Google officially launched Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking, introducing real-time background reasoning and asynchronous tool execution to speech-to-speech agents. Meanwhile, Shanghai AI Laboratory released Atria Dawn Preview, an open-weight 256K context model designed to automate entire research and engineering workflows end-to-end.

We’ll talk about:

• Standard and Extended Thinking variants deploying across Google AI Studio and Search, introducing continuous audio dialogue alongside parallel background API calls.
• An open-weight model released under MIT license scoring 96.0 on DeepSearchQA and targeting end-to-end hypothesis testing, code execution, and report generation.
• OpenAI acquiring camera startup Glass Imaging in a $300M+ deal to incorporate computational photogra ...