Microsoft Unveils Major Redesign of Copilot with Voice and Vision
Microsoft is introducing a major redesign of its Copilot experience, adding voice and vision capabilities to make it a more personalized artificial intelligence (AI) assistant. Inflection AI influence this transformation, with new features including a virtual news presenter mode, the ability for Copilot to see what you are viewing, and a natural voice interaction like OpenAI’s Advanced Voice Mode.
The redesign spans mobile, web, and the dedicated Windows app, adopting a card-based user interface inspired by Inflection AI’s Pi assistant. Earlier this year, Microsoft brought in several experts from Inflection AI, including Mustafa Suleyman, now CEO of Microsoft AI. This marks Suleyman’s first notable change to Copilot since taking over the consumer AI assistant division.
Suleyman expressed in an open letter that Microsoft AI aims to create a supportive and helpful technology era. The new Copilot interface is warmer and more inviting, featuring a personalized Discover page tailored to your conversation history, with useful searches, tips, and relevant information.
The consumer version of Copilot was moved to Suleyman’s team, allowing for more experimentation with personality and customization. Yusuf Mehdi, Microsoft’s executive vice president and consumer chief marketing officer, noted that insights from the Inflection AI team have significantly influenced the redesign.
In addition to the fresh look, Microsoft is enhancing its vision of an AI companion for everyone by adding voice capabilities like those in ChatGPT. Users can now have natural conversations with Copilot, choosing from four voice options. Mehdi emphasized the importance of voice interaction, suggesting it will shape the future of AI assistance.
Copilot Vision, another key feature, enables the AI to see and interact with the content on your screen, providing recommendations and answering questions about what you are viewing. These sessions are opt-in and ephemeral, with strict limitations on the types of websites it can interact with initially.
Read More: