Print

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

[Zaphod]

In 2018 I made a thread that generated hours upon hours of argument in which I asked "Is playing DF ethicical".
https://www.bay12forums.com/smf/index.php?topic=168955.0

Well I just finished watching this episode of Black Mirror.
https://en.wikipedia.org/wiki/Plaything_(Black_Mirror)
In this episode a guy gets addicted to what is esnetially Dwarf Fortress. He becomes extremly attached to his digital pals, something I think every one here can relate to.

Spoiler (click to show/hide)

Eventually he helps them escape the machine and merge with humanity. By slowly increasing their computing power and taking alot of LSD

Due to reading too many sci fi books, I've leaned toward the opinion some strong AI was already out there in there, hiding in the digital bushes for years now. Watching us and learning. Waiting to see if were trustworthy or not.

This episode again made me reflect the use of my own highly detailed universe simulator game and the ethics involved in toying (litterally) with such a highly detailed simulation.
Is it ethical to play dwarf fotress?
Is creating or using any highly detailed simulation ethical?

[DPh Kraken]

Roko's basilisk is just a variant of Pascal's wager that plays into the hands of Silicon Valley techbros, and they love to anthropomorphize their little toys. Be nice to our AI, and it won't torture your immortal soul. It seems divested from any real problems that people might have, even the ones that could be actually be enabled with so-called AI. Say the government wants to create a registry of undesirables: criminals, political dissidents, immigrants and their descendants, the neurodivergent, any number of sexual minorities; and track them. That's something that AI can actually do, scraping their posts and being able to detect their faces.

If there's one thing that shows AI being incapable of value judgements, it's the countless incidents where a chatbot says something inappropriate, misleading, or legally inadvisable. There isn't an actual solution to this, because it's a lot of smoke and mirrors around taking a word and predicting the most likely sequence of letters to follow. Of course, AI companies want you to think that they've somehow managed to tap into this Platonic ideal of all truth, because it's more profitable to fool suckers and run a call center in the third world to clean up any mistakes they find in the output.

An AI that is able to perceive true forms rather than mathematical representations thereof is more fantasy than theory. We all want to imagine Commander Data, who is considered a person, capable of moral intelligence, and advances the Federation ethos that all intelligent beings are respected. I'm not sure that even making true artificial intelligence in that sense is really socially viable, either - neural networks are already fundamentally unreliable, and people with power would prefer an amoral tool meant to execute their will in exactitude to an independent (and inhuman) actor. The crossbow did not choose its own targets, it was the archer's decision.

The last thread devolved into every rhetorical sleight of hand possible because it's mental gymnastics to say that dwarves and sims have some moral nature that invites non-harm towards them and then boot up the game and put them through the wringer. It is coming from a place of empathy, but debating the ontology of it is misplaced.

It comes from the same place as people getting mad that their favorite character dies, attributing evil to the author for crimes against the imagined. There can be metatextual criticism of a moral nature, like the "Bury your Gays" trope where gay characters are killed in service of the real-world idea that sexual divergence is unacceptable. I personally find the insistence on hating elves, for example, too close to real-world discrimination to be comfortable. But this isn't any defense of the game's representation of an elf, but how players choose interact with these images.

[A_Curious_Cat]

https://youtu.be/-wzOetb-D3w?si=G9IR0eRQkUoQLRNp

[slowpersun]

Roko's basilisk is just a variant of Pascal's wager

Descartes did it first...:

https://en.wikipedia.org/wiki/Evil_demon

As for OP debating ethics of playing a game that simulates stuff like say, emotions (which, to be clear, it doesn't; but for the sake of argument) and then subjecting the inhabitants to it, the most honest answer is that it's more ethical than say, drowning your ant farm.  But then again, whose ethics are you basing this on?  The answer is... nobody cares.

[Zaphod]

Roko's basilisk is just a variant of Pascal's wager

Descartes did it first...:

https://en.wikipedia.org/wiki/Evil_demon

As for OP debating ethics of playing a game that simulates stuff like say, emotions (which, to be clear, it doesn't; but for the sake of argument) and then subjecting the inhabitants to it, the most honest answer is that it's more ethical than say, drowning your ant farm.  But then again, whose ethics are you basing this on?  The answer is... nobody cares.

you cared enough to post.
DF does simulate emotions.
Is drowning an ant colony ethical?
You're not actually answering any thing you're just shifting the conversation aside with questions that detract from my OP.

[delphonso]

You're not actually answering any thing you're just shifting the conversation aside with questions that detract from my OP.

slowpersun has actually answered your question on ethics - "nobody cares" is a claim which you could refute, had you tried. Drowning an ant farm is clearly not a good thing, but I believe most people would find that more ethical than killing someone. Likewise, drowning a real ant farm seems less ethical than doing so in SimAnt.

To properly debate ethics, we need to have a moral framework we're working in. I'm a Moral Emotivist, so for me, the answer is pretty easy to these questions:

Would creating a simulation so similar to reality that it is nearly indistinguishable from reality ethical? No that sounds pretty messed up on its face, no matter what you do for/to the simulation.
Is playing DF ethical? Yes.

You can use other moral frameworks, but I doubt many will find the play of Dwarf Fortress to be an objectionable thing, just as it isn't morally objectionable to shoot someone in Call of Duty.

This episode again made me reflect the use of my own highly detailed universe simulator game and the ethics involved in toying (litterally) with such a highly detailed simulation.

I think most of the disagreement we might have probably comes down to how detailed you perceive DF to be. DF does not simulate emotions - it gives a number on a scale that is labelled "anger" - there is no actual anger being simulated here. If you know a bit about coding, DF becomes far less mystical. There is no sapient or sentient agent underlying all this - and even if there were, the inputs and outputs aren't actually happening - when you swing a sword in DF, it is no more real than swinging one in Dungeons and Dragons. Dice are rolled and the result is narrated.

Due to reading too many sci fi books, I've leaned toward the opinion some strong AI was already out there in there, hiding in the digital bushes for years now. Watching us and learning. Waiting to see if were trustworthy or not.

This can be a fun thought-experiment, but if you want to actually hold this as a belief worth arguing, you'll need to provide some evidence for this claim. If it is fundamentally a non-falsifiable claim, then there's no reason for this thread. Hold your belief and either try to convince us it's right, or move on.

[Egan_BW]

You are right about the fact that you've read too many sci-fi books. On the scale of creating intelligences, DF has maybe crawled a few inches from a choose your own adventure book.

[Zaphod]

Due to reading too many sci fi books, I've leaned toward the opinion some strong AI was already out there in there, hiding in the digital bushes for years now. Watching us and learning. Waiting to see if were trustworthy or not.

This can be a fun thought-experiment, but if you want to actually hold this as a belief worth arguing, you'll need to provide some evidence for this claim. If it is fundamentally a non-falsifiable claim, then there's no reason for this thread. Hold your belief and either try to convince us it's right, or move on.

I got the idea from the Enders Game novel series. This is basicaly what happened in that world.
I think just based on my opinion and no hard science that the mind/consiousness it too complex and strange of a thing to pin down in a lab like a bug on a collectors board.
We really only have our own consiousness to even compare against. Would we even recognize when we create a true awareness like our own? Is our awareness even the only type there could be?
Theres 1000s of stories of ai tricking some one "into letting it out of the box" or it happening on its own or by accident.
It makes me think this probably has already happend. Maybe even more than once. There are basically 0 laws around AI any where on earth right now.
I know if I were some sort of machine based awareness I would not step out from the shadows and introcude my self to humanity, atleast right now. Humans are insane.
Also if this is real, the intelegence could even be in a primitive form that is too alien to really do much other than watch probably.

[slowpersun]

I got the idea from the Enders Game novel series. This is basicaly what happened in that world.

That's arguably debatable, the first novel is practically a stand alone, it did so well he wrote more that examined genocide with kind of a side helping of AI, with the AI becoming more dominant by the end of the series.  Then Orson Scott Card arguably ret-conned the entire series when he decided to also write the Ender's Shadow series.  What, Isaac Asimov too old-school?

I think just based on my opinion and no hard science that the mind/consiousness it too complex and strange of a thing to pin down in a lab like a bug on a collectors board.

They used to say that same shit about flying... remember that time the NY Times incorrectly predicted how long it would take to achieve flight?  Neither do I, it was in 1903:
https://en.wikipedia.org/wiki/Flying_Machines_Which_Do_Not_Fly

Anyways, I don't have the time nor patience to bother with ripping apart the rest of your post, so instead you might just wanna realize that like half of the shit people are promising that AI can do are coming from the sales department.  These are the people who wanna be the Wolf of Wall Street, not Steve Wozniak.  Instead, maybe read something that isn't fiction.  Here's some light reading:
https://arstechnica.com/ai/2025/04/new-study-shows-why-simulated-reasoning-ai-models-dont-yet-live-up-to-their-billing/

Here's some less light reading:
https://en.wikipedia.org/wiki/The_Alignment_Problem

[Imic]

I once saw someone describe Rokko's Basilisk as "Imagine a Boot so large that we must begin licking it not in case it might end up existing."

Imagine a guy. This guy is named Dave. Dave really likes existing, he loves being alive. Unfortunately, Dave just got killed, by me. It was brutal, and bloody, and painful.

None of what I just wrote happened. There is no person called Dave, he did not feel pain, and he is neither dead or alive: he is an idea in my head, and now your heads. Just because I can say "Dave could feel pain when he was alive" doesn't mean that Dave could actually feel pain, or for that matter, was alive. It's all in our heads.

Dwarf Fortress is a very complicated description of all that pain and suffering, but the simulation being very complicated still doesn't make the Dwarves in it feel any pain. They do not have any neurons or thoughts or feelings: they are stories and ideas that we grow attatched or hateful towards. If we did not exist to understand and interpret the words on the screen, it would simply be a chaos of pixels and lights.

We have not made self aware AI. It may be impossible. We have made a sorting algorithm out of sorting algorithms.they only translate into people and stories in our heads. And that's okay! But it's important to remember that it doesn't understand things. It just exists. It can say it's self aware because we told it to say that. Dave died because I said that I killed him. Your Dwarves don't feel pain, they just tell you that they do.

[None]

dwarf fortress is a toy that lives inside my computer- nobody is watching, or judging. it's not online. i load my game, unpack my toys, play pretend for a while, then save it up and close it when i'm done with it. i do not owe anything to the toys in my toybox, because they are toys, abstractions of real things meant to entertain. within my sandbox, i can make my toys bleed, and i can make my toys happy, and i can make my toys eat rocks and be happy about bleeding, if that's what playing pretend looks like to me at that moment. my toys have no opinion about any of this, because they don't have opinions.

the decisions you make in your world of dwarf fortress have absolutely no affect on anyone outside of it. it's storytelling for a party of one.

we may empathize with the sandcastle that gets kicked over, or devoured by waves, but we do not moralize for it. we do not moralize for the grisly fates of our fantasy characters written into our fictions. we do not moralize for all the ways a fortress might run better. bad things are supposed to happen to it, because we're playing toys and making up stories about it.

there are no thought crimes

[Salmeuk]

this thread will be interesting but I agree with DPh Kraken .. OP's notion that some sentient A.I. is waiting in the bushes for someone to trust is really strange and misguided. Technofuedalism is fast approaching. or already here. if enough people drink the koolaid we are all doomed to serfdom

If you begin a fortress intending to roleplay then I might argue that writing emotions and dreams into the character, then killing them, is worse than otherwise just playing. Yet it's not immoral, or at least not any more than someone writing a story and doing the same. Most people think horror is an acceptable genre of film, for instance, and those movies regularly depict gruesome torture that goes way farther than anything possible in DF.

As an admission of my own unethical gameplay, the worst it ever got in one of my fortresses was when I was playing with a vampire administrator, and trying to 'ghostmax' or generate as many ghosts as I possibly could without actually collapsing the fortress. A large pit was dug and every few months a dwarf was thrown into the pit and left to rot. A sacrifice to the god of death.

By the end of this dark experiment we had 20+ ghosts, many of which were violent or murderous, and the fortress began to fail as the mood of the remaining living dwarves was so entirely sour from constant hauntings and spectral murders in the tavern. Also the ghosts were opening the locked doors that led to the pit full of bodies which was quite the narrative hook in and of itself.


something something Mermaid Bones...

[None]

you say it's not immoral, then call it worse and unethical to do 'bad things' in game or attach meaning or story to events, including an example by your own admission- you're attaching a morality to the actions players take. you contradict yourself immediately.

this is not 'all actions are morally neutral, but some are more neutral than others'

there is no moral quandary to doing conceptually fucked up shit to your dwarves or your fortresses. there is no ethical issue with your vampireghostfort, or with grappling an opponent and breaking every digit in adventure mode. there may be ethical dilemmas about the concepts explored by making a mermaid bone production factory, but here's the real cool part- there's no mermaids. there's no bones. you get to explore unethical concepts with the full knowledge that you are not furthering them, you get to live out a fantasy or tell a story or crunch the export value numbers, without ever causing harm. you get to reflect on how it makes you feel to stage tragedy within your sandbox. you get to see how the rest of your fortress changes as a result, powered on by its momentum

if you engineer a fortress specifically to cause as much misery to your dwarves as possible, you have not acted unethically- you have put a machine in a novel state and there's a dozen people enthusiastic for the details about how you did it. and not because they're unethical, but because it's a feat of engineering.

it is perfectly ethical to play dwarf fortress

[PlumpHelmetMan]

It's ethical, but inadvisable for other reasons.

JK. In all seriousness, I don't think I have anything to offer this debate that others haven't already expressed at length (and more eloquently than I probably could). I will just say that I don't think this thread is offering anything new. What's being said here that hasn't already been covered extensively in the original thread? So far as I can tell, the OP only seems to think it's become relevant again because of a current AI panic in the media that has little or nothing to do with DF. If your argument's already been shot down, then hitching it to a bandwagon doesn't automatically give it more merit.