On Tuesday, Microsoft unveiled a major update for Copilot, its conversational agent. The main new features include Copilot's ability to interpret images and its increased personalization capabilities.
Unveiled on Tuesday by Mustafa Suleyman, Microsoft's vice-president in charge of Artificial Intelligence, the new Copilot update should enable Microsoft to move up a gear in the face of strong competition from conversational agents. This update illustrates both the speed of innovation and its availability to the general public in the field of AI. For the record, Mustafa Suleyman, recruited by Microsoft last March, is a former head of Google DeepMind, one of the pioneers of AI...
Visual AI: a step towards natural interaction
This major update first introduces visual functionalities. Copilot can now not only respond to voice commands, but also analyze and interpret images. This will enable the AI to "see" what the user sees on a page, with the ability to answer questions in real time and suggest actions and adapt without interrupting the workflow.
Although this feature is still in the test phase, it promises to extend the uses of AI, particularly for the processing complex documents assures Microsoft. This feature will also be available, initially, "for a limited list of heavily-used sites to ensure that the experience is safe for everyone", insists Microsoft. The Redmont-based company insists on security, by guaranteeing that the data collected will be deleted after use.
Voice interaction and personalization: towards more intuitive AI
In addition to this visual capability, Microsoft is highlighting the new personalization tools from Copilot. Users can choose from four different voices assigned to the AIfor more natural interaction. This feature will initially be deployed in English-speaking countries, facilitating more fluid and intuitive communication.
A another major advance is the personalization of interactions Copilot will now be able to tailor its responses based on conversation history. However, this feature will not be available in the European Union and the UK, due to strict data protection regulations.
Towards more powerful AI: adding "deep thinking" to the mix
Microsoft finally announces a new "deep thinking" feature. Developed in partnership with OpenAI, this innovation enables Copilot to solve more complex problems. The AI will be able to adapt to more sophisticated scenariosreinforcing its ability to provide customized assistance in complex professional or technical situations. The integration of this functionality reflects Microsoft's desire to reduce the boundaries between professional and private use.
A fast-changing competitive environment
The arrival of this update comes against a backdrop of increasing competition in the AI market. Microsoft, a pioneer in the field thanks to its partnership with OpenAI, is now seeing Google and Apple pick up the pace. Google, with its Gemini AIand Apple, via its Apple Intelligencecapitalize on their presence dominant on smartphones to promote their own solutions. For its part, Meta relies on the popularity of its social networks to integrate its agent. Meta AI with its users. In this race for innovation, Microsoft focuses on diversifying its functionalities Copilot to stay one step ahead, particularly through its applications.
In short, with this update, Microsoft is taking a decisive step forward in the evolution of its consumer AI, bringing significant improvements in terms of visual interaction, personalization and complex problem solving. The Group is thus seeking to assert its position in an industry undergoing rapid transformation, while remaining attentive to data protection issues.
