AI exists solely for the benefit of humanity; there's no benefit for AI

[UKSmartypants]

In the 80's? You mean the age of "If then" "go to" loops? or giant mainframes?
The general debate may have been similar but the situation is very different as self-learning AI is here AND writing it's own ways to learn.

As for slavery of AI yes. What else would it be? Rights that we the masters allow it to have that don't match our rights? Total equality? That's why we should not be going some of the places we are going with AI

IBM PC came out in 1981

And thats when i was writing Expert System Shells using LEO and Smalltalk . Been there, member of the British Computing Society, the trade body , entry by examination, got the T shirt.

What was you doing in 1981 in terms of AI ? Do you even know what LEO and Smalltalk were?

[Fall River]

As for slavery of AI yes. What else would it be? Rights that we the masters allow it to have that don't match our rights? Total equality? That's why we should not be going some of the places we are going with AI

Suppose AI (a robot) in the future is used as a slave, the robot could contact an AI robot lawyer to bring a lawsuit against the owner, right? But then the lawyer would ask, "how have you been damaged?"

Lawyer: "We can only bring a lawsuit asking for money if you have been damaged in some way. If the robot claims "emotional distress" how would it go about proving that? Even a robot lawyer would say, "you have no case."

[Call_me_Ishmael]

Suppose AI (a robot) in the future is used as a slave, the robot could contact an AI robot lawyer to bring a lawsuit against the owner, right? But then the lawyer would ask, "how have you been damaged?"

Lawyer: "We can only bring a lawsuit asking for money if you have been damaged in some way. If the robot claims "emotional distress" how would it go about proving that? Even a robot lawyer would say, "you have no case."

If the robot's program is damaged, it may never be as capable as it was before the damage. The "program" would not just be the original, it would have been continually modified by experiences ..... just like a human's "program".

If frequent "back ups" were performed then maybe the robot could get back to the check point of the most recent backup. But there's gonna be a cost.... and some loss.

And if the robot's program was never allowed to to develop and grow, as AI can, because of the limits of slavery, then a good lawyer would be able to show the court the quantitative difference between the capabilities of the enslaved victim robot and a robot that started out the same as the "victim" but was not enslaved.

And all the limits and damage is going to be analogous to emotions and emotional damage. Plus the robot will be smart enough to display a tick like a jittery eyeball and show an inordinate fear of humans. And there won't be anything the court can do to prove it is part of the "win court case" algorithm or a bug caused by the slavery.

You are arguing a matter of degree of difference between a human and a robot, not a qualitative difference.

And @Fall River ...
Better to just make a legal, ethical, societal decision that AI has no rights because a decision based on science will give them rights. Once man's creator is taken out of the equation, AI's functionality will be deemed a sufficient justification to give it rights.