OpenAI CEO Sam Altman visits “Making Money with Charles Payne” at Fox Business Network Studios in New York on December 4, 2024.
Mike Coppola | Getty Images
OpenAI is taking its ChatGPT chatbot to the next level, adding a feature to automate tasks like planning a vacation, filling out forms, making restaurant reservations, and ordering groceries.
The tool, released Thursday, is called Operator. OpenAI describes it as “an agent that can access the network to perform tasks for you,” adding that it is trained to interact with the “buttons, menus, and text fields that people use every day” on the network.
It can also ask follow-up questions to further personalize the task it completes, such as login information for other websites. Users can control the screen at any time.
“Operators are one of our first agents, artificial intelligence that can work independently for you,” OpenAI wrote in a blog post Thursday. “You give it a task and it executes it.”
Currently, Operator is only available to ChatGPT Pro users. It can be accessed through Operator.ChatGPT.com. OpenAI said it eventually plans to expand to Plus, Team and Enterprise users, and integrate Operator into ChatGPT. The company also said it’s currently having trouble with certain tasks, such as managing calendars and creating slideshows.
OpenAI, supported by MicrosoftIt said that users can opt out of some of the company’s training data collection by turning off the “Improving models for everyone” setting in ChatGPT, which means that the data in the Operator will not be used to train its models. The company also said that users can delete all browsing data and exit all websites with “one click” in the privacy section.
Operator competes directly with earlier versions of Anthropic, Amazon– Support for Claude, the artificial intelligence startup behind the chatbot founded by a former OpenAI research director.
October, artificial Introducing “Computer Use”, This ability enables its artificial intelligence agents to use computers to complete complex tasks in the same way humans do. Anthropic says it can interpret content on a computer screen, select buttons, enter text, navigate websites and perform tasks through any software and real-time web browsing.
Jared Kaplan, Anthropic’s chief scientific officer, told CNBC at the time that the tool could “use computers in basically the same way we do.” He said it could complete the task in “tens or even hundreds of steps.”
The generative artificial intelligence market, including OpenAI and Anthropic and GoogleAmazon, Microsoft He Yuanyes Expected to exceed US$1 trillion Income over ten years.
Google recently agreed Anthropic will receive more than $1 billion in new investment, a person familiar with the matter confirmed to CNBC this week. Anthropic in late-stage talks to raise $2 billion Valuation: $60 billion The project is led by Lightspeed Venture Partners, CNBC reported earlier this month.
OpenAI’s latest announcement comes as the company works to advance the potential future of general artificial intelligence. AGI is a vaguely defined benchmark that refers to artificial intelligence that equals or exceeds human intelligence on a wide range of tasks. Experts disagree on whether it’s achievable and how long it would take.
Scale AI provides training data to major artificial intelligence players, CEO Alexandr Wang said on Thursday Interviewed by CNBC He defined AGI as “powerful artificial intelligence systems that can use computers like you and me.” He said it could take two to four years to reach this level of technology.
