A computer will be able to think in the same way a submarine can swim.
Thinking is something we've anthropomorphized, and the moment a computer can do something we previously considered 'thinking', we'll move the goalposts again.
Thinking is something we've anthropomorphized, and the moment a computer can do something we previously considered 'thinking', we'll move the goalposts again.
A mechanical computer can "think" just like an electrical one. A series of gears and pulleys can produce a difference of information to the observer. Reducing thought away from the human gives you a recursive spiral down to the bottom of physics. It needs to be embodied in a human or you hit philosophical problems.
Ideology and beliefs aside, I think it's absolutely fine to change one's definition of something when one learns more. I think we all can benefit from better definitions in our thoughts and conversations.
What you may call "moving goalposts" may actually be incremental refinement of an idea -- or a belief that is increasing in sophistication (or absurdity in some cases!) as more is understood.
The authors are certainly moving the goal posts when they concede that 4/5 humans in a vegetative state have no consciousness, but a full simulation of a healthy human mind never should be considered self aware. And yet in the same breath claim this lack of consciousness allows you to destroy a robot in your front yard, but stay startlingly silent on what that means for those same in the vegetative state.