Farm fields might just be the ultimate proving ground for physical AI: zero room for error, chaotic environments, and real-world impact.
Had a blast joining @gbrulte and @ReservoirFarms on @RoadToAutonomy to discuss how @GoogleDeepMind is thinking about robotics, using multimodal models for things like crop scanning, and bridging the gap from working-in-a-lab to working-in-real-dirt. 🤖🌱
Farm robots that work in simulation sometimes fall apart in the field, and on average ag robotics startups wait six to nine months after raising venture capital before they even get access to a farm. @DynamicWebPaige, AI Developer Relations Lead at @GoogleDeepMind, and Danny Bernstein, Founder & CEO of @ReservoirFarms, joined @gbrulte on The Road to Autonomy to discuss rugged physical AI and automation on farms. Episode Chapters 00:00 Rugged Physical AI Goes Mainstream 01:02 Google DeepMind's Robotics Ambitions 04:11 Bonsai Robotics' $60,000 Autonomous Tractor 06:43 Why Startups Wait Months to Reach a Farm 10:17 Building Trust Between Growers and Engineers 14:45 Physical AI as a Growth Market 16:24 How Robotics Moved to the Center at Google 18:31 Broccoli Bred for Automation 22:44 Gemini Vision for Crop Scanning 24:39 Is Agriculture Ready for Automation? 26:34 Physics Replacing Chemicals 29:57 Weather Next and Google Earth Engine 33:17 The Future of Physical AI 35:19 @AUTNMYAI
👩💻 Paige Bailey · x.comPublished
Original title: Farm fields might just be the ultimate proving ground for physical AI: zero room for error, chaotic environments, and real-world impact.
