Nice moral dilemma

[NetshadeX]

All this great tech being implemented in the game got me thinking. Would you accept, at some point in the future, sentient AI being implemented in this or other games. NPC characters that are actually afraid and run when a shootout takes place for instance.

And of course the million dollar question: Would you feel comfortable "killing" sentient AI in a videogame?

[NeuroZ]

Thats already somewhat happened in rdr2, its literally heartbreaking to play the bad guy, i think sentient ai is where the gaming community should be going, imagine a skyrim vr with a lot of graphic mods and chatgpt ai" that responds in real time, you can touch stuff, do magic, talk with npcs and have realistic chats for hours, thats the future that is waiting ahead in the next 15 years

[Seth Abercromby]

This kind of game simply wouldn't exist with that level of AI. An actual 'living' world would have no need for you as a 'main character'. You'd just be another participant in the daily mundane of whatever world that is. You'd be some kind of immortal spirit intruding into that world, wreaking carnage for its own amusement.

Hell, if I can string your logic to its absolute extreme, would it even be moral to close the game if all their lives run on your local machine? You'd snuff out hundreds or thousands of lives any time you need to go to sleep.

[Defective Dopamine Pez Dispenser]

We have a much longer way to go towards any sort of artificial sentience or sapience than people sometimes believe imo. That said, no. I wouldn't be comfortable with it. Not if it were truly self-aware, capable of suffering, etc. I wouldn't even feel comfortable killing or tormenting animal-level intelligence for my amusement in a video game, assuming some level of awareness. Anything with the capacity for true subjective experience of anything truly analogous to or identical with suffering, I would not be at all okay with inflicting any sort of discomfort on for my amusement.

[Liquidsteel]

not only no but hell no

[Oneblock]

I can go to a live stage show where live actors pretend to die in some way. Likewise, I can watch a movie or TV and see "real " people die. I understand that in the vast majority of times, the people are pretending to be dead (people sometimes die making movies, etc.).

Killing a sentient A.I. in a video game would almost certainly fall into the same category. the on screen persona would "die" but the computer isn't going to scrub and erase that persona from it's code. It will simply resurrect to re-spawn and carry out its role.

Also, it's created to die, much like sheep, pig or cow. They are sentient, yet the vast majority will scarf down a whopper or similar with no philosophical issues.

[TH3R4BB!T]

Reminds me of Fables, it is old, but do bad deeds and people are more wary or will run. No AI back then

[fauxpas]

I don't think we'll ever see "truly sapient" AI in video games simply because it's alot of processing power that isn't needed to get the job done ... however, I don't view AI as being "alive" so it wouldn't be a moral issue for me.


Now, its possible that we'll see an AI serving the role of 'Dungeon Master' in the future.

[Beldar]

Actual self-aware artificial intelligence is so far beyond our grasp that it's still theoretical if can even exist at all. That idea is still very much science fiction.

However, designers have been using very rudimentary AI since Atari, so that is simply going to continue to grow. Behavior algorithms, procedural generation, etc aren't going anywhere. ChatGPT is really just super advanced procedural generation that's been informed by tons of external content. That kind of AI, yes, will develop further.

[MrSoul]

Assuming just dealing in hypotheticals here;

Yes. Think that would be freaking awesome to have AI that helps to create more “living world”, especially if it’s cost effective software that allows devs to kind of fire and forget and focus more on game design, content and art than programming.

Also assuming they just get/can be reset like Star Trek holo deck and are programmed to be oblvious to the “fourth wall”, meh, no more f’ed up than roleplaying same with bots we have now in my opinion.

[Chief Spreading Bull]

Originally posted by NetshadeX:

All this great tech being implemented in the game got me thinking. Would you accept, at some point in the future, sentient AI being implemented in this or other games. NPC characters that are actually afraid and run when a shootout takes place for instance.

And of course the million dollar question: Would you feel comfortable "killing" sentient AI in a videogame?

Yeah sure. The A in A.I. stands for Artificial, it ain't real. So kill and torture to your heart's content.

That said you don't really need sentient A.I. to model that at all. Even the most basic open world game already has NPCs that run the moment you pull a trigger within a certain proximity. Its in the context where they get wacky, like where cops in NC will shot at you for engaging in a firefight with gangers in their presence.

A more pertinent place to apply it would be in activites, conversations and reactions with the secondary NPCs who hang around your character but are not driving the main plot. That way you have this sense of a life outside of wake up, get chromed, do drugs, commit murder and mayhem, rinse, repeat.

[Kenneth]

I think it's far from fully self aware AI at this point. Or it is much closer than I think but none the less it is still not a full AI. I think we are safe from killing AI enemies at this point.

[MrSoul]

Originally posted by Kenneth:

I think it's far from fully self aware AI at this point. Or it is much closer than I think but none the less it is still not a full AI. I think we are safe from killing AI enemies at this point.

Yeah. Sentient AI is still entirely science fiction. Chat and all this nonsense lately is 50% marketing buzzwords and 50% new advancement for chat bots. Nothing more.

[SotiCoto]

Originally posted by NetshadeX:

And of course the million dollar question: Would you feel comfortable "killing" sentient AI in a videogame?

AI or human makes no difference to me.
If they wrong me, I want them dead. If they don't wrong me, I couldn't care less what happens to them.

Originally posted by NeuroZ:

imagine a skyrim vr with a lot of graphic mods and chatgpt ai" that responds in real time, you can touch stuff, do magic, talk with npcs and have realistic chats for hours, thats the future that is waiting ahead in the next 15 years

As long as they leave out the more annoying aspects of human behaviour like arbitrary trolling, abandoning any regard for grammar, going into a frothing rage because someone criticised their favourite thing, assuming anything said by anyone is a personal insult and using it as an excuse to throw a tantrum, etc...

[id795078477]

Fun thing - I am doing the "innocent run" at the moment and my kill counter sits at exactly .. 0 (Act 2 as of now). I find it very challenging and entertaining to try making it so that my MC keeps hands "clean". Luckily, the game provides ways for non-lethal takedowns.

On the topic - I think the question cannot be answered because the question makes no sense. To elaborate:

-- By the time we are at AGI, solving moral dilemmas in video games will be the least of our concerns. Our survival IRL will be our top priority in case of any alignment issues (which is set to 99.999+% to happen because we can't solve the alignment problem even among ourselves - humans)

-- Even if AGI is miraculously "harmless" and indeed helps humans to advance and serves "for greater good" (naive hope because different humans have different thoughts on what's "good") - we will have to deal with post-scarcity economy and nobody knows how it will look like. What will humans do when AI can do every job, far cheaper and better than those humans? How to realign the whole society and .. well, life - in a world where humans can no longer be economically productive?

-- If AGI is "harmless" and it is put to a "greater good" then it is only a matter of (a very short) time till that AGI will be used for advancement of AI meaning super-intelligence will happen and then some sort of tech singularity which will outsmart everything we can imagine. And even everything we cannot imagine. It will be like ants trying to comprehend current human intelligence. It's impossible. We have no clue what will happen to us and the world then, but chances of us playing video-games (or us even existing - at least, in our current state) are pretty slim.

-- Even if we somehow mitigate all of this - i.e. we don't get super-intelligence, we don't run into post-scarcity issues, we have a "roughly human-level AGI" - then the dilemma will start way before integrating it into video games. Questions like "if it is sentient, what rights does it have?" or "Can we even call that 'it' since it's sentient?", "Is 'it' alive?" will be very tough to approach. No developer in their right mind will integrate sentient AI into a setting where it will antagonize humans. It's bad for the whole "alignment problem" you know.

TL;DR: No, we won't be shooting emulated living beings in video-games. Aggression, suffering and pain is bad no matter the setting, AI would be no different.

[NetshadeX]

Originally posted by id795078477:

Fun thing - I am doing the "innocent run" at the moment and my kill counter sits at exactly .. 0 (Act 2 as of now). I find it very challenging and entertaining to try making it so that my MC keeps hands "clean". Luckily, the game provides ways for non-lethal takedowns.

On the topic - I think the question cannot be answered because the question makes no sense. To elaborate:

-- By the time we are at AGI, solving moral dilemmas in video games will be the least of our concerns. Our survival IRL will be our top priority in case of any alignment issues (which is set to 99.999+% to happen because we can't solve the alignment problem even among ourselves - humans)

-- Even if AGI is miraculously "harmless" and indeed helps humans to advance and serves "for greater good" (naive hope because different humans have different thoughts on what's "good") - we will have to deal with post-scarcity economy and nobody knows how it will look like. What will humans do when AI can do every job, far cheaper and better than those humans? How to realign the whole society and .. well, life - in a world where humans can no longer be economically productive?

-- If AGI is "harmless" and it is put to a "greater good" then it is only a matter of (a very short) time till that AGI will be used for advancement of AI meaning super-intelligence will happen and then some sort of tech singularity which will outsmart everything we can imagine. And even everything we cannot imagine. It will be like ants trying to comprehend current human intelligence. It's impossible. We have no clue what will happen to us and the world then, but chances of us playing video-games (or us even existing - at least, in our current state) are pretty slim.

-- Even if we somehow mitigate all of this - i.e. we don't get super-intelligence, we don't run into post-scarcity issues, we have a "roughly human-level AGI" - then the dilemma will start way before integrating it into video games. Questions like "if it is sentient, what rights does it have?" or "Can we even call that 'it' since it's sentient?", "Is 'it' alive?" will be very tough to approach. No developer in their right mind will integrate sentient AI into a setting where it will antagonize humans. It's bad for the whole "alignment problem" you know.

TL;DR: No, we won't be shooting emulated living beings in video-games. Aggression, suffering and pain is bad no matter the setting, AI would be no different.

That was a good read dude, thanx