5 ms·

Put it this way: if we ever do create AIs that can (for want of a better yardstick) pass the Turing Test reliably, someone somewhere will stick them in a comput

Put it this way: if we ever do create AIs that can (for want of a better yardstick) pass the Turing Test reliably, someone somewhere will stick them in a computer game for people to shoot at them. Sit and think about the philosophical implications of that for a minute.

castellar 9y ago
The Turing test isn't really a good indicator of AI development. I don't disagree with the idea that using sufficiently advanced AI for entertainment could create ethical quandaries though.
- codeulike 9y ago
  When people imagine what a Turing Test conversation would look like, they frequently underestimate the conversation. I find Dennet's example of an imaginary Turing Test from Consciousness Explained to be a good counterexample: Judge: Did you hear about the Irishman who found a magic lamp? When he rubbed it a genie appeared and granted him three wishes. “I’ll have a pint of Guiness!” the Irishman replied and immediately it appeared. The Irishman eagerly set to sipping and then gulping, but the level of Guiness in the glass was always magically restored. After a while the genie became impatient. “Well, what about your second wish?” he asked. Replied the Irishman between gulps, “Oh well, I guess I’ll have another one of these.” CHINESE ROOM: Very funny. No, I hadn’t heard it– but you know I find ethnic jokes in bad taste. I laughed in spite of myself, but really, I think you should find other topics for us to discuss. J: Fair enough but I told you the joke because I want you to explain it to me. CR: Boring! You should never explain jokes. J: Nevertheless, this is my test question. Can you explain to me how and why the joke “works”? CR: If you insist. You see, it depends on the assumption that the magically refilling glass will go on refilling forever, so the Irishman has all the stout he can ever drink. So he hardly has a reason for wanting a duplicate but he is so stupid (that’s the part I object to) or so besotted by the alcohol that he doesn’t recognize this, and so, unthinkingly endorsing his delight with his first wish come true, he asks for seconds. These background assumptions aren’t true, of course, but just part of the ambient lore of joke-telling, in which we suspend our disbelief in magic and so forth. By the way we could imagine a somewhat labored continuation in which the Irishman turned out to be “right” in his second wish after all, perhaps he’s planning to throw a big party and one glass won’t refill fast enough to satisfy all his thirsty guests (and it’s no use saving it up in advance– we all know how stale stout loses its taste). We tend not to think of such complications which is part of the explanation of why jokes work. Is that enough? Dennett: "The fact is that any program that could actually hold up its end in the conversation depicted would have to be an extraordinary supple, sophisticated, and multilayered system, brimming with “world knowledge” and meta-knowledge and meta-meta-knowledge about its own responses, the likely responses of its interlocutor, and much, much more…. Maybe the billions of actions of all those highly structured parts produce genuine understanding in the system after all."
  - Retric 9y ago
    The problem is not the quality of interaction, it's a question of what else an AI needs to pass the test. Living organisms have evolved to go about the process of life and avoid pointless death etc, but Thinking and Self preservation are separate. Even people chose to die for various reasons, so AI could be really strange.
    - wst_ 9y ago
      This is exactly what I've been thinking. Above mentioned conversation is so rich that a lot of humans couldn't follow it in such a manner. And yet, there wouldn't be any doubts when talking with them face to face. Is this the right test after all? Or rather the question is, what are we testing actually? It seems to be nothing more than a mimicking human behavior. In this case - very smart and eloquent human. And this is exactly what Turing test is[1]. But does AI need to be able to follow such sophisticated dialogue to be considered an independent, self aware being? [1] Turing: "Are there imaginable digital computers which would do well in the imitation game?"
CM30 9y ago
Eh, I see where you're coming from, but I just consider fiction to be seperate from reality by category. It's like the difference between you imagining you're committing a crime and hurting a bunch of people and you actually committing a crime and hurting a bunch of people. Those in the dream might have arguably better AI than any video game character could hope to have and be as realistic as possible. But they're also not real. They don't matter and simply cannot matter. Same goes for virtual acts towards real people's avatars. If they're within the rules of the game and what the creators of it intended, then they're not against the law. Because you haven't really done anything to them. Otherwise you may as well start saying it's wrong to kill players in a MMORPG setting or what not. Then again, I'm also someone who thinks fictional worlds should be treated as if they were completely seperate universes to the 'real' one in general. Murder, theft, assault, terrorism... right down to copyright infringement shouldn't be banned in a fictional world unless that world's setup explicitly sets laws against those things. We have to keep fiction and reality seperate.
- codeulike 9y ago
  I don't think I'm talking about fiction. OK, how about looking at it this way: At some point game designers will need to start simulating things like 'fear' to make NPCs more realistic. Then imagine the simulation getting better and better and better. Its easy to imagine them simulating a whole complex range of feelings and responses, simulating an inner-world for the NPCs. Then apply moore's law and let the simulations exponentially improve in accuracy and fidelity year on year. At what point would simulations of fear and terror in NPCs start to make you feel uneasy? Unless you believe in dualism or some sort of bio-centrism that means machines can never be conscious, I think there are interesting ethical issues there.
  - Zarath 9y ago
    Can't believe you are receiving downvotes my friend. It seems very likely that consciousness is an emergent property of a sufficiently complex system. Given that this is the case, if we have a computational system such that it sufficiently mimics all nonphysical human behavior, it's very tough to argue that it does not share consciousness, especially if you are not a solipsist.
NeutronBoy 9y ago
Is there a difference between shooting at a character controlled by an AI, and shooting at a character controlled by another human?
- codeulike 9y ago
  Yes of course. I'm hypothesising character AIs that don't know they are in a game. From their point of view, it's all 'real'. They might get built that way to maximise the realism of their behaviour.
StavrosK 9y ago
You think shooting at them is bad? What about quitting the game?
- wst_ 9y ago
  I think it depends on how much can you get yourself into the world and story. When you play CS/Quake death match the goal is simple, there's no second thought before pulling the trigger - there's not story to begin with, just frags/scores. But when you are playing some good RPG and come to some village, you could start the killing spree, see people around you crying, bleeding, crawling at your feet and begging for their lives... would you have any thoughts before pulling the trigger then? My level of immersion usually is quite high if I like a game. Even today. Despite crud graphics and AI behavior. I'll use an imagination to tell the missing part of the story or give NPCs goals and desires. Simply because if feels better. By analogy, I would not kill innocent because that feels bad to me. So, if the day will come when all this missing part will get filled by brilliant AI I am not going to kill it either. Probably the reason that people have no quarrels with killing everyone in a game is that it's very difficult to convince player that NPCs have a feelings. Even in modern AAA titles when you start shooting in the middle of the city it looks more ludicrous than real. People are frightened, sure, but very often they make really fake poses, run randomly with hands up, or do some other stupid things. It's hardly convincing, so shooting them is easy and players have no remorse about it. You can easily see they are fake. But should they behave really human-like, should you see their pain... would you still pull the trigger?
  - StavrosK 9y ago
    I mean that, if they're living, quitting the game is literally murder. Keep in mind that an intelligent AI wouldn't be hurt when you shot an NPC any more than a show writer would be hurt when a character got shot in the show. There wouldn't be one AI that felt pain per NPC, it would be a general orchestrator deciding which NPC to make cry and how they should act.
wst_ 9y ago
If they would be so advanced as you say, they would probably be conscious enough to be able to play any game asked. If that's the case, we could just create them outside of a game and ask them to play against humans. They would, sort of, role play as NPC or as an adversary. That would solve the problem of acting violently against them. They wouldn't get hurt, since they wouldn't be the suffering characters per se. They would just play the same game and understood the implications.

castellar 9y ago

The Turing test isn't really a good indicator of AI development. I don't disagree with the idea that using sufficiently advanced AI for entertainment could create ethical quandaries though.

codeulike 9y ago
When people imagine what a Turing Test conversation would look like, they frequently underestimate the conversation. I find Dennet's example of an imaginary Turing Test from Consciousness Explained to be a good counterexample: Judge: Did you hear about the Irishman who found a magic lamp? When he rubbed it a genie appeared and granted him three wishes. “I’ll have a pint of Guiness!” the Irishman replied and immediately it appeared. The Irishman eagerly set to sipping and then gulping, but the level of Guiness in the glass was always magically restored. After a while the genie became impatient. “Well, what about your second wish?” he asked. Replied the Irishman between gulps, “Oh well, I guess I’ll have another one of these.” CHINESE ROOM: Very funny. No, I hadn’t heard it– but you know I find ethnic jokes in bad taste. I laughed in spite of myself, but really, I think you should find other topics for us to discuss. J: Fair enough but I told you the joke because I want you to explain it to me. CR: Boring! You should never explain jokes. J: Nevertheless, this is my test question. Can you explain to me how and why the joke “works”? CR: If you insist. You see, it depends on the assumption that the magically refilling glass will go on refilling forever, so the Irishman has all the stout he can ever drink. So he hardly has a reason for wanting a duplicate but he is so stupid (that’s the part I object to) or so besotted by the alcohol that he doesn’t recognize this, and so, unthinkingly endorsing his delight with his first wish come true, he asks for seconds. These background assumptions aren’t true, of course, but just part of the ambient lore of joke-telling, in which we suspend our disbelief in magic and so forth. By the way we could imagine a somewhat labored continuation in which the Irishman turned out to be “right” in his second wish after all, perhaps he’s planning to throw a big party and one glass won’t refill fast enough to satisfy all his thirsty guests (and it’s no use saving it up in advance– we all know how stale stout loses its taste). We tend not to think of such complications which is part of the explanation of why jokes work. Is that enough? Dennett: "The fact is that any program that could actually hold up its end in the conversation depicted would have to be an extraordinary supple, sophisticated, and multilayered system, brimming with “world knowledge” and meta-knowledge and meta-meta-knowledge about its own responses, the likely responses of its interlocutor, and much, much more…. Maybe the billions of actions of all those highly structured parts produce genuine understanding in the system after all."
- Retric 9y ago
  The problem is not the quality of interaction, it's a question of what else an AI needs to pass the test. Living organisms have evolved to go about the process of life and avoid pointless death etc, but Thinking and Self preservation are separate. Even people chose to die for various reasons, so AI could be really strange.
  - wst_ 9y ago
    This is exactly what I've been thinking. Above mentioned conversation is so rich that a lot of humans couldn't follow it in such a manner. And yet, there wouldn't be any doubts when talking with them face to face. Is this the right test after all? Or rather the question is, what are we testing actually? It seems to be nothing more than a mimicking human behavior. In this case - very smart and eloquent human. And this is exactly what Turing test is[1]. But does AI need to be able to follow such sophisticated dialogue to be considered an independent, self aware being? [1] Turing: "Are there imaginable digital computers which would do well in the imitation game?"

CM30 9y ago

Eh, I see where you're coming from, but I just consider fiction to be seperate from reality by category. It's like the difference between you imagining you're committing a crime and hurting a bunch of people and you actually committing a crime and hurting a bunch of people. Those in the dream might have arguably better AI than any video game character could hope to have and be as realistic as possible. But they're also not real. They don't matter and simply cannot matter. Same goes for virtual acts towards real people's avatars. If they're within the rules of the game and what the creators of it intended, then they're not against the law. Because you haven't really done anything to them. Otherwise you may as well start saying it's wrong to kill players in a MMORPG setting or what not. Then again, I'm also someone who thinks fictional worlds should be treated as if they were completely seperate universes to the 'real' one in general. Murder, theft, assault, terrorism... right down to copyright infringement shouldn't be banned in a fictional world unless that world's setup explicitly sets laws against those things. We have to keep fiction and reality seperate.

codeulike 9y ago
I don't think I'm talking about fiction. OK, how about looking at it this way: At some point game designers will need to start simulating things like 'fear' to make NPCs more realistic. Then imagine the simulation getting better and better and better. Its easy to imagine them simulating a whole complex range of feelings and responses, simulating an inner-world for the NPCs. Then apply moore's law and let the simulations exponentially improve in accuracy and fidelity year on year. At what point would simulations of fear and terror in NPCs start to make you feel uneasy? Unless you believe in dualism or some sort of bio-centrism that means machines can never be conscious, I think there are interesting ethical issues there.
- Zarath 9y ago
  Can't believe you are receiving downvotes my friend. It seems very likely that consciousness is an emergent property of a sufficiently complex system. Given that this is the case, if we have a computational system such that it sufficiently mimics all nonphysical human behavior, it's very tough to argue that it does not share consciousness, especially if you are not a solipsist.

NeutronBoy 9y ago

Is there a difference between shooting at a character controlled by an AI, and shooting at a character controlled by another human?

codeulike 9y ago
Yes of course. I'm hypothesising character AIs that don't know they are in a game. From their point of view, it's all 'real'. They might get built that way to maximise the realism of their behaviour.

StavrosK 9y ago

You think shooting at them is bad? What about quitting the game?

wst_ 9y ago
I think it depends on how much can you get yourself into the world and story. When you play CS/Quake death match the goal is simple, there's no second thought before pulling the trigger - there's not story to begin with, just frags/scores. But when you are playing some good RPG and come to some village, you could start the killing spree, see people around you crying, bleeding, crawling at your feet and begging for their lives... would you have any thoughts before pulling the trigger then? My level of immersion usually is quite high if I like a game. Even today. Despite crud graphics and AI behavior. I'll use an imagination to tell the missing part of the story or give NPCs goals and desires. Simply because if feels better. By analogy, I would not kill innocent because that feels bad to me. So, if the day will come when all this missing part will get filled by brilliant AI I am not going to kill it either. Probably the reason that people have no quarrels with killing everyone in a game is that it's very difficult to convince player that NPCs have a feelings. Even in modern AAA titles when you start shooting in the middle of the city it looks more ludicrous than real. People are frightened, sure, but very often they make really fake poses, run randomly with hands up, or do some other stupid things. It's hardly convincing, so shooting them is easy and players have no remorse about it. You can easily see they are fake. But should they behave really human-like, should you see their pain... would you still pull the trigger?
- StavrosK 9y ago
  I mean that, if they're living, quitting the game is literally murder. Keep in mind that an intelligent AI wouldn't be hurt when you shot an NPC any more than a show writer would be hurt when a character got shot in the show. There wouldn't be one AI that felt pain per NPC, it would be a general orchestrator deciding which NPC to make cry and how they should act.

wst_ 9y ago

If they would be so advanced as you say, they would probably be conscious enough to be able to play any game asked. If that's the case, we could just create them outside of a game and ask them to play against humans. They would, sort of, role play as NPC or as an adversary. That would solve the problem of acting violently against them. They wouldn't get hurt, since they wouldn't be the suffering characters per se. They would just play the same game and understood the implications.