Line #126 Headscratchers/IRobot


1	!Per wiki policy, Administrivia/SpoilersOff applies here and all spoilers are unmarked. Administrivia/YouHaveBeenWarned.
2
3	[[AC:Stories]]
4	* In "Little Lost Robot", the titular robot hides itself in a group of other robots because its designated supervisor lost his temper and told it--profanely and at length--to get lost. Fine. But why on earth are they addressing this with complicated plans intended to identify the robot, instead of just having the supervisor address the group and tell the lost robot to stop hiding and step forward? Surely robotic programming isn't so clunky that it doesn't allow for giving the robot new orders?
5	** That's exactly the problem. No, I don't recall if Asimov explained why they couldn't give an overriding order, but something about the wording of the order made it so that it wasn't that simple.
6	*** Basically, the profanity and volume of the "get lost" instruction made the robot interpret it as such a high-priority order that there was no instruction they could have given it that would have taken precedence over "get lost".
7	*** ^ Exactly that, combined with the fact that the modified Three Laws had stressed the positronic brain to borderline instability. When the profanity-laden order was given, the robot's mind snapped and it 'lost itself'.
8	*** AND Susan Calvin said something along the lines of robots having resentment for considering themselves better than humans, but having to abide by the 3 Laws anyway. All the ordeal helped strengthen that resentment.
9	* In 'Little Lost Robot', the second test that Calvin comes up with involves dropping a weight towards a human, but with (fake) electrified cables between the robots and the person. However, the robots stay seated, and one of the Nestors points out afterwards that they couldn't save the human anyway, and if they stayed seated they might save a human some time in the future. So, why didn't Calvin arrange a test where the robots would be killed 'after' saving a human? That would have caught out the rogue Nestor, since then a normal robot would be able to save a human (at the cost of its life) and so would do so, but the rogue one wouldn't.
10	** Because it didn't occur to her that the rogue Nestor would think of something like that.
11	* If the stories are set in the same universe as the other Robot novels, then how is it that some of the robots and AIs seem to be more advanced personality-wise that those featured in the other novels, which are set decades and centuries into the future? Cutie, the Brain, and in particular Stephen Byerly appear to demonstrate mannerisms and behavior that is more human-like than say Daneel and Giskard.
12	** Powell and Donovan are testers, and the USR crew get called in to deal with problematic cases hot off the assembly line. They're all dealing with bleeding-edge robots and emergent behavior - Cutie, the first robot built to supervise other robots, and the Brain, USR's latest and greatest experimental AI. Besides, Cutie's behavior isn't really any more advanced than Dave's in "Catch That Rabbit", and the Brain ''only'' thinks, it doesn't have to fit in or run a body. As for Byerly, he's a one-off built explicitly to imitate humanity by what's implied to be an extremely gifted roboticist (and moreover, to imitate his builder specifically).
13	** Also, keep in mind that in "Literature/TheBicentennialMan", a later story in TheVerse, USR starts making their robots deliberately less human so as to not accidentally make another Andrew. That Daneel is of the same general 'level' as Byerly or Andrew isn't surprising in that case.
14	* In Evidence was Stephen Byerlyan actual robot or a person? I know that there was deliberate ambiguity to the story in that regard but I want to know.

Context Headscratchers / IRobot