Imagine what it would be like to have a robot friend that can do things like take selfies, toss a ball, eat popcorn and play air guitar? Well, you might not have to wait too long. Researchers at the University of Tokyo have created a robot that can do all that and more, thanks to the power of GPT-4, the latest and most advanced large language model (LLM) in the world. Alter3 is a humanoid robot that was first introduced in 2016 as a platform for exploring the concept of life in artificial systems. It has a realistic appearance and can move its upper body, head and facial muscles with 43 axes controlled by air actuators. It also has a camera in each eye that allows it to see and interact with humans and the environment. But what makes Alter3 really special is that it can now use GPT-4, a deep learning model that can generate natural language texts from any given prompt, to control its movements and behaviors. This means that instead of having to program every single action for the robot, the researchers can simply give it verbal instructions and let GPT-4 generate the corresponding Python code that runs the Android engine. For example, to make Alter3 take a selfie, the researchers can say something like: “Create a big, joyful smile and widen your eyes to show excitement. Swiftly turn the upper body slightly to the left, adopting a dynamic posture. Raise the right hand high, simulating a phone. Flex the right elbow, bringing the phone closer to the face. Tilt the head slightly to the right, giving a playful vibe.” And GPT-4 will produce the code that makes Alter3 do exactly that.

Full report : Meet the humanoid robot that learns from natural language, mimics human emotions.