Google DeepMind, the AI research division of Google, announced on Wednesday the launch of new AI models named Gemini Robotics, aimed at helping real-world machines interact with objects and navigate their surroundings.
The lab shared several demonstration videos showcasing robots using Gemini Robotics to fold paper, place glasses into a case, and perform various tasks based on voice commands. DeepMind stated that Gemini Robotics was designed to generalize behaviors across different robotic hardware and link visible objects to potential actions.
ICYMT: 17,000 MoMo businesses collapsed due to E-Levy – Agents association reveals
According to DeepMind, tests showed that Gemini Robotics enabled robots to excel in unfamiliar environments not covered during training.
READ ALSO: Ministry of Health kicks off 2025 Glaucoma awareness week
The lab also introduced a streamlined version, Gemini Robotics-ER, for researchers to develop their own robotics control models, along with a benchmark called Asimov to assess risks associated with AI-driven robots.
SOURCE: TECH CRUNCH