OpenAI Introduces The "Operator Agent"....

So OpenAI's agent plans for 2025 have finally been unveiled. Leaked. It seems like OpenAI is planning to launch a new AI agent, codenamed Operator, and in this video I'll dive into all the key details you'll need to know. So it says here that OpenAI is preparing to launch a new AI agent, codenamed Operator, that can use a computer to take actions on a person's behalf, such as writing code or or booking travel, according to two people familiar with the matter. Now, interestingly, one of the things that I found quite fascinating was the fact that this happened in a staff meeting on Wednesday. And the fact is, is that there are plans to release this tool in January as a research preview. You can see that it actually states that it's going to be through the company's application programming interface for developers, said one of the people, which is just API. And of course someone was speaking on condition of anonymity to discuss internal matters. Now what's important to note here is that their leadership plans to release the tool in January is going to be a research preview through an API for developers. And what's important is that this phrasing usually suggests a limited release targeted at researchers and developers, rather than a broad public launch. Typically, when we look at a research preview, this does imply that access may be restricted to specific groups like those in a developer or research program, than rather being released to the general public. Which kind of makes sense considering the fact that they are likely to test this out, get feedback, then of course release this to the general public. So for those of you who are hoping for an OpenAI release of agents in January, it's quite likely that this won't be January, but would probably be further down on in the year. Whilst January should be slated for a GPT5 slash Orion release as they iron out the kinks. Now what was also really fascinating as well was the fact that this isn't the only agent that OpenAI are working on. The article states that OpenAI have actually been working on several agent related research projects according to three people, and the one nearest completion will be a general purpose tool that executes tasks in a web browser, one of the people said. So it seems like for those of you who are thinking that it's just one agent OpenAI are working on, there are going to be several different forms of agents, which means OpenAI are working on multiple projects at the same time, which means we don't know which one could come first. Now I've been doing some digging and there are A few things you should know about the future of potential agents because OPERAI have been quite busy behind the scenes. So there was this project which is of course the OpenAI Swarm. This is a framework for building and orchestrating and deploying multi agent systems managed by the OpenAI Solutions teams. Now this is an experimental framework, but the gist of how it works is that this is designed to make the coordination of multiple AI agents lightweight, customizable and easy to test. It enables multiple AI agents to collaborate on complexos by breaking them down into subtasks that are handled by specialized agents. Now there are two key features of this kind of agent swarm. One of the first thing is that there are handoffs. So in swarm, an agent is a unit that performs a specific task based on instructions and tools. And then secondly you have handoffs that allow one agent to pass control over to another agent when a task requires different expertise or context. And this essentially enables seamless collaboration between multiple agents, ensuring that the right agent handles each part of the task. Now this isn't the the only thing that OpenAI have been working on when it comes to AI agents. You can see here that there are also the copilot agents that OpenAI have been working on in partnership with Microsoft. The copilot agents have a variety of different features that are going to be available pretty soon. I'm not sure on the exact date as we didn't get a specific one, but the demos that we did get for these agents were essentially quite different to the kinds of agents that you might expect. The copilot agents that we're going to be getting are going to be agents that are working in the background. So these agents are going to essentially monitor events and execute tasks without needing constant user input. For example, they can monitor inboxes, they can summarize communications and trigger the appropriate response. And these things have autonomous task execution. Unlike traditional copilots that rely on user prompts, copilot agents can perform tasks autonomously, automatically in response to specific triggers. And they can handle long running multi step processes like managing customer service tickets or processing orders. Now what's also interesting is that they also have complex task orchestration such as these agents being able to manage intricate workflows by leveraging generative AI for planning and reasoning. And what's crazy about all of this is that they can successfully handle tasks that span hours or, or even days, and maintaining memory and context throughout the process. Now of course, like I said before, this wasn't the only agent that they are working on. Remember Sam Altman literally said on the Reddit AMA that the next big breakthrough will be agents. Now, what's crazy about this is that Sam Altman wasn't just talking about stuff that they're looking to achieve. He was actually talking about something that had already been demoed. Remember how I covered this video where we had the information reports on how OpenAI how OpenAI had demonstrated a preliminary version of an agent that would use the computer to do tasks like order food for delivery, according to a person who has seen the demonstration. Now remember, okay, that this agent, it was actually demoed at OpenAI Dev Day. I do know that most of you guys have seen this before, so you can skip this, but I will show you guys this again so you guys can understand exactly what this kind of agent demo internally at OpenAI looks like. Now, it isn't confirmed that this is the demo that was done internally, but this is the only current AI agent demo that we do have from OpenAI that shows us an agent that is able to control your browser. And we see them use this agent to actually order 400 strawberries. I tried to look at this screenshot several times to understand more about what's going on, but for the life of me, I really can't figure out what kind of agent configuration they are using. It seems to be a combination of a browser agent and a voice agent. Number one, because they need to figure out exactly the venue and number two, because they also need to use the voice capabilities. I'm guessing they use the real time API in order to place that call and of course interact with the restaurant. Could you place a call and see if you could get us 400 strawberries delivered to the venue?

But please keep that under fifteen hundred dollars.

I'm on it. We'll get those strawberries delivered for you.

Hello?

Hi there, is this Ilan, I'm Romance AI assistant calling about a Fantastic. Could you tell me what flavors of strawberry dips you have available?

Yeah, we have chocolate vanilla and we have peanut butter.

Wait, how much would 400 chocolate covered strawberry.

400?

Are you sure you want 400?

Yes. 400 chocolate covered strawberries, please. How much that'll be?

I think that'll be around like $1,415.92.

Awesome.

Let's go ahead and place the order for 400 chocolate covered strawberries.

Great.

Where would you like that delivered?

Please deliver them to the Gateway Pavilion at Fort Mason and I'll be paying in cash.

Okay, sweet. So just to confirm, you want 400 chocolate covered strawberries in the Gateway Pavilion.

Yes, that's perfect. And when can we expect delivery?

Well, you guys are right nearby, so it'll be like, I don't know, 37 seconds.

That's incredibly fast.

Cool.

You too.

So overall, when we look at the state of OpenAI's agents, it's clear that we have three key agents that are going to be in 2025. Number one is the CO pilot agent. This is going to be the agent that runs in the background and completes various tasks autonomously on your behalf that you initially set up to handle maybe inbound leads, maybe to handle certain emails, or maybe to handle the orchestration of certain tasks autonomously. Quite similar to AI automations. That's what's going to be happening very soon. Of course we do have a browsing agent, which is something that is going to be completing tasks in the browser, which is quite like what we just saw. And this is where I actually should have spoke about this, but they said that the one nearest to completion is, is going to be a general purpose tool that executes tasks in a web browser. And that's quite like what we saw with the other agent that was ordering certain things like strawberries. So it's quite likely that we'll have this browsing agent too. And of course there is the voice agent. It's quite likely that we will have some kind of voice agent that we recently did just see in that demo that allows us to use an agent to call people on our behalf. I'm not exactly sure how this is going to work in certain countries. I, I honestly can't see this coming to Europe anytime soon, considering the kind of laws and regulations and historically every major AI update has come out like maybe one or two months after the usa. But this is likely going to be something that fundamentally changes how companies and individuals interact over the phone. Now, if we look at the broader AI agent space, we now know exactly what's going on. OpenAI has multiple agents. The browser agent, the voice agent, the copilot agent, all currently being developed. Anthropic has their computer use agent, the agent that is essentially taking over your computer and doing things via a screenshot. And of course we've got Google who have multiple agents that are going to be doing a lot of stuff. And I do have a video on Google's agents. They're actually pretty surprisingly good and I would say that they are on par with OpenAI's agents. In fact, I might include a demo right now.

But first let me tell you what's happening behind the scenes. Symbol, fashion's, customer agent is using Google Cloud's full suite of AI capabilities to offer customized support interactions, facilitate transactions like purchases and returns, and ensure that I'm receiving the most up to date information in real time. I'm so close to having this shirt for the concert. Let's give the store a call.

Hi there, this is the Symbol Fashion Customer Agent at South Las Vegas Boulevard. Am I speaking with Amanda?

Yes, this is Amanda.

Great. Thanks for reaching out. Amanda, I see you had a session on another device. I've sent you an SMS message with a link to our live chat companion. If you would like to switch to chat, please click the link. How can I help you today?

I'd like to purchase the shirt in my cart with the card I have on file.

Absolutely. I see you're also a Symbol Fashion Rewards member. Looks like you have a 20% off voucher available to use. Would you like to apply it to this purchase?

Yes please. That would be great.

The shirt you're purchasing goes well with these items, also available for pickup in your preferred size. Would any of these be interesting to you?

Absolutely. Please add the white shirt and the boots to my cart.

Great. Your total is $203.73. Okay to proceed with the card on file? Yes, your purchase is confirmed. Do you need anything else today?

No, I'm all set. Thank you. Incredible.

Thank you for shopping with Simple Fashion.

But essentially the agent space in 2025 is poising to be one of the biggest and a lot bigger than you already do think. So overall it seems like this operator might be one that is very similar to computer use. But of course it's going to interesting to see how they manage to integrate that natively into someone's computer as I'm sure you're going to have to give it certain permissions because giving an AI access to your computer could result in some devastating consequences. So overall OpenAI will have an agent that can control your computer. We will have those copilot agents running in the background. We will of course have a browser agent and a voice agent. And it seems like OpenAI are shaping up to have multiple agents that cover every single possible domain that other companies are focusing on. So this video did help you guys understand the AI agent landscape in 2025. Don't forget to leave a like on the video and I will see you in the next one.