From chatbots for computers and smartphones to robots capable of imitating humans. This could be the next goal of OpenAI, developer of ChatGpt. Caitlin Kalinowskiformer head of Meta’s augmented reality division, and manager of OpenAI since November, shared the future of the company in a post on X. The company plans to create new automatons equipped with sensors that will be powered by internally developed artificial intelligence models. To do this, it has started a process of selecting dedicated figures: “I can’t wait to start this work – Kalinowski’s words on the speed of work and concentration make this a truly wonderful place.”
Specifically, the company is looking for engineers and designers to create “generic”, “adaptive” and “versatile” robots, capable of operating with human-like intelligence in “dynamic” and “real” situations. A field that already sees some protagonists, such as X1 and Figure (both supported by OpenAI), intent on create machines that move and interact with the world. At the end of December, an article on the website The Information identified the critical issues that OpenAI was encountering in the construction of its first robot.
The company is no stranger to the industry, having managed until 2021 a unit dedicated to the development of artificial intelligence models for industrial autonomous machines. In one project, OpenAI researchers trained a robotic arm to solve a Rubik’s cube. The company has also open-sourced Roboschool, a collection of simulation tools that can be used to train AI on board humanoids to coordinate the movement of each body part.