Singularity Hub

The range of movement of this robot is impressive, but when will we see one of these robots doing something constructive, like picking up a pen off the floor? Or a newspaper? When can my robot grab a beer from the fridge and make a grill cheese sandwich? Dancing robots...it's getting old.

When they start using an HTM, perhaps?

HTM Sensors? Well, Dan Berry does suggest that being able to experience the real world may be the best way for AI to develop. That would include the senses.
Oh,wait. Was that a joke?

Well produced. But when talking about Big Dog, I wonder whether the response to the robot being kicked was like that of amazement of its ability to recover, rather than a response of sympathy for the robot over having being kicked.
I do agree however, that it won't really matter whether the robot actually has consciousness, but whether we perceive the robot having consciousness.

It also matters whether the robot can "feel" (record) "pain" (negatively experienced state). In the present case the e-dog is not experiencing discomfort as an animal would.

That being said, I am a big believer in robot rights--if they are sentient. No robot even approaches mammal level sentience today. And it is vital for the future that we do not take the Turing test literally--it DOES matter if the robot has consciousness. For two reasons. 1. instrumentally for not all situations can be anticipated, so a bot with intelligent behavior and ability to learn would be more useful than a rule driven bot, and 2. ethically, as artificial sentient beings would need to be treated in a way that does not inflict the equivalent of "pain', not merely because the robot appears undisturbed.

So sympathy in this case would be more of an automatic empathy response, than a rational one. In fact the robot of sufficient complexity could be programed to experience "pleasure" (positively experienced state) to take righting itself as a challenge, much like my ferret enjoys being tossed on the bed.... Read More

To properly empathize with an intelligent robot you need to translate the internal state of a robot to one of your own internal states (emotions) and ask yourself what your reaction would be. The golden rule would be do unto robots as you would wish to be done unto you --if you were a robot.