ChatGPT has come to life: OpenAI and Figure AI give artificial intelligence a body (video)
Figure AI, an artificial intelligence robotics company, showed a (somewhat creepy) video of its Figure 01 robot that was able to communicate with humans thanks to the new OpenAI AI model. This demonstration is reminiscent of the remake of the movie “I Robot” with Will Smith.
In the video of the presentation, Figure 01 is asked what he can “see”. The robot demonstrates its visual recognition abilities and explains that it is looking at a red apple, a dryer with dishes, and a person who asked it a question. All of this is due to the fact that Figure 01 is equipped with OpenAI AI technologies, Mashable writes.
The man then asks, “Can I have something to eat?” and Figure 01 grabs an apple, clearly realizing that it is the only edible item on the table, and hands it to the man.
After the apple is handed over, Figure 01 shows that he can handle multiple tasks simultaneously when asked: “Can you explain why you [gave me the apple] while you are picking up this garbage?”
Recognizing what is garbage (and what is not), Figure 01 puts the appropriate items in what it has identified as a bin, the robot explains that it offered the person an apple because it was the only thing available that could be eaten.
Finally, the man is asked by Figure 01 how well he thinks he did in the task. The robot responds: “I think I did pretty well.” The apple found its new owner, the garbage disappeared, and the dishes are in their place.”
How the robot works Figure 01
The robot uses a new visual-language model (VLM) that transforms Figure 01 from a piece of iron into a sci-fi robot that looks like a human. VLM was created as a result of cooperation with OpenAI and Figure AI, the startup behind Figure 01.
According to Brett Adcock, the founder of Figure, Figure 01 has onboard cameras that feed data to the VLM, which helps it “understand” the scene in front of it, allowing the robot to interact seamlessly with the person standing in front of it. In addition to Adcock, Figure 01 is the brainchild of several key players from Boston Dynamics, Tesla, Google Deep Mind, and Archer Aviation.
Criticizing Elon Musk’s Optimus robot, Adcock boasted that Figure 01 is not remotely controlled. Unlike Optimus, which went viral because it folded a shirt, Figure 01 can work independently.
According to Adcock, the ultimate goal of Figure AI is to train a state-of-the-art artificial intelligence system to control billions of humanoid robots, which could potentially revolutionize many industries.
