional virtual assistant — assume Google Assistant or Siri however with ChatGPT’s extra complex language features.
Duties works by means of letting customers inform ChatGPT what they want and when they want it finished. Need a day by day climate record at 7AM? A reminder about your passport expiration? Or possibly only a knock-knock shaggy dog story to inform your youngsters ahead of bedtime? ChatGPT can now deal with all of that via scheduled one-time or ordinary duties.
To make use of the characteristic, subscribers want to make a choice “4o with scheduled duties” in ChatGPT’s style picker. From there, it’s so simple as typing out what you wish to have ChatGPT to do and when you wish to have it finished. The gadget too can proactively recommend duties according to your conversations, regardless that customers need to explicitly approve any tips ahead of they’re created. (In truth, I think just like the tips have the prospective to create annoying slop unintentionally).
All duties will also be controlled both without delay in chat threads or via a brand new Duties segment (to be had simplest by the use of the internet) within the profile menu, so it’s simple to switch or cancel any activity you’ve arrange. Upon crowning glory of those duties, notifications will alert customers on internet, desktop, and cellular. There’s additionally a prohibit of 10 energetic duties that may run concurrently.
OpenAI hasn’t specified when (or if) the characteristic may come to unfastened customers, suggesting Duties may stay a top class characteristic to assist justify ChatGPT’s subscription prices. The corporate has per month $20 and $200 subscription tiers.
Whilst scheduling features are a commonplace characteristic in virtual assistants, this marks a shift in ChatGPT’s capability. Till now, the AI has operated only in genuine time, responding to quick requests quite than dealing with ongoing duties or long run making plans. The addition of Duties suggests OpenAI is increasing ChatGPT’s position past dialog into territory historically held by means of digital assistants.
OpenAI’s ambitions for Duties seem to stretch past easy scheduling, too. Bloomberg reported that “Operator,” an self reliant AI agent in a position to independently controlling computer systems, is slated for liberate this month. In the meantime, reverse engineer Tibor Blaho found that OpenAI seems to be operating on one thing codenamed “Caterpillar” that might combine with Duties and make allowance ChatGPT to seek for explicit knowledge, analyze issues, summarize knowledge, navigate internet sites, and get entry to paperwork — with customers receiving notifications upon activity crowning glory.
The upward push of “agentic” AI in 2025 isn’t with regards to technological development — it’s about economics
As I previously wrote back in October, the upward push of “agentic” AI in 2025 isn’t with regards to technological development — it’s about economics. Those agent-like options constitute a strategic option to monetize pricey AI infrastructure. Whilst OpenAI’s choice to position this capability in the back of ChatGPT’s paywall was once predictable, the actual query stays: will it ship dependable effects? The final time I were given an OpenAI agent demo, it produced faulty knowledge. The approaching months will expose whether or not the staff has solved those basic reliability demanding situations.
I additionally recall to mind this new characteristic as a reasonably extra subtle script, however on the finish of the day, Duties is following a easy, rote set of directions, similar to an ordinary bot. The purpose of many frontier AI labs like OpenAI is to adapt those options into one thing that is in a position to engage with environments, be informed from comments, and make choices with out consistent human enter.
Then again, questions stay about how dependable those scheduled duties shall be and what occurs if ChatGPT fails to ship time-sensitive knowledge. OpenAI’s choice to release Duties in beta suggests it’s nonetheless understanding those main points and desires to collect real-world comments ahead of a much wider rollout.
For now, in the event you’re a paying ChatGPT consumer, you’ll be able to get started experimenting with Duties by means of in search of the “4o with scheduled duties” possibility for your style picker. Simply needless to say it’s nonetheless in beta — so possibly don’t depend on it for that tremendous essential assembly reminder simply but.