Summary
Openai announced on Thursday the launch of an artificial intelligence assistant known as “Chatgpt Agent”, marking the greatest incursion of the technological company in task management applications through AI, a field that is gaining popularity as users seek a more practical use of their virtual assistants.
Key data
- Chatgpt Agent can do diverse jobs and tasks for users, such as creating presentations, buying ingredients for a meal or sending expenses, depending on the announcement.
- Users can also program tasks so that they are automatically repeated, such as updating spreadsheets with new data at the beginning of each week, Openai said.
- Openai acknowledged that there are “new risks” with these capabilities, and although he has implemented security measures to handle sensitive information on the web and other factors, he warned that malicious indications on web pages could deceive Chatgpt Agent to share private data with a hacker or “perform a harmful action in a place where the user has logged in.”
- Users can interrupt Chatgpt Agent at any time to clarify instructions or completely change the task, allowing them to resume from where he stayed without losing progress.
Soser context: OpenAI launches ‘chatgpt agent’, its most advanced assistant: what you should know
When will chatgpt agent be available?
The AI wizard will be available for Chatgpt Pro users from Thursday. Chatgpt Plus and Chatgpt Team users will get “in the next few days”, while users of Enterprise and Education plans will have it available in the coming weeks.
What privacy measures can users take against chatgpt agent?
Openai indicated that users can erase all their navigation history and close all active sessions on websites linked to their tasks with a single click. However, website cookies will continue to be stored according to the policy of each site.
What to observe
Openai emphasized that Chatgpt Agent is still in an initial stage and can make mistakes when generating presentations or delivering basic results. The company expects to improve its efficiency and depth while seeking to balance its capacities with user security.
Additional data
The development of OpenAI in artificial intelligence for task management has been accompanied by an avalanche of financing. In March, the company closed an investment round for 40 billion dollars, the largest ever registered in the private technological sector. In addition, Openai has compromised billions of dollars in the “Stargate Project”, a new company created with SoftBank, Oracle and MGX that plans to invest 500 billion dollars in AI infrastructure in the United States.
MOST CONTEXT: OpenAI will launch a web browser that will compete with Google Chrome
Key history
Chatgpt Agent is not Openai’s first attempt at homework management assistants. In January, the company launched “Operator”, a system capable of web tasks for users, such as online supermarket orders, purchase of tickets, reimbursement calculations and more. Operator is not completely autonomous and Openai describes him as one of the three pillars that gave rise to Chatgpt Agent. The new tool combines the ability to interact with websites, the information synthesis of Chatgpt Deep Research and Chatgpt’s conversational skills. The task management applications released this year bring the company closer to the creation of a general artificial intelligence (AGI), a concept that refers to machines capable of self -apprehending and completing tasks comparable to humans.
This note was originally posted in Forbes US
Follow technology information in our specialized section
