Source:
ChatGPT:
Microsoft AI CEO Mustafa Suleyman explained the feature was created in response to feedback that people feel more comfortable speaking to a “face.” Unlike the earlier Copilot Appearances feature, which used cartoon-like blobs, Portraits focuses on human-inspired designs. Importantly, the avatars are not photorealistic but intentionally stylized to reduce the risk of misuse while still offering expressive variety.
The technology powering Portraits comes from VASA-1, a Microsoft Research project that generates lifelike animations from a single image without the need for complex 3D modeling. This lets Copilot create responsive visual interactions quickly and at scale.
The move follows similar developments from rivals, such as X’s Grok chatbot, which recently introduced customizable 3D avatars, including anime-inspired companions with controversial NSFW options. Microsoft, however, is rolling out Portraits cautiously, citing safety concerns around AI-human interactions. Safeguards include age restrictions (18+), time limits, and clear indicators reminding users they’re talking to AI.
By combining expressive visuals with trusted Copilot intelligence, Microsoft aims to make AI conversations more engaging, while still prioritizing safety and responsible use. The experiment underscores a broader industry trend: turning chatbots into more personable companions that blend utility with emotional connection.
No comments:
Post a Comment