Google's DeepMind , Pens 'Robot Constitution' , to Govern Helper Robots.<br />On January 4, Google announced a set of rules <br />meant to govern the artificial intelligence system <br />behind the company's intelligent machines.<br />Gizmodo reports that the so-called <br />'Robot Constitution' is based on <br />Isaac Asimov's 'Three Laws of Robotics.'.<br />Gizmodo reports that the so-called <br />'Robot Constitution' is based on <br />Isaac Asimov's 'Three Laws of Robotics.'.<br />Before robots can be integrated <br />into our everyday lives, they need <br />to be developed responsibly with <br />robust research demonstrating <br />their real-world safety, Google DeepMind robotics team, via blog post.<br />The 'Robot Constitution' was written by Google's DeepMind, the tech company's primary AI research wing.<br />Google reportedly has plans for intelligent robots <br />to soon help around homes and offices in the near future.<br />These robotic helpers would run on a system of large language <br />models (LLMs) and neural networks that would allow <br />them to make decisions and be aware of their surroundings.<br />Google says the 'Robot Constitution' is meant to act as part of a <br />system of "layered safety protocols" which govern the machines <br />that have already been deployed on the company's campus.<br />DeepMind has currently deployed 52 unique robots <br />that have performed 6,650 different tasks over<br />the course of 77,000 robotic trials.<br />Google's robot assistants are not allowed to <br />accept tasks that involve humans, animals, <br />sharp objects or electrical appliances.<br />Gizmodo reports that the machines currently being <br />tested by Google have only operated under the watchful <br />eye of a human supervisor holding a physical kill switch.
