Skip to main content

ChatGPT just dipped its toes into the world of AI agents

OpenAI's ChatGPT blog post is open on a computer monitor, taken from a high angle.
Photo by Alan Truly / Digital Trends

OpenAI appears to be just throwing spaghetti at this point, hoping it sticks to a profitable idea. The company announced on Tuesday that it is rolling out a new feature called ChatGPT Tasks to subscribers of its paid tier that will allow users to set individual and recurring reminders through the ChatGPT interface.

Tasks does exactly what it sounds like it does: It allows you to ask ChatGPT to do a specific action at some point in the future. That could be assembling a weekly news brief every Friday afternoon, telling you what the weather will be like in New York City tomorrow morning at 9 a.m., or reminding you to renew your passport before January 20. ChatGPT will also send a push notification with relevant details. To use it, you’ll need to select “4o with scheduled tasks” from the model picker menu, then tell the AI what you want it to do and when.

Recommended Videos

The new feature will begin rolling out in beta to Plus, Team, and Pro users over the next few days. You’ll be able to set up to 10 separate tasks on desktop, web, and mobile during the beta phase, though the tasks manager menu will only be available through the web interface. The AI will be able to push reminder notifications through the macOS desktop app, your browser, or the iOS and Android mobile apps. The idea behind Tasks seems to be nearly identical to what the myriad native and third-party reminder apps on the market already do, just through a different brand’s interface.

Tasks looks to be OpenAI’s initial foray into AI agents, specialized models designed to take independent action while automating repetitive daily tasks. It’s a field that is expected to grow massively in 2025. Anthropic’s Claude already unveiled its Computer Use API in November, as did Microsoft with its Copilot Actions. Google is reportedly working on an agent of its own, code-named Project Jarvis, and announced earlier this week that its automotive agent will power the next generation of Mercedes-Benz MBUX navigation and entertainment systems.

Andrew Tarantola
Former Computing Writer
Andrew Tarantola is a journalist with more than a decade reporting on emerging technologies ranging from robotics and machine…
5 AI apps with deep research features to rival ChatGPT
Deep Research option for ChatGPT.

Artificial intelligence brands are in fierce competition, and their next steps are to make AI tools smarter by allowing them to execute deep search functions that can provide expert-level results and analyze larger amounts of information in a shorter time. Several companies have announced deep research features in recent weeks and months that excel in areas such as finance, science, marketing, and academics. Research that would have taken a person weeks or months can be achieved in a fraction of the time, with a properly detailed prompt. 

Deep research features are considered AI agents that can work independently and will allow you to make a query and let the AI process for several minutes while it generates the information and returns when it is finished to display the results. They are considered the first steps toward the concept of artificial general intelligence (AGI), which some define as a model that can process a query based on novel data that it has not been trained on, and it can produce unique content. However, we’re not quite there yet, and the main premise of deep research tools is processing large amounts of data and making it easier to understand.

Read more
Meta’s new AI app lets you share your favorite prompts with friends
Meta AI WhatsApp widget.

Meta has been playing the AI game for a while now, but unlike ChatGPT, its models are usually integrated into existing platforms rather than standalone apps. That trend ends today -- the company has launched the Meta AI app and it appears to do everything ChatGPT does and more.

Powered by the latest Llama 4 model, the app is designed to "get to know you" using the conversations you have and information from your public Meta profiles. It's designed to work primarily with voice, and Meta says it has improved responses to feel more personal and conversational. There's experimental voice tech included too, which you can toggle on and off to test -- the difference is that apparently, full-duplex speech technology generates audio directly, rather than reading written responses.

Read more
It’s not your imagination — ChatGPT models actually do hallucinate more now
Deep Research option for ChatGPT.

OpenAI released a paper last week detailing various internal tests and findings about its o3 and o4-mini models. The main differences between these newer models and the first versions of ChatGPT we saw in 2023 are their advanced reasoning and multimodal capabilities. o3 and o4-mini can generate images, search the web, automate tasks, remember old conversations, and solve complex problems. However, it seems these improvements have also brought unexpected side effects.

What do the tests say?

Read more