Chatgpt Can Now Handle Reminders And To-dos

OpenAI is launching a caller beta characteristic successful ChatGPT called Tasks that lets users schedule early actions and reminders.

The feature, which is rolling retired to Plus, Team, and Pro subscribers starting today, is an effort to make nan chatbot into point personification to a accepted integer adjunct — deliberation Google Assistant aliases Siri but pinch ChatGPT’s overmuch precocious relationship capabilities.

Tasks useful by letting users show ChatGPT what they petition and erstwhile they petition it done. Want a regular upwind study astatine 7AM? A reminder astir your passport expiration? Or perchance conscionable a knock-knock joke to show your kids earlier bedtime? ChatGPT tin now grip each of that done scheduled one-time aliases recurring tasks.

To usage nan feature, subscribers petition to premier “4o pinch scheduled tasks” successful ChatGPT’s exemplary picker. From there, it’s arsenic elemental arsenic typing retired what you want ChatGPT to do and erstwhile you want it done. The strategy tin too proactively propose tasks based connected your conversations, though users personification to explicitly o.k. immoderate suggestions earlier they’re created. (Honestly, I consciousness for illustration nan suggestions personification nan imaginable to create annoying slop by accident).

All tasks tin beryllium managed either consecutive successful chat threads aliases done a caller Tasks conception (available only via nan web) successful nan level scheme menu, truthful it’s easy to modify aliases cancel immoderate task you’ve group up. Upon completion of these tasks, notifications will alert users connected web, desktop, and mobile. There’s too a limit of 10 progressive tasks that tin tally simultaneously.

OpenAI hasn’t specified erstwhile (or if) nan characteristic mightiness recreation to free users, suggesting Tasks mightiness enactment a premium characteristic to thief warrant ChatGPT’s subscription costs. The institution has monthly $20 and $200 subscription tiers.

While scheduling capabilities are a communal characteristic successful integer assistants, this marks a displacement successful ChatGPT’s functionality. Until now, nan AI has operated solely successful existent time, responding to contiguous requests alternatively than handling ongoing tasks aliases early planning. The summation of Tasks suggests OpenAI is expanding ChatGPT’s domiciled beyond reside into territory traditionally held by virtual assistants.

OpenAI’s ambitions for Tasks look to agelong beyond elemental scheduling, too. Bloomberg reported that “Operator,” an autonomous AI supplier tin of independently controlling computers, is slated for merchandise this month. Meanwhile, reverse technologist Tibor Blaho recovered that OpenAI appears to beryllium moving connected point codenamed “Caterpillar” that could merge pinch Tasks and fto ChatGPT to hunt for circumstantial information, analyse problems, summarize data, navigate websites, and entree documents — pinch users receiving notifications upon task completion.

As I antecedently wrote backmost successful October, nan emergence of “agentic” AI successful 2025 isn’t conscionable astir technological advancement — it’s astir economics. These agent-like features correspond a strategical measurement to monetize costly AI infrastructure. While OpenAI’s determination to put this functionality down ChatGPT’s paywall was predictable, nan existent mobility remains: will it coming reliable results? The past clip I sewage an OpenAI supplier demo, it produced inaccurate information. The coming months will uncover whether nan squad has solved these basal reliability challenges.

I too deliberation of this caller characteristic arsenic a somewhat overmuch blase script, but astatine nan extremity of nan day, Tasks is pursuing a simple, rote group of instructions, overmuch for illustration a emblematic bot. The extremity of galore frontier AI labs for illustration OpenAI is to germinate these features into point that is tin to interact pinch environments, study from feedback, and make decisions without changeless value input.

However, questions enactment astir really reliable these scheduled tasks will beryllium and what happens if ChatGPT fails to coming time-sensitive information. OpenAI’s determination to motorboat Tasks successful beta suggests it’s still moving retired these specifications and wants to stitchery real-world feedback earlier a wider rollout.

For now, if you’re a paying ChatGPT user, you tin commencement experimenting pinch Tasks by looking for nan “4o pinch scheduled tasks” action successful your exemplary picker. Just retrieve that it’s still successful beta — truthful perchance don’t spot connected it for that ace important gathering reminder conscionable yet.