Google DeepMind VP says robotics doesn't need to wait for AGI to become useful
retained text
Carolina Parada, VP of Robotics at Google DeepMind, talks about how the company is approaching the next phase of robotics. Their bet isn’t that one robot form factor wins. It’s that there will be a lot of different robots, and the bigger opportunity is building an AI layer that can work across all of them. She also gets into the split we’re seeing now between companies trying to deploy robots for narrow tasks today and those waiting for something closer to physical AGI. Her view: you don’t need full AGI for these systems to start being useful. She points to Boston Dynamics using Gemini Robotics with Spot for inspection tasks, and says manufacturing and logistics will likely see this kind of capability first because the environments are more structured. Full: [https://www.youtube.com/watch?v=YQmVLjt2PCA](https://www.youtube.com/watch?v=YQmVLjt2PCA)
document fetched 2026-09-18 00:39 · canonical https://reddit.com/r/robotics/comments/1wivifk/google_deepmind_vp_says_robotics_doesnt_need_to