Tag
NPCTalk 是一个开源项目，可将普通游戏 NPC 变成能用麦克风实时语音对话、自带音色并做出身体反应的角色，首批应用于 GTA V，并打包为 AI 编码代理可执行的指令，支持任意游戏接入。
An indie developer launched Historai, an app that turns any historical topic into a researched two-host podcast with period artwork, supporting mid-episode voice questions. The builder used bake-offs across GPT, Claude and Gemini to assign the best model to each task (research, scriptwriting, artwork, voices).
Flam App has announced a $40M Series B funding round led by QED Investors to develop interactive AI characters. These characters use screen and voice context, powered by the Fable 23B model for fast token streaming.
A playable demo uses the GPT-Live-1 API to enable real-time world generation in games via voice commands, illustrating a new form of interactive media.
This paper presents SurgicalRoomAgent, a voice-interactive multi-agent system for smart operating rooms that uses large language models to enable natural language device control, focusing on architecture design and key technologies like KV Cache optimization and progressive prompt disclosure to reduce latency in sterile environments.
GPT-Live-1 is now available in the API, enabling developers to create voice-based applications with real-time conversational capabilities similar to ChatGPT.
This paper introduces S³-Bench, a systematic evaluation framework for assessing speech interaction models as scientific voice assistants, covering 10 major disciplines and analyzing performance tradeoffs in scientific interactions.
The tweet highlights that Gemini audio models allow users to interact with their Agent using voice commands in Gmail and Google Docs, enhancing productivity.
The article discusses how AI voice agents should handle situations where customers disagree with backend system data, exploring design options like additional verification or human handoff.
A South Korean AI app goes viral for enabling lifelike video conversations with AI characters that use voice, lip sync, facial expressions, and camera context, signaling a shift from text-based interfaces to real-time video-native interactions.
The article demonstrates ChatGPT's 'Work' mode in the desktop app, enabling voice-controlled task initiation and background automation for activities like brainstorming and generating visual proposals.
A demonstration of using ChatGPT Voice on desktop to manage tasks like drafting blog posts, debugging code, and coordinating launch preparations through natural speech.
OpenAI released the GPT-Live model, which has persistent presence and temporal awareness capabilities, supports interruptible voice interaction and real-time timing functions, enhancing the naturalness and practicality of voice assistants.
Google DeepMind is experimenting with a Gemini-powered AI pointer that understands user pointing intent, context, and voice, and performs actions across applications, thereby reshaping human-computer interaction.