The giant in the field of Open AI technology, announced a new service: Chatgpt Agent, an assistant who can complete complex tasks. Given the advanced capabilities of the new AI agent, the company also introduced new safety measures.
Chatgpt Agent comes in the context of a fierce competition among Silicon Valley companies for the development of more and more performing agents, according to Mediafax.
The artificial intelligence assistant can automatically browse the user calendar, generate editable presentations, run code and access third -party applications such as Gmail or Github via Chatgpt connectors, according to OpenAI. At the same time, the agent can plan shopping, analyze competition and create complex reports or presentations, only on the basis of orders in natural language introduced by the user, notes Techcrunch.
How Chatgpt Agent works
The Chatgpt agent from Openai combines advanced functionalities from the previous instruments developed by the company, such as operator and Deep Research, and is available from Thursday for Pro Subscribers Pro, Plus and Team.
Chatgpt Agent has access to the terminal for code execution and uses APIs to access external applications. His performance scores far exceed the previous models.
Chatgpt Agent can do things in your place. Photo shutterstock
Chatgpt Agent comes with new safety measures
Because the new agent AI has advanced capabilities, Openai has introduced new safety measures: temporarily disabled memory function to prevent possible type attacks “Prompt Injection”, And each request is analyzed in real time, checking if the prompt is related to biology or chemistry, areas considered sensitive.
What can, in everyday life, chatgpt agent
The new service not only can answer your questions, but can do things in your place, notes ProTV.
You can ask them, for example, to plan a meal and you can order ingredients if you give it access if you give it access to your account. Likewise, you can buy clothes and many other things including clothes.
