Gemini expands, robotics gains

Google is widening Gemini’s reach in two directions at once: more countries for its Personal Intelligence feature, and smarter reasoning for robots in the physical world. (techcrunch.com) Personal Intelligence is an opt-in setting that lets Gemini pull context from a user’s own Google services, including Gmail, Google Photos, YouTube and Search, to answer questions with personal details instead of only public web knowledge. Google introduced it in the United States as a beta on January 14, 2026, then expanded it in March across the United States in Search, the Gemini app and Chrome. (blog.google, blog.google) On April 14, Google said it was launching Personal Intelligence in India, where it starts with Google AI Pro and Google AI Ultra users, with free-user access planned in the coming weeks. TechCrunch reported the broader international rollout excludes the European Economic Area, Switzerland and the United Kingdom, and The Verge said paid subscribers in those newly added regions get it first. (blog.google, techcrunch.com, theverge.com) The feature pushes Gemini closer to the role Google has long wanted for its assistant software: not just answering generic questions, but using a person’s inbox, photos and search history to help with travel plans, schedules and recommendations. Google says users choose which apps to connect and can turn those links off at any time, while also warning that Gemini can still misread context or connect unrelated details. (blog.google, techcrunch.com) The second half of Tuesday’s news is about robotics, where the problem is different: software has to interpret a messy real-world scene, decide what matters, and guide a machine through it safely. Google DeepMind said its new Gemini Robotics-ER 1.6 model is built for that “embodied reasoning,” meaning a robot uses camera views and sensor data to understand its surroundings before acting. (blog.google) Google said Gemini Robotics-ER 1.6 adds a new instrument-reading skill, so robots can interpret pressure gauges and sight glasses in industrial settings. The company said the model is available starting now through the Gemini Application Programming Interface and Google AI Studio, along with a Colab notebook for developers. (blog.google, deepmind.google) Boston Dynamics said it is integrating Gemini and Gemini Robotics-ER 1.6 into Spot, its four-legged inspection robot, through the Orbit platform’s artificial intelligence and visual inspection tools. The companies said that should help Spot handle tasks like reading gauges and adapting to changing industrial environments with less hand-coded behavior. (therobotreport.com, theverge.com) Taken together, the updates show Google applying the same Gemini push to two very different jobs: one assistant that knows more about your digital life, and one robot model that knows more about the room it is standing in. Both rollouts arrived on April 14, with Google widening access for consumers while giving developers and robotics partners a new model to test in the field. (techcrunch.com, deepmind.google)