High Robotics · 1 min read
Figure 01 + OpenAI: first end-to-end LLM-driven humanoid demo
In one sentence Figure publishes a video of its Figure 01 humanoid conversing, recognizing objects, and manipulating them using OpenAI models for language and vision, in an end-to-end pipeline.
Reading level
Figure, a humanoid-robot startup, shows an impressive video: their robot Figure 01 stands at a table; a person speaks to it, and the robot replies aloud and picks up objects on request.
"Give me something to eat," the person says. The robot looks at the table, picks an apple (the only food there), and hands it over. Then it clears dirty dishes into a bin, all while replying verbally.
All in near-real time, no teleoperation. The brain is an OpenAI multimodal model that sees, understands, and decides.
Companies
Figure, OpenAI
Tools
Figure 01
Tags
FigureOpenAIHumanoidEmbodied AIVLM
Sources