Figure's robot/Official site, Figure
The Figure AI company, which develops humanoid robots, unveiled today (Thursday) Figure 01, a robot capable of having full conversations with people.
This impressive breakthrough was made possible thanks to the integration of the artificial intelligence model from OpenAI, which gives Figure 01 unprecedented advanced language capabilities.
In fact, as you can see in the published video, the robot is not satisfied only with short or predefined responses, but is able to conduct fluid and complex conversations, with a deep understanding of nuances and subtleties in the language.
Figure 01 is able to carry a weight of up to 20 kg and move at a speed of 1.2 meters per second (about 4.3 km/h).
These capabilities make it a versatile and functional tool, which can be used for a wide variety of tasks.
Last month, Figure AI raised $675 million, and investors include tech giants such as Microsoft, Open AI, OpenAI's startup fund, NVIDIA, and Jeff Bezos' investment company (Bezos Expeditions).
The company's CEO, Brett Adcock, said that the cameras integrated in Figure 01 send information to a powerful vision-language model developed by OpenAI. At the same time, the robot's independent neural networks are able to "receive images at a rate of 10 images per second." OpenAI is also responsible for The ability to understand speech, and all this information is translated into precise and agile robotic actions by Figure's neural network.
In addition, Adcock emphasized that the robot was not remotely activated during the demonstration, and the video was shot at normal speed. This is an impressive achievement for a collaboration that is less than two weeks old. "Our goal is train a global model that will operate billions of humanoid robots," Adcock said. "At this rate, we won't have to wait long."
More on the same topic:
robot