Google Deepmind unveiled, Wednesday, March 12, two versions of its artificial intelligence models (AI) for robotics: Gemini Robotics and Gemini Robotics-Er. Their goal? “Make AI useful in the physical world,, Explain the company. So far, capacities [de leurs modèles d’IA à mieux résoudre des requêtes complexes en texte, image, audio et vidéo] have been largely confined to the digital domain. »»
Unlike conversational robots and virtual assistants like Chatgpt or Gemini, these models aim to perform tasks in the real world, for industrial or domestic use. The announcement of Google Deepmind is accompanied by videos of demonstration of actions carried out with two robotic arms: put a pencil in a pot, erase a slate, put tools on a wall, move fruit in a basket or mark with a mini ball in a mini basketball basket … these models “Lay the basics of a new generation of robots that can be useful in various applications”,, A predict, on X, the CEO of Google, Sundar Pichai.
You have 83.34% of this article to read. The rest is reserved for subscribers.