Artificial intelligence will continue to evolve. In the near future, we will be able to use technologies which will not only answer questions, but which will also be able to carry out actions for us on the internet. This is a new type of artificial intelligence called an “agent” that is capable of taking control of the user’s computer pointer and input, to navigate the internet or applications, in order to Complete tasks like ordering items online or making reservations.
Many companies are already on board
In other words, we will only have to give an order to the AI, in natural language, for it to execute on the PC. And currently, many players are already developing this technology. Among these, there is the Anthropic laboratory which, in October, presented its functionality called “computer use”. This is already offered in the form of an API to developers who, according to the startup’s explanations, can ask artificial intelligence Claude “to use computers the way people do, by looking at a screen, moving a cursor, clicking buttons and typing text.”
This December, Google also lifted the veil on its Mariner project, which will be able to integrate Google Chrome to take control of it and accomplish online tasks. For example, the user can give a list of companies to Mariner, and request a list of contacts. To obtain this list, the AI will carry out searches on Google, explore the sites of the companies on the list, in order to provide the requested information to the user. It will also be possible to ask Mariner to add items to an Amazon cart, instead of performing this task manually. But for the moment, Mariner is still only a prototype and we don’t know when it will be available in a stable version.
“Project Mariner is an early research prototype built with Gemini 2.0 that explores the future of human-agent interaction, starting with your browser. As a research prototype, it is able to understand and reason through information on your browser screen, including pixels and web elements, such as text, code, images and forms, then use that information via an experimental Chrome extension to perform tasks for you”indicated the firm in the presentation of this technology.
OpenAI and Apple are also on board
OpenAI, the creator of ChatGPT, could also unveil its AI agents in 2025. Kevin Weil, product manager at OpenAI, is convinced that this type of artificial intelligence could become popular. “These more agentic systems are going to become possible, and that’s why I think 2025 will be the year when agentic systems finally become mainstream,” he said, according to an article in the Financial Times, in October.
Furthermore, if Apple does not use the term “AI agent”, it also intends to rely on artificial intelligence to allow users of its platforms to carry out tasks more quickly. Among the new Apple Intelligence features that the firm plans to deploy in 2025, there is a new version of Siri which will be able to understand the information displayed on the screen and which will be able to perform “hundreds of actions” on Apple applications and on third-party applications.
