Figure, a Bay Area robotics firm, has announced a new machine learning model called Helix, which allows its humanoid robot to take voice orders and perform tasks in the home environment. The model is a Vision-Language-Action (VLA) model, which leverages vision and language commands to process information. Helix is designed to bridge the gap between vision and language processing, enabling the robot to visually assess its environment and perform tasks based on natural language voice prompts. The company has showcased the capabilities of Helix by demonstrating its humanoid robot, 02, performing various household tasks, such as picking up objects and placing them in a designated location. Figure's goal is to make robots more useful in households by enabling them to generate intelligent new behaviors on-demand, especially for objects they've never seen before.