Last week, Tesla held a We, Robot event to showcase its new self-driving cars Cybercab and Robovan. The company also paid considerable attention to its Optimus humanoid robots, which were a significant part of the show.
During the event, the robots mingled with the crowd, served drinks and played with guests, and danced in the gazebo. The most amazing thing is that they could even talk. But for the most part, it was just a show. You can see what it was like by watching various videos from the We, Robot presentation.
So far, however, Optimus robots have not yet reached a level of autonomy that allows them to respond to verbal and visual cues while talking face-to-face with people in dimly lit crowds. Event participant Robert Scoble learned that people «remotely assist» robots. He later clarified that, according to the engineer, robots use AI to walk. Morgan Stanley analyst Adam Jonas wrote that the «robots relied on teleoperation (human intervention) of the» robots.
There is obvious evidence to support these claims, such as the fact that all the robots have different voices or that their responses were instantaneous, with appropriate gestures.
There is no sense that Tesla was doing anything to make anyone think that the Optimus cars were acting on their own. In another video, in response to a question about how much the robot is controlled by AI, the Optimus voice jokingly said that «can be a little bit».
Another robot — or a human voice — said to the participant in a disappointing synthetic voice: «Today I have a person helping me», adding that he is not fully autonomous (the voice stumbled over the word «autonomous»).
Musk first announced Tesla’s humanoid robot by bringing a man in a robot suit on stage. So it is not surprising that Optimus at last week’s event was hyperbolic in its presentation. And the people who left didn’t seem upset or betrayed by it. However, it’s clear that the We, Robot event failed to give a sense of how far Tesla has really come in its work with humanoid robotics.
Source: The Verge