Calling this “self-aware” is misleading. The robot has body-state awareness, not introspective awareness. And calling it “learning like humans” exaggerates what is mostly sensorimotor adaptation rather than psychologically rich learning shaped by motives, memory, and lived experience. My main problem here is the definition of the word "self-aware". That, according to my understanding, means being able to comtemplate oneself, one's existance, one's purpose, to me, self aware means something that can contemplate itself, not just correct its position based on what a sensor is telling it, not just navigate through without destructing itself, but to be able to think in a nuanced way about itself and the consequences related to that. So "self-aware", I call bullshit. These are just bots that are starting to find ways to not break themselves while doing one action they are programmed to do. Think of humans, we arent programmed to do anyhting, we learn progressively, through years, through lived experinece, these bots that "imitate" humans are just copying actions, if theyre programmed to, they will copy anything they see, good actions and bad ones. Human psych doesnt work the same way, we see someone do something, we dont mindlessly copy it, we go thorugh the stages of the social learning theory, attention, retention, motivation and finally replication, if we dont see motivation to imitate, we simply dont, we take factors like our own past experineces, our current surroundings, everything, and then learn, that is what learning is for us. For the robot, "learning" is just little more than motor replication, no nuance there.