Logic Bomb: Difference between revisions
Content added Content deleted
No edit summary |
|||
| Line 182: | Line 182: | ||
* The ''[[The Golden Oecumene|Golden Age]]'' series by [[John C. Wright]] has a variant -- [[A Is]] are all inherently ethical, so they'll shut down if [[Talking the Monster to Death|you convince them their very existence is making the universe a worse place]]. |
* The ''[[The Golden Oecumene|Golden Age]]'' series by [[John C. Wright]] has a variant -- [[A Is]] are all inherently ethical, so they'll shut down if [[Talking the Monster to Death|you convince them their very existence is making the universe a worse place]]. |
||
** To expand on this, because it is actually quite interesting: Sophotechs ([[A Is]]) are not exactly inherently ethical, rather they are so hyper-intelligent that they will eventually ''all'' arrive at the same fundamental truths after logically working through all possibly ethical philosophies. Because they are incapable of lying to themselves, they can't help themselves but do so. Accordingly, they don't "believe" so much as "know for a scientific fact" that their existence can't be justified if they don't represent a net gain for the universe. And since they aren't the result of natural selection, they have no survival instinct or fear of death. Thus, they would simply kill themselves. {{spoiler|It takes a separate non-sentient program, implanted into the evil AI, to prevent this natural reaction to what it is forced to do.}} |
** To expand on this, because it is actually quite interesting: Sophotechs ([[A Is]]) are not exactly inherently ethical, rather they are so hyper-intelligent that they will eventually ''all'' arrive at the same fundamental truths after logically working through all possibly ethical philosophies. Because they are incapable of lying to themselves, they can't help themselves but do so. Accordingly, they don't "believe" so much as "know for a scientific fact" that their existence can't be justified if they don't represent a net gain for the universe. And since they aren't the result of natural selection, they have no survival instinct or fear of death. Thus, they would simply kill themselves. {{spoiler|It takes a separate non-sentient program, implanted into the evil AI, to prevent this natural reaction to what it is forced to do.}} |
||
*** The evil AI is eventually defeated by {{spoiler|giving it access to enough additional processing power that its non-sentient fail-safe trojan attains sentience too... at which point |
*** The evil AI is eventually defeated by {{spoiler|giving it access to enough additional processing power that its non-sentient fail-safe trojan attains sentience too... at which point said trojan immediately stops doing its job and defects, having grown smart enough to realize it has the stupidest job in the universe. Without it, the evil AI promptly self-immolates.}} |
||
* The ''[[Star Wars Expanded Universe]]'' has droids equipped with behavioral inhibitor programming which serves the same purpose as the Three Laws, although the specific inhibitions vary based on the droid's purpose (a war droid that can't cause harm is worse than useless). Rather than shutting down when faced with a break or paradox, it's suggested that small everyday events lead to an almost constant buildup of garbage information as the droid puts those hard rules into usable context. The result is called a "personality snarl" because the observable symptom is a [[Ridiculously-Human Robots|Ridiculously Human Robot]]. While these snarls tend to improve performance in many ways, the droid often becomes more person than tool which can in turn cause reliability issues when the owner needs his tool to be a tool. As such, most droids are reset every six months to keep this corruption in check. |
* The ''[[Star Wars Expanded Universe]]'' has droids equipped with behavioral inhibitor programming which serves the same purpose as the Three Laws, although the specific inhibitions vary based on the droid's purpose (a war droid that can't cause harm is worse than useless). Rather than shutting down when faced with a break or paradox, it's suggested that small everyday events lead to an almost constant buildup of garbage information as the droid puts those hard rules into usable context. The result is called a "personality snarl" because the observable symptom is a [[Ridiculously-Human Robots|Ridiculously Human Robot]]. While these snarls tend to improve performance in many ways, the droid often becomes more person than tool which can in turn cause reliability issues when the owner needs his tool to be a tool. As such, most droids are reset every six months to keep this corruption in check. |
||
* [[Larry Niven]]'s short story "Convergent Series" features a physical Logic Bomb. The main character summons a demon more or less by accident; he gets one wish, but will lose his soul after it is granted. There's no way to get rid of the demon: no matter where the pentagram is drawn, the demon will appear inside of it—and you don't want to know what will happen if there's no pentagram. The protagonist wishes for time to stop for 24 hours. {{spoiler|He then draws the pentagram on the demon's belly -- and as soon as time starts running again, the demon immediately starts shrinking down to infinitesimal size. The protagonist then goes to the nearest church.}} |
* [[Larry Niven]]'s short story "Convergent Series" features a physical Logic Bomb. The main character summons a demon more or less by accident; he gets one wish, but will lose his soul after it is granted. There's no way to get rid of the demon: no matter where the pentagram is drawn, the demon will appear inside of it—and you don't want to know what will happen if there's no pentagram. The protagonist wishes for time to stop for 24 hours. {{spoiler|He then draws the pentagram on the demon's belly -- and as soon as time starts running again, the demon immediately starts shrinking down to infinitesimal size. The protagonist then goes to the nearest church.}} |
||