AI Model Introduction - Google DeepMind introduces AlphaGenome, a new DNA sequence model accessible via API [1] - AlphaGenome aims to help scientists understand DNA, the instruction manual for life [1] Potential Applications - The model can quickly predict the impact of genetic changes, aiding in generating new hypotheses and driving biological discoveries [1] - AlphaGenome's predictions can be used to understand genome function and develop new treatments [1]

RT Robby Stein (@rmstein)AI Mode is coming to India – our first international expansion! Labs users in India can now start testing our most powerful AI search, designed to help with your toughest questions. Exciting times ahead for Search … stay tuned for more this summer. ...

Robotics & AI Advancements - Google DeepMind is bringing powerful AI directly onto robots with Gemini Robotics On-Device [1] - It's the company's first vision-language-action model designed to enhance robot speed, efficiency, and adaptability [2] - The model enables robots to perform tasks and adapt to environments without a constant internet connection [2] Performance & Capabilities - Gemini Robotics demonstrates generality and dexterity in a model small enough to run directly on a robot [1] - The technology offers incredible speed and performance, even in areas with low connectivity [1] Future Implications - The company aims to make robots more helpful and useful to people by continuing to advance this technology [1]

Historical Significance - Alan Turing was born 113 years ago [1] - Alan Turing cracked the German's Enigma code in WWII [1] - Alan Turing proposed the "Turing Test" to judge machine intelligence [1] - Alan Turing designed a chess algorithm by hand in 1950 [1] Athletic Achievement - Alan Turing ran a 2 hours 46 minutes marathon [1] - Alan Turing placed 5th in the 1948 Olympic trials [1]

Performance Improvement - Gemini API 缓存基础设施重大更新，视频缓存命中时，TTFT (Time To First Token) 速度提升 3 倍 (300% faster) [1] - Gemini API 缓存基础设施重大更新，PDF 文件缓存命中时，TTFT 速度提升 4 倍 (400% faster) [1] - 缩小了隐式缓存和显式缓存之间的速度差距 [1] Ongoing Development - 正在进行大型音频文件的优化工作 [1]

AI内容趋势 - AI生成的ASMR内容在网络上迅速普及 [1] - 某些频道仅凭几个完全由AI制作的视频就获得了数千名粉丝 [1] 用户行为 - 观众表示AI ASMR听起来非常逼真，会重复播放 [1] 示例 - 玻璃早餐切割是AI ASMR的一个例子 [1]

Promotion & Eligibility - Google AI Pro 为符合条件的学生提供一年免费 Gemini 服务 [1] - 注册截止日期为 6 月 30 日 [1]

RT Robby Stein (@rmstein)1/ If you haven't tried AI Mode, now's the time. https://t.co/1WVa1u9JZP 🧵 https://t.co/9nUSqR0RAw ...

Product Update - Gemini App 现在支持在 Android 和 iOS 上上传视频 [1] - 此功能已向所有用户开放 [1]

AI Model Capabilities - Google DeepMind 的 Gemini 2.5 Flash-Lite 能够根据前一个屏幕的内容，快速编写用户界面 (UI) 的代码 [1] - 该过程的速度非常快，几乎与点击按钮的速度相同 [1]