AI agents learn when—and how—to interact with people
融中财经, republished by 36氪 and 钛媒体, argues that AI devices must do more than understand requests and context: they must judge whether to engage, how to intervene and when to stop. The report cites Thinking Machines Lab's Interaction Models research preview, Sesame's low-latency voice work and SYNAPUFF's human-machine co-existence research as examples of an emerging effort to treat interaction as a distinct capability requiring new evaluation methods.
As AI glasses, earbuds, vehicle systems and robots become more continuously present, appropriate initiation and restraint become product concerns beyond basic task completion.
当智能开始拥有不同形态的身体,如何与人类共同生活,正在成为AI 产业必须回答的新问题。
一台机器人发现你口渴了,于是走过来递给你一杯水;一副 AI 眼镜识别出你眼前的建筑,主动告诉你它的历史;一只智能耳机注意到你今天的日程很满,于是提醒你该休息了。从技术上看,这些行为都没有什么问题。机器正确理解了信息,完成了推理,并给出了符合预期的回应。
但如果你正在与客户谈判、和朋友争吵,或者只是想安静地待一会儿呢?机器可能理解了你说的话、看见了你眼前的世界,甚至推测出了你的情绪,却依然不知道自己是否应该在此刻出现。
过去几年,人工智能的发展主要围绕一个问题展开:如何让机器变得更聪明?而今,另一个问题正在浮现:当越来越多的智能体进入人类的生活,它们应该如何与人相处?
这并不只是机器人行业的问题。
#Thinking Machines Lab#Interaction Models#Sesame#SYNAPUFF#Google#Gemini#Meta#Ray-Ban Display#Disney#BDX#Olaf