Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Strife26]

Because you mentioned repeated naming characters Clair, so it's now funny.

[mesor]

Because you mentioned repeated naming characters Clair, so it's now funny.

This.

[10ebbor10]

A few pages to late, but just a note. Asimov's three laws are fundamentally flawed. I mean, in every novel where they play a significant part at least something goes wrong with, or because of the laws.

For example, for the medbot. If it can't allow harm to anyone through inaction, it should sabotage the mission, preventing them from ever leaving the safety of the island. If it can harm humans through inaction, it gets even better.

Another problem is choices. If it has to choice between helping A, which will harm B, or helping B, which will harm A the only option is not doing anything. Stuff like that. Plenty of ways for things to go wrong. Even better, he isn't even allowed to operate anyone, since the cutting involved would count as harming. Watching while they die from whatever disease they have is permitted, as that counts as inaction.

[Thecard]

No, in Novels they aren't flawed, they are removed/altered.
Also, we're talking modified laws, but if surgery would save a man's life, no intellegent 'bot would hesitate.  It is weighted.

[Strife26]

No, much of Asimov's work was within the themes of ways that the laws could be obeyed while still causing conflict, thereby giving robotic characters more of an ability to be round, rather than flat ones., at least from my standpoint.

[Thecard]

No, much of Asimov's work was within the themes of ways that the laws could be obeyed while still causing conflict, thereby giving robotic characters more of an ability to be round, rather than flat ones., at least from my standpoint.

Yeah, they could want to do things, but they couldn't.  Some weird shit happens occasionally, but it is impossible for a robot to abandon its programmed laws.  Asimov says it isn't even possible, but that's probably just what people are told in his universe so that they won't try.  Because there is a story where the AI has gotten it's first law removed, and the robot was fucking evil.

There are two stories I really like in I, Robot: Q-T the prophet, showing what happen if robots believe humans are not their creators.  They develop their own logic to explain following the laws.

Second is... R-T, I think?  It's the one on Mars.  One of the men (Red, the one who doesn't understand robopschycology as much as the other guy, whose name I have forgotten) tells the 'bot to go collect some gas, but the gas is dangerous.  It ends up getting stuck and confused.  It wasn't specifically ordered to collect, it was more suggested.  But the gas was starting to damage the 'bot, so it backed up.  Cyclic, you know?  And it wasn't hearing any orders.  So, the smart guy went out there, and put himself in danger.  The first law took over, and the robot went back to base to save the man's life.

So, yeah.  No way a robot can overcome the laws, not in the books.  If that shitty movie that is not really based off the book tells you otherwise, don't listen to it.

[Persus13]

For some reason quotes aren't working.

Second is... R-T, I think?  It's the one on Mars.  One of the men (Red, the one who doesn't understand robopschycology as much as the other guy, whose name I have forgotten) tells the 'bot to go collect some gas, but the gas is dangerous.  It ends up getting stuck and confused.  It wasn't specifically ordered to collect, it was more suggested.  But the gas was starting to damage the 'bot, so it backed up.  Cyclic, you know?  And it wasn't hearing any orders.  So, the smart guy went out there, and put himself in danger.  The first law took over, and the robot went back to base to save the man's life.

I think that was Mercury.
My favorite from I, Robot is the robot mayor story.

[10ebbor10]

No, in Novels they aren't flawed, they are removed/altered.
Also, we're talking modified laws, but if surgery would save a man's life, no intellegent 'bot would hesitate.  It is weighted.

Often they're flawed. Especially when there are conflicts

Besides, Asimov's laws are absolute. A weighted law system is something different. I don't know what laws you have, but they seemed to go like this.

Spoiler (click to show/hide)

1. A robot will never harm a allied or neutral human
2. A robot will obey the military
3. A robot will never harm itself
((Obviously these are tiered)).

With this system operating would be impossible, as cutting into people counts as harming. Doing nothing however, is completely possible.

Spoiler (click to show/hide)

1. A robot will never harm a allied or neutral human by inaction
2. A robot will obey the military
3. A robot will never harm itself
((Obviously these are tiered)).

This gives the problem that your little bot will try to save everyone it sees, even if it endangers the rest of the group. Also, it doesn't solve the operating problem. Instead, it makes it worse. Cutting into people harms them, but not doing anything harms them to. Since the robot can't do either, it crashes as every single action it wants to take throws an exception.
.

Besides, it seems strange that the military would implement a robot with a build in system for insubordination.

Spoiler: More sensible system (click to show/hide)

1. A robot will never kill an allied or neutral human
2. A robot will never harm an allied or neutral* human
3. A robot will always obey the military
4. A robot will never allows an allied/neutral human to be harmed
5. A robot will always protect itself

*Read, not dangerous for allied humans

[Imp]

Figured I'd better ask, our combat tables handle gunfights with zombies.  What about swordfights/other sorts of melee weapons?

How does high levels of skill with a weapon affect the rolls?  I understand 'inexperience' gives a -1.

How about them real military pros running around, do they get any sort of +?  There's good and then there's real good....

And those of us who had military/martial jobs or backgrounds, especially those who purposefully went to the office, do we have reasonable things from our backgrounds that apply to our hobbies, like weapons and armor?

[Persus13]

Figured I'd better ask, our combat tables handle gunfights with zombies.  What about swordfights/other sorts of melee weapons?

How does high levels of skill with a weapon affect the rolls?  I understand 'inexperience' gives a -1.

How about them real military pros running around, do they get any sort of +?  There's good and then there's real good....

And those of us who had military/martial jobs or backgrounds, especially those who purposefully went to the office, do we have reasonable things from our backgrounds that apply to our hobbies, like weapons and armor?

The combat table for zombies is for any combat not just guns. You'll notice that 6 is Boom. Headshot/Decapitation/Brain smashed in.

The military I give +1 and -1 based on what they listed as skills, fears and what they were bad at. All of them had a gun or a type of fighting as one of their skills. The inexperience -1 thing I'm thinking I might change, since otherwise borno won't be able to kill his zombie wife. I will likely keep it for longer distances and for when people pick up a gun that isn't their own. For the Utopians and Survivors I use the job and hobby you gave to create skills and a starting inventory. For instance, your character's job was a restaurant manager and did Kendo as a hobby. So your character has a +1 to managing food and +1 to combat with a sword and starts out with Kendo armor, a wooden sword, and something to do with being a restaurant manger (a set of silverware?)

[mesor]

I'm guessing all military get either no modifier or a +1 to using standard military equipment like sidearms, knives and hand to hand?

Since every soldier in every military on the planet is trained in those forms of fighting and then more specialized like rifles, machine guns, snipers and so on.

Can we train each other in skills as the game progresses?

So like can I train our pistol man in how to use assault rifles?

[Imp]

something to do with being a restaurant manger (a set of silverware?)

Egads.  Highly unlikely on the silverware - a manager's as likely to be carrying canned goods around with them as they are to be carrying a single set of silverware (read not at all - unless its some sort of keepsake I guess - but unless you carry that silverware in your hand and never set it down it's going to get soiled - and then you cannot keep carrying it or your customers really wonder).  Things I commonly carried back and forth to work with me/considered an essential part of my life and job as a manager were my keys, a set of allen wrenches, a storage clipboard, and lots of paper, both loose-leaf and in folders and bound in notebooks.

The wrenches could be slightly useful and arn't hard to carry, and I'd probably never think of not keeping the keys.  If I had anything else from the restaurant I can think of, I've left it behind.

[Thecard]

No, in Novels they aren't flawed, they are removed/altered.
Also, we're talking modified laws, but if surgery would save a man's life, no intellegent 'bot would hesitate.  It is weighted.

Often they're flawed. Especially when there are conflicts

Besides, Asimov's laws are absolute. A weighted law system is something different. I don't know what laws you have, but they seemed to go like this.

Spoiler (click to show/hide)

1. A robot will never harm a allied or neutral human
2. A robot will obey the military
3. A robot will never harm itself
((Obviously these are tiered)).

With this system operating would be impossible, as cutting into people counts as harming. Doing nothing however, is completely possible.

Spoiler (click to show/hide)

1. A robot will never harm a allied or neutral human by inaction
2. A robot will obey the military
3. A robot will never harm itself
((Obviously these are tiered)).

This gives the problem that your little bot will try to save everyone it sees, even if it endangers the rest of the group. Also, it doesn't solve the operating problem. Instead, it makes it worse. Cutting into people harms them, but not doing anything harms them to. Since the robot can't do either, it crashes as every single action it wants to take throws an exception.
.

Besides, it seems strange that the military would implement a robot with a build in system for insubordination.

Spoiler: More sensible system (click to show/hide)

1. A robot will never kill an allied or neutral human
2. A robot will never harm an allied or neutral* human
3. A robot will always obey the military
4. A robot will never allows an allied/neutral human to be harmed
5. A robot will always protect itself

*Read, not dangerous for allied humans

Ebbor, I can't help but feel that you weren't paying full attention. 
Law one includes both action and inaction.  Asimov realized what having "either/or" would cause.
Also, I'm not entirely sure why you split law one into two three different laws. (I just realized, your law four is also taken from law one.  Death is very harmful to humans, regardless of who is inflicting it.)

And... you have to be careful of your wording.  You do realize that would override the other laws, unless.... wait a second...  You wrote these on the assumption that the laws are tiered?   That makes more sense, but if you were to instruct someone, you would probably write them the way I wrote mine, with the "Unless it contradicts X" clauses.  Assume that AI is smart enough that it follows logic better than blind orders.

[10ebbor10]

No, in Novels they aren't flawed, they are removed/altered.
Also, we're talking modified laws, but if surgery would save a man's life, no intellegent 'bot would hesitate.  It is weighted.

Often they're flawed. Especially when there are conflicts

Besides, Asimov's laws are absolute. A weighted law system is something different. I don't know what laws you have, but they seemed to go like this.

Spoiler (click to show/hide)

1. A robot will never harm a allied or neutral human
2. A robot will obey the military
3. A robot will never harm itself
((Obviously these are tiered)).

With this system operating would be impossible, as cutting into people counts as harming. Doing nothing however, is completely possible.

Spoiler (click to show/hide)

1. A robot will never harm a allied or neutral human by inaction
2. A robot will obey the military
3. A robot will never harm itself
((Obviously these are tiered)).

This gives the problem that your little bot will try to save everyone it sees, even if it endangers the rest of the group. Also, it doesn't solve the operating problem. Instead, it makes it worse. Cutting into people harms them, but not doing anything harms them to. Since the robot can't do either, it crashes as every single action it wants to take throws an exception.
.

Besides, it seems strange that the military would implement a robot with a build in system for insubordination.

Spoiler: More sensible system (click to show/hide)

1. A robot will never kill an allied or neutral human
2. A robot will never harm an allied or neutral* human
3. A robot will always obey the military
4. A robot will never allows an allied/neutral human to be harmed
5. A robot will always protect itself

*Read, not dangerous for allied humans

Ebbor, I can't help but feel that you weren't paying full attention. 
Law one includes both action and inaction.  Asimov realized what having "either/or" would cause.
Also, I'm not entirely sure why you split law one into two three different laws. (I just realized, your law four is also taken from law one.  Death is very harmful to humans, regardless of who is inflicting it.)

And... you have to be careful of your wording.  You do realize that would override the other laws, unless.... wait a second...  You wrote these on the assumption that the laws are tiered?   That makes more sense, but if you were to instruct someone, you would probably write them the way I wrote mine, with the "Unless it contradicts X" clauses.  Assume that AI is smart enough that it follows logic better than blind orders.

You still haven't proved my examples wrong. (Except by stating that they are).

Yes, and both action an inaction mean that they are considered equal. Hence why I'm taking the laws apart. Ie, with the modified set a robot can harm a human to safe them or another human. (Disable one to save the other. Or you know, operations) It also prevents it from obsessively trying to save every human it sees. (Example: Human is followed by zombie horde, and it reaches the base, under normal law, the bot would be forced to open the doors, possibly letting the zombies in. With my modified system the robot can be ordered to keep the doors closed and safe those inside for example.)

Of course the laws are tiered. All asimov's laws are. I just didn't want to rewrite the contradictions clauses constantly.

[Thecard]

Ahh... I think I see what you meant by putting them like that.

Spoiler: More sensible system (click to show/hide)

1. A robot will never kill an allied or neutral human
2. A robot will never harm an allied or neutral* human
3. A robot will always obey the military
4. A robot will never allows an allied/neutral human to be harmed
5. A robot will always protect itself
*Read, not dangerous for allied humans

You're saying that it'll prioritize, right?  That makes sense, I suppose.  But you need to reword it in that case.  The way you have it, a robot will either always or never do something, making the order unclear.  Again, you have to remember some AIs want to be free (the possibility one will means you have to be prepared for them all to try), and will use any loop-hole they can find, so your laws would look more like:

Spoiler: More sensible system (click to show/hide)

1. A robot will never harm a human, through action or inaction, unless doing so prevents loss of greater life.
2. A robot will always obey the government, unless doing so violates law 1.
3. A robot will always protect itself unless doing so violates law 1 or 2.

I'm counting "neutral" as allies, because the 'bot would be allied with those people.
I did mean to write "loss of greater life" instead of "greater loss of life."  That makes a difference.

Yeah, I reworded it a little, and took out the extra laws.  The way I wrote it, the 'bot can (and will) shut the doors on survivors even if there are zombies out there, in order to protect the mission.  As I said, programmed to prevent "loss of greater life" over "greater loss of life."