This Thursday, OpenAI announced Operator, an AI agent that can complete tasks on partner websites and apps (23). Now, you can ask the AI to reserve a restaurant or place a delivery order through Uber. But OpenAI isn’t the only one diving into the era of AI agents.
4 AI Agents to Know:
- OpenAI Operator
- Copilot
- IA Claude
- Gemini
OpenAI Operator
Operator, an AI agent from OpenAI, utilizes the GPT-4 model and can navigate websites and apps to complete specific tasks for you. It can perform actions like typing, clicking, and scrolling through pages on the internet.
Users can ask the agent to make purchases on online marketplaces, fill out forms, or even generate images based on their requests.
The AI agent is still in the research phase and has some limitations, according to OpenAI.
This feature is only available to ChatGPT Pro plan subscribers (around R$1,200 per month) in the United States. OpenAI has stated that they plan to expand access to users on the Plus, Team, and Enterprise plans.
Copilot
Copilot, Microsoft’s AI, is integrated with Office tools like Word, Excel, PowerPoint, Teams, and Outlook. With Copilot, users can generate complex reports in spreadsheets and request meeting summaries, among other tasks.
In October 2024, Copilot Studio introduced the ability to create autonomous agents. This allows companies using Microsoft tools to build their AI agents for daily tasks.
Copilot access is paid and included in plans like Microsoft 365 for Business.
Claude
In October 2024, Anthropic announced an API for Claude capable of interacting with the computer by typing commands.
This way, AI can automate specific tasks based on commands sent by the user. For example, it is possible to ask AI to use data on your computer and the internet to answer a form.
Anthropic agents can also be used in customer service platforms to automate responses or ask for decision-making help in big data analytics.
Gemini
Gemini is Google’s AI platform with multimodal capabilities, meaning it can interpret text, images, and graphics.
At the Galaxy Unpacked event last Wednesday (22), Gemini Live was introduced. With this feature, users can send photos, files, and even YouTube videos to the AI, providing more context for conversations with the assistant.
Google’s AI can now handle tasks that integrate with applications like Google Maps and Gmail, and it will also work with Samsung apps. For example, you can ask Gemini to search for the date of an event on Google and set an alarm on your phone as a reminder.
What are AI agents?
Artificial intelligence agents are software programs that collect and use data to perform tasks most efficiently.
The agent will autonomously determine the best approach to deliver a response based on a user’s request.
For example, an AI agent integrated into a customer service center can leverage data from previous conversations, the internet, and its own system to provide the most comprehensive answer to the user’s query.
. Now, you can ask the){kind=link}