In early 2025, Microsoft will add the «translator» feature to Teams, which will be powered by artificial intelligence and allow each participant to speak (or listen to others) in the language of their choice.
The function can imitate your voice in real time, but it will sound in a different language. Initially, the translator will be available in preview only to Microsoft 365 subscribers and will offer up to 9 languages: English, French, German, Italian, Japanese, Korean, Portuguese, Chinese, and Spanish.
«The feature will reproduce the speaker’s message as accurately as possible without adding assumptions or extraneous information», a Microsoft spokesperson told TechCrunch. «Voice simulation can only be enabled if users give consent through notifications during the meeting or by enabling “Voice Simulation Consent” in the settings».
In the future, they also promise to add transcriptions of multilingual meetings with support for 31+ languages.
Microsoft is also exploring the possibility of Copilot in Teams «understanding and repeating any visual content» played on the screen from PowerPoint or the web during a video call.Similarly, AI will be able to summarize the content of any shared file in a chat to save you time.
In other new features, Microsoft is activating Teams Super Resolution on Copilot Plus PCs, which uses the local NPU chip to improve the quality of video calls; and starting in January, Windows app developers will be able to use similar super-resolution APIs to improve blurry images.
Digital voice imitation technologies have become widespread in recent years: Meta recently announced that it was launching a translation tool that can automatically translate voices in Instagram Reels, while ElevenLabs launched a fairly popular platform for dubbing, where your voice can voice anything (which at the same time has repeatedly caused problems with diplomatic facsimiles). Despite the fact that AI translations are usually less lexically rich than the work of real experts, this potential cost savings looks quite attractive to companies. For example, according to data Markets and Markets, the natural language processing technology sector, including translation technologies, could be worth $35.1 billion by 2026.
Source: The Verge, TechCrunch