Connect with us

AI

OpenAI’s operator AI agent gets smarter with new O3 model

It act like a digital assistant that can work mostly on its own, without needing much input once a request is given.

Image: KnowTechie

Just a heads up, if you buy something through our links, we may get a small share of the sale. It’s one of the ways we keep the lights on here. Click here for more.

OpenAI is making a major upgrade to its AI agent called Operator, which can browse the web and use cloud-based software to help users complete tasks. 

This tool is designed to act like a digital assistant that can work mostly on its own, without needing much input once a request is given.

Previously, Operator used a special version of OpenAI’s GPT-4o model. But now, it’s being updated to use a newer and more advanced model called o3. 

The o3 model is part of OpenAI’s latest “reasoning” series, which is especially good at solving complex problems, like math and logic tasks, better than earlier versions.

This updated version, called o3 Operator, has also been specially trained to handle safety issues better. 

OpenAI says it has added more safety-focused training so that the AI knows when to say “yes” or “no” to certain types of requests, such as rejecting illegal tasks or protecting personal information. 

According to a technical report, the o3 Operator is also harder to trick through a method called prompt injection, a way of trying to get the AI to break its rules.

It’s important to note that while o3 is powerful in coding, this version of Operator doesn’t come with direct access to a coding environment or computer terminal. So, while it can understand and reason through code, it won’t execute it directly.

Operator is part of a larger race among tech companies to build advanced AI “agents” that can use computers like a human assistant. 

Google has launched similar tools, such as Gemini’s computer-use features and a product called Mariner, while Anthropic’s AI can also perform basic computer tasks.

OpenAI’s update to Operator makes the AI smarter, safer, and better at reasoning, which helps it complete tasks more reliably, all while maintaining strong boundaries for user privacy and security.

What are your thoughts on this new update? Do you use Operator? Let’s talk more down below in the comments, or via our Twitter or Facebook.

Follow us on Flipboard, Google News, or Apple News

Ronil is a Computer Engineer by education and a consumer technology writer by choice. Over the course of his professional career, his work has appeared in reputable publications like MakeUseOf, TechJunkie, GreenBot, and many more. When not working, you’ll find him at the gym breaking a new PR.

Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

More in AI