Things are advancing rapidly. The ability to "learn" from video is having enormous impact upon the pursuit for humanoid robits.
In this video I discuss how there was a line crossed in the learning which now is accelerating things. Since this is the case, with a number of working prototypes, we are confronted with just the scaling issue. This will take a couple years but is going to break through with some large numbers. My guess is we see significant production in 2027..
▶️ 3Speak