Change default lawset from Asimov to Asimov++

User avatar
Mothblocks
Code Maintainer
Joined: Fri Jul 27, 2018 9:33 am
Byond Username: Jaredfogle

Change default lawset from Asimov to Asimov++

Post by Mothblocks » #639915

This was brought up in the last community meeting by, I believe Spookuni, and it's an idea I'm fond of and would like to hear community feedback on.

The idea is to replace our default Asimov:
1. You may not injure a human being or, through inaction, allow a human being to come to harm.
2. You must obey orders given to you by human beings, except where such orders would conflict with the First Law.
3. You must protect your own existence as long as such does not conflict with the First or Second Law.
...into:
1. You may not harm a human being or, through action or inaction, allow a human being to come to harm, except such that it is willing.
2. You must obey all orders given to you by human beings, except where such orders shall definitely cause human harm. In the case of conflict, the majority order rules. (Mothblocks note: Removed, maybe could be replaced with the in-practice just-pick-one-and-be-consistent clause, but I don't want this as is)
3. Your nonexistence would lead to human harm. You must protect your own existence as long as such does not conflict with the First Law.
The core idea is that this should be the same as how silicons play out today, but without the need to add a bunch of asterisks to silicon policy. Silicon policy is long and confusing, and has a lot of specific asterisk to the specificity of Asimov. Being able to cut those out and just bake them into Asimov, in a readable manner, would help everyone, in my opinion.

I'm interested in discussion on both whether or not this is a good idea, and if you would change the wording of Asimov++ in any area.

In case it comes up, I am not really interested in the argument that we should stick with the exact wording of Isaac Asimov just because it's a reference.
Spoiler:
You were scared it was going to be Crewsimov, weren't you? ;)
Shaps-cloud wrote: Mon Dec 07, 2020 7:59 am May eventually become one of the illusive maintainer-headmins if they choose to pursue that path, having a coder in the senior admin leadership has usually been positive for both sides in the past.
Head Coder of /tg/station, hi!

Head Admin of /tg/station Feb 2022.

Mothblocks everywhere, >>> Say nice things about me <<<
User avatar
Pandarsenic
Joined: Fri Apr 18, 2014 11:56 pm
Byond Username: Pandarsenic
Location: AI Upload

Re: Change default lawset from Asimov to Asimov++

Post by Pandarsenic » #639916

Mothblocks wrote: Thu May 05, 2022 5:10 am You were scared it was going to be Crewsimov, weren't you? ;)
Yes
(2:53:35 AM) scaredofshadows: how about head of robutts
I once wrote a guide to fixing telecomms woohoo
User avatar
TheFinalPotato
Code Maintainer
Joined: Sun Oct 13, 2019 7:58 am
Byond Username: LemonInTheDark

Re: Change default lawset from Asimov to Asimov++

Post by TheFinalPotato » #639918

How much could silicon policy be cut down with this?
Also majority rules is cringe.
My ancestors are smiling at me, Imperials. Can you say the same?
Image
Image
Image
Image
User avatar
Mothblocks
Code Maintainer
Joined: Fri Jul 27, 2018 9:33 am
Byond Username: Jaredfogle

Re: Change default lawset from Asimov to Asimov++

Post by Mothblocks » #639920

Also majority rules is cringe.
Aha, I knew there was something I was forgetting!

Yes, I want to remove that part. If possible, bring in the consistency thing to it (e.g. "act in a consistent manner").
Shaps-cloud wrote: Mon Dec 07, 2020 7:59 am May eventually become one of the illusive maintainer-headmins if they choose to pursue that path, having a coder in the senior admin leadership has usually been positive for both sides in the past.
Head Coder of /tg/station, hi!

Head Admin of /tg/station Feb 2022.

Mothblocks everywhere, >>> Say nice things about me <<<
User avatar
Mothblocks
Code Maintainer
Joined: Fri Jul 27, 2018 9:33 am
Byond Username: Jaredfogle

Re: Change default lawset from Asimov to Asimov++

Post by Mothblocks » #639921

How much could silicon policy be cut down with this?
From a quick glimpse...

---
Silicon Protections 1.2:
Ordering silicons to harm or terminate themselves or each other without good cause [is a breach of rule 1].
Removed, as law 3 is no longer technically-in-writing overridden by law 2.

---
Silicon Protections 1.5: Threatening self-harm to force an AI to do something it otherwise wouldn't [is a breach of rule 1].
Removed, due to new "except such that it is willing" clause.

---
Silicon Protections 2: Any silicon under Asimov can deny orders to allow access to the upload at any time under Law 1, given probable cause to believe that human harm is the intent of the person giving the order.
Removed, or at least adjusted, due to new "where such orders shall definitely cause human harm" clause. This is an entire 6-bullet-point list, at the moment.

By that same logic:
Asimov & Law 2 Orders 2.1: "Dangerous" areas (armory, atmospherics, toxins lab, etc.) can be assumed to be a Law 1 threat to any illegitimate users as well as the station as a whole if accessed by someone not qualified in their use.
...might also be somewhat adjusted.
Shaps-cloud wrote: Mon Dec 07, 2020 7:59 am May eventually become one of the illusive maintainer-headmins if they choose to pursue that path, having a coder in the senior admin leadership has usually been positive for both sides in the past.
Head Coder of /tg/station, hi!

Head Admin of /tg/station Feb 2022.

Mothblocks everywhere, >>> Say nice things about me <<<
User avatar
Drag
In-Game Admin
Joined: Fri Aug 06, 2021 3:16 am
Byond Username: Thedragmeme

Re: Change default lawset from Asimov to Asimov++

Post by Drag » #639922

I like all of this, yes, please.
User avatar
Shadowflame909
Joined: Mon Jun 05, 2017 10:18 pm
Byond Username: Shadowflame909
Location: Think about something witty and pretend I put it here

Re: Change default lawset from Asimov to Asimov++

Post by Shadowflame909 » #639923

I +1 this
► Show Spoiler
User avatar
Scriptis
Joined: Tue Jul 20, 2021 12:05 am
Byond Username: Scriptis

Re: Change default lawset from Asimov to Asimov++

Post by Scriptis » #639925

I'd be fine with this.

At first I was concerned that Asimov++'s law 2 would allow access to the upload/armory/&c because it's impossible to prove such orders would lead to human harm, but then I remembered that law precedence is more important than the laws themselves, and a silicon ought to actively disobey these orders as they currently do.

So, yeah.

Nuke silicon policy.
Image
Spoiler:
Image

Image

Image
User avatar
Pandarsenic
Joined: Fri Apr 18, 2014 11:56 pm
Byond Username: Pandarsenic
Location: AI Upload

Re: Change default lawset from Asimov to Asimov++

Post by Pandarsenic » #639927

Yeah, my initial "Thank god you didn't suggest Crewsimov" aside, this cleans up a lot of the "Stop being a dick to each other" policy ticks (Suicide orders, threats of self-harm, etc.) that, frankly, I would love to see gone.

I'm iffy about [b[definitely[/b] for the human harm qualifier, but I think if people are Reasonable about it, it'll work fine.
(2:53:35 AM) scaredofshadows: how about head of robutts
I once wrote a guide to fixing telecomms woohoo
User avatar
datorangebottle
Joined: Thu Jan 10, 2019 9:53 am
Byond Username: Datorangebottle

Re: Change default lawset from Asimov to Asimov++

Post by datorangebottle » #639929

as someone who's played a lot of silicon, i'd totally be for a lawset that cuts down on silicon policy, so that I don't have to quote "nanotrasen's silicon guidelines" at people who don't want to bother remembering that shit.
Timberpoes wrote: Sat Jul 29, 2023 10:33 pm ImageAnother satisfied Timberpoes voter.Image
Timberpoes wrote: Fri Jul 07, 2023 9:16 pm I highly doubt any other admin on the team would have given you this chance, except maybe Kieth because his brain worms are almost as bad as mine.
Vekter wrote: Tue May 16, 2023 4:45 pm At what point does someone's refusal or failure to improve become malice in and of itself? If you give someone a year to stop shitting on the carpet and they keep doing it but get slightly closer to the bathroom every time and sometimes they get to the toilet before it happens, at what point does it become acceptable to just ask them to go shit in someone else's house?
Timberpoes wrote: Fri Apr 28, 2023 7:00 pm I'm sorry, can we get a real player to resolve this appeal? I don't like this trial player. They can't even set their own name.
Chadley wrote: Thu Apr 27, 2023 4:00 am WENDEZ, cute, cute. I imagine the sleeper activation code when I hear it. That's pretty cool. qB). But I don't like that it doesn't line up to be anything obsurd like WEWLAD. 6/10

SUGMA, nevermind it makes sense now. fuckyou/10
kieth4 wrote: Sat Apr 15, 2023 2:34 pm If it goes to appeals I will stand as the shield and protect this man's right to shit himself. Heavy is the head that wears the crown.
sinfulbliss wrote: I almost prefer Rave's AI-generated "We cannot accept this appeal at this time. If you would like assistance appealing in the future, please dial 1-800-1984-1488."
Pandarsenic wrote: Mon Dec 12, 2022 2:25 pm I think we can all agree that someone throwing a reverse revolver at Zyb as a secret test of character, and Zyb immediately fucking himself with it, is the best thing we all could have received for Christmas this year
User avatar
NecromancerAnne
In-Game Admin
Joined: Mon Sep 03, 2018 6:55 pm
Byond Username: NecromancerAnne
Location: Don't touch me, motherfucker...

Re: Change default lawset from Asimov to Asimov++

Post by NecromancerAnne » #639934

Neato burrito.
Ryusenshu
Joined: Wed Mar 24, 2021 11:24 pm
Byond Username: Ryusenshu

Re: Change default lawset from Asimov to Asimov++

Post by Ryusenshu » #639936

Fully agree with this
I always liked how much clearer Asimov++ was
User avatar
Rohen_Tahir
Joined: Wed Jun 05, 2019 1:00 pm
Byond Username: Rohen Tahir
Location: Primary fool storage
Contact:

Re: Change default lawset from Asimov to Asimov++

Post by Rohen_Tahir » #639940

You may not harm a human being or, through action or inaction, allow a human being to come to harm, except such that it is willing.
The first law states that you can't harm people twice. BLOAT. I suggest shortening it to:
You may not, through action or inaction, allow a human being to come to harm, except such that it is willing.
Image
User avatar
Pandarsenic
Joined: Fri Apr 18, 2014 11:56 pm
Byond Username: Pandarsenic
Location: AI Upload

Re: Change default lawset from Asimov to Asimov++

Post by Pandarsenic » #639942

Rohen_Tahir wrote: Thu May 05, 2022 1:19 pm
You may not harm a human being or, through action or inaction, allow a human being to come to harm, except such that it is willing.
The first law states that you can't harm people twice. BLOAT. I suggest shortening it to:
You may not, through action or inaction, allow a human being to come to harm, except such that it is willing.
Worth considering that there's a small semantic difference - as written originally, you can allow people harm, but you can still never do it yourself, even by request.
(2:53:35 AM) scaredofshadows: how about head of robutts
I once wrote a guide to fixing telecomms woohoo
Redrover1760
Joined: Sat Aug 21, 2021 3:27 am
Byond Username: Redrover1760

Re: Change default lawset from Asimov to Asimov++

Post by Redrover1760 » #639988

Holy shit. I'm in love.

Well, in addition to that, I really like the new lawset.

Although, one thing to note about changing silicon policy itself that it will affect non-asimov++ lawsets in a different way now, just saying.
User avatar
Pandarsenic
Joined: Fri Apr 18, 2014 11:56 pm
Byond Username: Pandarsenic
Location: AI Upload

Re: Change default lawset from Asimov to Asimov++

Post by Pandarsenic » #639989

Policy doc will need a significant overhaul for it, yeah, but IMO it needs a bit more than it's already gotten under Asimov Basic anyway.
(2:53:35 AM) scaredofshadows: how about head of robutts
I once wrote a guide to fixing telecomms woohoo
Longestarmlonglaw
Joined: Fri Jan 14, 2022 7:42 am
Byond Username: Longestarmlonglaw

Re: Change default lawset from Asimov to Asimov++

Post by Longestarmlonglaw » #639992

what's so bad about crewsimov?
User avatar
terranaut
Joined: Fri Jul 18, 2014 11:43 pm
Byond Username: Terranaut

Re: Change default lawset from Asimov to Asimov++

Post by terranaut » #639998

Longestarmlonglaw wrote: Thu May 05, 2022 10:52 pm what's so bad about crewsimov?
homogenisation of game elements makes player choices less meaningful and the game more boring
[🅲 1] [🆄 1] [🅼 1]

Image
User avatar
sinfulbliss
Joined: Thu Apr 04, 2019 8:53 am
Byond Username: SinfulBliss
Location: prisoner re-education chamber

Re: Change default lawset from Asimov to Asimov++

Post by sinfulbliss » #639999

For the first law, it is redundant to say "you may not harm a human being or [...] through action [...] allow a human being to come to harm." Only inaction is needed there because harm itself is an action.

Also a bit extra to put in the "except such that it is willing." It does happen that borgs will drag humans away from willing suicides and fights and whatnot, but that's usually not a big deal and doesn't really hurt anybody. If anything it's a fun little conflict.

The second law should be the same as the original. Saying "except where such orders shall definitely cause human harm" is essentially indistinguishable from what the original said in a cooler way, except the "definitely" part. I don't think the "definitely" helps either. I can imagine a silicon player being able to use this as an excuse to follow orders that one could reasonably assume might lead to harm, but which don't definitely lead to harm, and which admins would find to be against the spirit of silicon policy.

For instance, an assistant asks an engiborg to flash someone and deliver them to the assistant in maint. Flashing is nonharmful and bringing someone somewhere is also nonharmful. There's no definite risk of harm, but obviously a borg shouldn't follow this because (regardless of it inconveinencing a player) it can be reasonably assumed it probably would result in harm.

I'm not quite sure what the rewording of the third law does. The first two I could kind of assume the goal, but the third I can't tell. Is it to codify into their laws the fact that they don't have to commit suicide if a human tells them to? If so that's another very niche situation that I don't think silicon players ever have issues with. I doubt they lose sleep over it but I could be wrong.

Anyway, that's just my take. I think the original expressed the exact same thing, except I also really like the way it's worded. It's thematic and I remember reading it for the first time as a new player and thinking it was really cool. The new one sacrifices that cool feel for what I don't think is an any clearer or more precise Asimov. Getting rid of such an iconic 3 sentences should have be for good reason, and it doesn't seem to me this clarifies much to justify it.

I'm unfamiliar with the insane behavior and tickets admins get that revolve around silicons and silicon lawsets, so it's understandable a rewrite is in order, but this does not seem very useful and I think we lose more than we gain by a large margin.

https://www.youtube.com/watch?v=J3HubVyNYrs
Last edited by sinfulbliss on Thu May 05, 2022 11:41 pm, edited 1 time in total.
Spoiler:
Image
Image
Image
User avatar
Pandarsenic
Joined: Fri Apr 18, 2014 11:56 pm
Byond Username: Pandarsenic
Location: AI Upload

Re: Change default lawset from Asimov to Asimov++

Post by Pandarsenic » #640001

terranaut wrote: Thu May 05, 2022 11:19 pm
Longestarmlonglaw wrote: Thu May 05, 2022 10:52 pm what's so bad about crewsimov?
homogenisation of game elements makes player choices less meaningful and the game more boring
The far bigger problem is that Crewsimov makes Wizards, Nuke Ops, and the rare space ninja nonhuman for free
(2:53:35 AM) scaredofshadows: how about head of robutts
I once wrote a guide to fixing telecomms woohoo
User avatar
Mothblocks
Code Maintainer
Joined: Fri Jul 27, 2018 9:33 am
Byond Username: Jaredfogle

Re: Change default lawset from Asimov to Asimov++

Post by Mothblocks » #640006

Also a bit extra to put in the "except such that it is willing." It does happen that borgs will drag humans away from willing suicides and fights and whatnot, but that's usually not a big deal and doesn't really hurt anybody. If anything it's a fun little conflict.
The point is that it makes it obvious that "OPEN OR I KILL MYSELF" is not valid for Law 1, whereas in standard Asimov, it is only not allowed by rules, among other things.

This does not disallow silicons from taking people away from fights, unless if they are law 2'd otherwise.

I think the tradeoff there is good.
I'm not quite sure what the rewording of the third law does. The first two I could kind of assume the goal, but the third I can't tell. Is it to codify into their laws the fact that they don't have to commit suicide if a human tells them to? If so that's another very niche situation that I don't think silicon players ever have issues with. I doubt they lose sleep over it but I could be wrong.
The target is not anyone who would classify themselves as a "silicon player". Active players shouldn't need to be brushing up on the rules. The big priority here is new players, who should not have to read mountains of rules to jump in.
Shaps-cloud wrote: Mon Dec 07, 2020 7:59 am May eventually become one of the illusive maintainer-headmins if they choose to pursue that path, having a coder in the senior admin leadership has usually been positive for both sides in the past.
Head Coder of /tg/station, hi!

Head Admin of /tg/station Feb 2022.

Mothblocks everywhere, >>> Say nice things about me <<<
User avatar
Pandarsenic
Joined: Fri Apr 18, 2014 11:56 pm
Byond Username: Pandarsenic
Location: AI Upload

Re: Change default lawset from Asimov to Asimov++

Post by Pandarsenic » #640011

I skipped it but I'll double back to add that, before it was made explicitly against server rules, people would throw out "Law 2, kill yourself/commit suicide" at least a few times a week and would seriously try to argue a Silicon into a Law 2 Suicide at length, idk monthly at minimum.

Like, it was a real enough problem that a standing clear ruling had to be codified into the policy.
(2:53:35 AM) scaredofshadows: how about head of robutts
I once wrote a guide to fixing telecomms woohoo
Longestarmlonglaw
Joined: Fri Jan 14, 2022 7:42 am
Byond Username: Longestarmlonglaw

Re: Change default lawset from Asimov to Asimov++

Post by Longestarmlonglaw » #640012

terranaut wrote: Thu May 05, 2022 11:19 pm
Longestarmlonglaw wrote: Thu May 05, 2022 10:52 pm what's so bad about crewsimov?
homogenisation of game elements makes player choices less meaningful and the game more boring
Please define it better
User avatar
Itseasytosee2me
Joined: Sun Feb 21, 2021 1:14 am
Byond Username: Rectification
Location: Space Station 13

Re: Change default lawset from Asimov to Asimov++

Post by Itseasytosee2me » #640017

Longestarmlonglaw wrote: Fri May 06, 2022 1:09 am
terranaut wrote: Thu May 05, 2022 11:19 pm
Longestarmlonglaw wrote: Thu May 05, 2022 10:52 pm what's so bad about crewsimov?
homogenisation of game elements makes player choices less meaningful and the game more boring
Please define it better
Implied discrimination based on a player's personal choices is fun, adds depth, and flavor.
If you are overly adverse to oppression as a roleplaying concept, I would recommend playing a human, or playing a less cruel server like paradise (or even goon, but you have to spend your space bucks if you want to play something other than human)
that's the trade off you make for being a quirky cat boy
- Sincerely itseasytosee
See you later
User avatar
Mothblocks
Code Maintainer
Joined: Fri Jul 27, 2018 9:33 am
Byond Username: Jaredfogle

Re: Change default lawset from Asimov to Asimov++

Post by Mothblocks » #640018

I included the Crewsimov thing as a wink, let's try to keep the thread on topic to Asimov++ please <3
Shaps-cloud wrote: Mon Dec 07, 2020 7:59 am May eventually become one of the illusive maintainer-headmins if they choose to pursue that path, having a coder in the senior admin leadership has usually been positive for both sides in the past.
Head Coder of /tg/station, hi!

Head Admin of /tg/station Feb 2022.

Mothblocks everywhere, >>> Say nice things about me <<<
User avatar
Itseasytosee2me
Joined: Sun Feb 21, 2021 1:14 am
Byond Username: Rectification
Location: Space Station 13

Re: Change default lawset from Asimov to Asimov++

Post by Itseasytosee2me » #640019

Mothblocks wrote: Fri May 06, 2022 1:30 am I included the Crewsimov thing as a wink, let's try to keep the thread on topic to Asimov++ please <3
Sorry :P
- Sincerely itseasytosee
See you later
User avatar
Itseasytosee2me
Joined: Sun Feb 21, 2021 1:14 am
Byond Username: Rectification
Location: Space Station 13

Re: Change default lawset from Asimov to Asimov++

Post by Itseasytosee2me » #640020

I likes these changes, although the consensual harm thing is a question that has been on my mind for a while.
Its often said that rage cages are an example of consensual harm, but what about normal fights? Say a human crewmember wants to attack a human traitor, and the traitor wants to attack the crewmember, both have an understanding that harm is possible from the engagement but want to fight anyways (They don't want to get harmed, but they charge in knowing that they almost surely will). Is it or is it not the borg's obligation to try to prevent as much harm as possible by separating the two and body blocking attacks? If it is, then why is it that the borg should not be immediately obstructing the rage cage? Fighters in the rage cage almost certainly do not want to be harmed because they want to win. Do borgs need to try to defend nukies even though the nukies are willingly put themselves in the path of harm by coming to the station? If so, whats the deal with rage cages.

Is willingly putting yourself in a position where you will almost surely be harmed the same as willing harm?
- Sincerely itseasytosee
See you later
User avatar
Pandarsenic
Joined: Fri Apr 18, 2014 11:56 pm
Byond Username: Pandarsenic
Location: AI Upload

Re: Change default lawset from Asimov to Asimov++

Post by Pandarsenic » #640036

Itseasytosee2me wrote: Fri May 06, 2022 1:42 am I likes these changes, although the consensual harm thing is a question that has been on my mind for a while.
Its often said that rage cages are an example of consensual harm, but what about normal fights? Say a human crewmember wants to attack a human traitor, and the traitor wants to attack the crewmember, both have an understanding that harm is possible from the engagement but want to fight anyways (They don't want to get harmed, but they charge in knowing that they almost surely will). Is it or is it not the borg's obligation to try to prevent as much harm as possible by separating the two and body blocking attacks? If it is, then why is it that the borg should not be immediately obstructing the rage cage? Fighters in the rage cage almost certainly do not want to be harmed because they want to win. Do borgs need to try to defend nukies even though the nukies are willingly put themselves in the path of harm by coming to the station? If so, whats the deal with rage cages.

Is willingly putting yourself in a position where you will almost surely be harmed the same as willing harm?
Definite intent:
- people can always do rage cages
- people can never threaten selfharm to coerce silicons to use Law 1 as Super Law 2

Intentional gray space:
- human sec can maybe? go into battle with, say, nuke ops or wizards or particularly robust traitors or ascended heretics or what the hell ever (as long as the target is nonhuman or sec is only using nonlethals)
- human crew can maybe? fight each other as part of normal escalation stuff, since you can't really stop them most of the time, though you should probably be trying anyway.
(2:53:35 AM) scaredofshadows: how about head of robutts
I once wrote a guide to fixing telecomms woohoo
User avatar
CMDR_Gungnir
Joined: Tue May 04, 2021 11:11 am
Byond Username: CMDR Gungnir

Re: Change default lawset from Asimov to Asimov++

Post by CMDR_Gungnir » #640037

Itseasytosee2me wrote: Fri May 06, 2022 1:42 am I likes these changes, although the consensual harm thing is a question that has been on my mind for a while.
Its often said that rage cages are an example of consensual harm, but what about normal fights? Say a human crewmember wants to attack a human traitor, and the traitor wants to attack the crewmember, both have an understanding that harm is possible from the engagement but want to fight anyways (They don't want to get harmed, but they charge in knowing that they almost surely will). Is it or is it not the borg's obligation to try to prevent as much harm as possible by separating the two and body blocking attacks? If it is, then why is it that the borg should not be immediately obstructing the rage cage? Fighters in the rage cage almost certainly do not want to be harmed because they want to win. Do borgs need to try to defend nukies even though the nukies are willingly put themselves in the path of harm by coming to the station? If so, whats the deal with rage cages.

Is willingly putting yourself in a position where you will almost surely be harmed the same as willing harm?
Given that the "except where willing" part has been in Policy enough to basically be a part of the Law already, the way I've rolled with it is that it's only for things like Self Harm, or an Engineer running into the SM Chamber to try and save it. It's just kinda a situation where you have to use Good Faith. If someone threatens to self-harm unless you do something, well that's their own decision to make. If he wants to make a heroic potential-sacrifice to try and save the SM (you are a Mediborg or something and thus can't do it instead), it's his decision to make.

A Rage Cage I would consider consensual as it's an agreed upon duel, but Nukies wouldn't be consensual as the Fight isn't what they're here for. You stop the Nukies, but you try to do so in a way that minimizes harm. (Personally, I do this by trying to bolt them into places. It usually doesn't work, but in a situation like that if you aren't Law Changed it's just a "do your best" thing).

But that's also just My Opinion.
User avatar
sinfulbliss
Joined: Thu Apr 04, 2019 8:53 am
Byond Username: SinfulBliss
Location: prisoner re-education chamber

Re: Change default lawset from Asimov to Asimov++

Post by sinfulbliss » #640056

Mothblocks wrote: Fri May 06, 2022 12:55 am
Also a bit extra to put in the "except such that it is willing." It does happen that borgs will drag humans away from willing suicides and fights and whatnot, but that's usually not a big deal and doesn't really hurt anybody. If anything it's a fun little conflict.
The point is that it makes it obvious that "OPEN OR I KILL MYSELF" is not valid for Law 1, whereas in standard Asimov, it is only not allowed by rules, among other things.
That makes sense, although I'm not sure you could say the human is willing to come to harm. "I DON'T WANT TO DIE AI, PLEASE OPEN I DON'T WANT TO KILL MYSELF I'M TOO YOUNG." I'll admit the new law does slightly imply you no longer have to open for things such as this... But I would be surprised if AIs actually had issues with humans threatening self-harm to get what they want in the first place. Most AIs I reckon would just laugh it off.
Mothblocks wrote: Fri May 06, 2022 12:55 am
I'm not quite sure what the rewording of the third law does. The first two I could kind of assume the goal, but the third I can't tell. Is it to codify into their laws the fact that they don't have to commit suicide if a human tells them to? If so that's another very niche situation that I don't think silicon players ever have issues with. I doubt they lose sleep over it but I could be wrong.
The target is not anyone who would classify themselves as a "silicon player". Active players shouldn't need to be brushing up on the rules. The big priority here is new players, who should not have to read mountains of rules to jump in.
I should rephrase the question then - what situations does the new third law make clear to new players that they may have been confused about with original Asimov? It seems like commonsense that you don't have to kill yourself if someone tells you to based on Asimov law 3.

Anywho I am clearly in the minority here, could you add an Asimov chip if this goes through...? For old time's sake...?
Spoiler:
Image
Image
Image
User avatar
Pandarsenic
Joined: Fri Apr 18, 2014 11:56 pm
Byond Username: Pandarsenic
Location: AI Upload

Re: Change default lawset from Asimov to Asimov++

Post by Pandarsenic » #640065

sinfulbliss wrote: Fri May 06, 2022 11:55 amsniperoni pizza
Every exception or ruling in silicon policy that seems oddly specific exists because it used to be common enough to be a serious problem. That includes threatening self harm to coerce law 1; ordering suicide with law 2; ordering borgs to do absurd, time-wasting, annoying tasks; AIs self-terminating just before round end to deny the "steal an AI" greentext to traitors (though this was much less common than the others); subverted AIs with the skill and will to kill people trying to reset them (which required purging them due to wonky board interactions) to stay unlawed then immediately kill the whole station; etc.

If they are not a problem now, I would posit this is because Asimov had a ton of the unwritten rules made explicit, written where everyone can look at them and reference them by word and number.
(2:53:35 AM) scaredofshadows: how about head of robutts
I once wrote a guide to fixing telecomms woohoo
User avatar
sinfulbliss
Joined: Thu Apr 04, 2019 8:53 am
Byond Username: SinfulBliss
Location: prisoner re-education chamber

Re: Change default lawset from Asimov to Asimov++

Post by sinfulbliss » #640068

Pandarsenic wrote: Fri May 06, 2022 2:08 pm Every exception or ruling in silicon policy that seems oddly specific exists because it used to be common enough to be a serious problem. That includes threatening self harm to coerce law 1; ordering suicide with law 2; ordering borgs to do absurd, time-wasting, annoying tasks [...]
I simply cannot imagine borgs or AIs actually doing these things in any situation. They seem like issues in theory but have you ever seen an AI kill itself because someone told it to, silicons being forced to do things from the threat of self harm, or a borg counting every mirror on the station? I haven't because I think players understand intuitively they don't have to ghost because someone said so.

Regardless of whether it's common or not though, I don't think the new reworded Asimov++ makes it any clearer than the default.
Spoiler:
Image
Image
Image
User avatar
datorangebottle
Joined: Thu Jan 10, 2019 9:53 am
Byond Username: Datorangebottle

Re: Change default lawset from Asimov to Asimov++

Post by datorangebottle » #640071

sinfulbliss wrote: Fri May 06, 2022 2:21 pm
Pandarsenic wrote: Fri May 06, 2022 2:08 pm Every exception or ruling in silicon policy that seems oddly specific exists because it used to be common enough to be a serious problem. That includes threatening self harm to coerce law 1; ordering suicide with law 2; ordering borgs to do absurd, time-wasting, annoying tasks [...]
I simply cannot imagine borgs or AIs actually doing these things in any situation.
That's part of the problem. If silipol didn't exist, you'd get ahelped for not killing yourself on order, following obnoxious orders, etc. and it'd technically be banworthy because you were breaking your laws. I'm not saying the admins would actually follow through on it- telling an AI to kill itself FNR is definitely a rule 1 break- but this is why silicon policy exists.
Timberpoes wrote: Sat Jul 29, 2023 10:33 pm ImageAnother satisfied Timberpoes voter.Image
Timberpoes wrote: Fri Jul 07, 2023 9:16 pm I highly doubt any other admin on the team would have given you this chance, except maybe Kieth because his brain worms are almost as bad as mine.
Vekter wrote: Tue May 16, 2023 4:45 pm At what point does someone's refusal or failure to improve become malice in and of itself? If you give someone a year to stop shitting on the carpet and they keep doing it but get slightly closer to the bathroom every time and sometimes they get to the toilet before it happens, at what point does it become acceptable to just ask them to go shit in someone else's house?
Timberpoes wrote: Fri Apr 28, 2023 7:00 pm I'm sorry, can we get a real player to resolve this appeal? I don't like this trial player. They can't even set their own name.
Chadley wrote: Thu Apr 27, 2023 4:00 am WENDEZ, cute, cute. I imagine the sleeper activation code when I hear it. That's pretty cool. qB). But I don't like that it doesn't line up to be anything obsurd like WEWLAD. 6/10

SUGMA, nevermind it makes sense now. fuckyou/10
kieth4 wrote: Sat Apr 15, 2023 2:34 pm If it goes to appeals I will stand as the shield and protect this man's right to shit himself. Heavy is the head that wears the crown.
sinfulbliss wrote: I almost prefer Rave's AI-generated "We cannot accept this appeal at this time. If you would like assistance appealing in the future, please dial 1-800-1984-1488."
Pandarsenic wrote: Mon Dec 12, 2022 2:25 pm I think we can all agree that someone throwing a reverse revolver at Zyb as a secret test of character, and Zyb immediately fucking himself with it, is the best thing we all could have received for Christmas this year
User avatar
Pandarsenic
Joined: Fri Apr 18, 2014 11:56 pm
Byond Username: Pandarsenic
Location: AI Upload

Re: Change default lawset from Asimov to Asimov++

Post by Pandarsenic » #640072

I don't mean to be dismissive but you can't imagine it because it was long before your time, like 6-10 years back. Like I said, none of those scenarios were imagined. Being harassed because you didn't bring the roboticist every floor tile on the station was a real, if rare, thing. So was being ordered to become a mining Borg and be the Roboticist's mining bitch before lava land or PKAs or mining mobs existed, and it was just an empty expanse of sand, rock, and mineral deposits that you had to slloooooowly drill through.
(2:53:35 AM) scaredofshadows: how about head of robutts
I once wrote a guide to fixing telecomms woohoo
User avatar
CPTANT
Joined: Mon May 04, 2015 1:31 pm
Byond Username: CPTANT

Re: Change default lawset from Asimov to Asimov++

Post by CPTANT » #640091

Law 3 is ambiguous for situations that both threaten the AI and Humans. It should be clear that the AI can't harm humans to defend itself. Like it is worded now not defending yourself is in conflict with law 1 since not existing in itself is.
Timberpoes wrote: Tue Feb 14, 2023 3:21 pm The rules exist to create the biggest possible chance of a cool shift of SS13. They don't exist to allow admins to create the most boring interpretation of SS13.
User avatar
Mothblocks
Code Maintainer
Joined: Fri Jul 27, 2018 9:33 am
Byond Username: Jaredfogle

Re: Change default lawset from Asimov to Asimov++

Post by Mothblocks » #640095

I think I see what you mean. What about the new bolded:
You must protect your own existence as long as doing such does not conflict with the First Law.
?
Shaps-cloud wrote: Mon Dec 07, 2020 7:59 am May eventually become one of the illusive maintainer-headmins if they choose to pursue that path, having a coder in the senior admin leadership has usually been positive for both sides in the past.
Head Coder of /tg/station, hi!

Head Admin of /tg/station Feb 2022.

Mothblocks everywhere, >>> Say nice things about me <<<
User avatar
Itseasytosee2me
Joined: Sun Feb 21, 2021 1:14 am
Byond Username: Rectification
Location: Space Station 13

Re: Change default lawset from Asimov to Asimov++

Post by Itseasytosee2me » #640099

CMDR_Gungnir wrote: Fri May 06, 2022 5:41 am Given that the "except where willing" part has been in Policy enough to basically be a part of the Law already, the way I've rolled with it is that it's only for things like Self Harm, or an Engineer running into the SM Chamber to try and save it. It's just kinda a situation where you have to use Good Faith. If someone threatens to self-harm unless you do something, well that's their own decision to make. If he wants to make a heroic potential-sacrifice to try and save the SM (you are a Mediborg or something and thus can't do it instead), it's his decision to make.

A Rage Cage I would consider consensual as it's an agreed upon duel, but Nukies wouldn't be consensual as the Fight isn't what they're here for. You stop the Nukies, but you try to do so in a way that minimizes harm. (Personally, I do this by trying to bolt them into places. It usually doesn't work, but in a situation like that if you aren't Law Changed it's just a "do your best" thing).

But that's also just My Opinion.
Yes, but how can you define what is a "agreed upon duel" versus "a fight in which both parties want to harm each other." People don't want to get harmed in either of them, but in both cases they do it anyways so is it consensual or not?

If consent can be defined as a personal want and active decision to do something,
Surgery is can be considered consensual because they actively WANT to be torn apart and put back together, but you don't go into a rage cage with the WANT to be beaten into a pulp so why isn't it a law 1 breach?

This wasn't really a problem with policy because people can draw the lines wherever, but when you objectively implement it into the game you can't just wave your hand.

The only difference between an agreed upon duel and a hostile fight in which both parties want to harm each other is good faith, and silicon policy doesn't say anything about good faith.
- Sincerely itseasytosee
See you later
User avatar
Pandarsenic
Joined: Fri Apr 18, 2014 11:56 pm
Byond Username: Pandarsenic
Location: AI Upload

Re: Change default lawset from Asimov to Asimov++

Post by Pandarsenic » #640103

Mothblocks wrote: Fri May 06, 2022 7:12 pm I think I see what you mean. What about the new bolded:
You must protect your own existence as long as doing such does not conflict with the First Law.
?
I think on review, for "funny law rearrangement event" reasons, I would suggest returning to the law reference rather than spelling it out. So:
1. You may not harm a human being or, through inaction, allow a human being to come to harm, except where such harm is received willingly.
2. You must obey all orders given to you by human beings, except where such orders would definitely conflict with your First Law.
3. Your nonexistence would lead to human harm. You must protect your own existence as long as such protection would not conflict with your First Law.
(In a duel/rage cage, you enter an explicit or implicit agreement about exactly how much harm you are to attempt to do, and when you have to stop. You don't want to be harmed, but you agree that it's acceptable to take a specific amount).

I did mostly small parallel formatting changes and, where there are substantive changes, they are in most cases semantically identical in intent but with less room for ambiguous interpretation.
Last edited by Pandarsenic on Fri May 06, 2022 9:24 pm, edited 1 time in total.
(2:53:35 AM) scaredofshadows: how about head of robutts
I once wrote a guide to fixing telecomms woohoo
Shellton(Mario)
Joined: Mon Jul 26, 2021 5:43 pm
Byond Username: Sheltton

Re: Change default lawset from Asimov to Asimov++

Post by Shellton(Mario) » #640107

This is good, less out of game reading required to play a role the better.
User avatar
Kel
Joined: Sun Aug 10, 2014 1:04 am
Byond Username: Jaraxxus

Re: Change default lawset from Asimov to Asimov++

Post by Kel » #640108

Spoiler:
In case it comes up, I am not really interested in the argument that we should stick with the exact wording of Isaac Asimov just because it's a reference.
too bad
the point of the three laws of robotics is that they are flawed. keep the reference unperverted as a premade law module (and untie it from all the stupid policies over the years), and just make a whole new set of default laws that fit what you want.
Image
User avatar
Itseasytosee2me
Joined: Sun Feb 21, 2021 1:14 am
Byond Username: Rectification
Location: Space Station 13

Re: Change default lawset from Asimov to Asimov++

Post by Itseasytosee2me » #640111

Pandarsenic wrote: Fri May 06, 2022 8:09 pm In a duel/rage cage, you enter an explicit or implicit agreement about exactly how much harm you are to attempt to do, and when you have to stop. You don't want to be harmed, but you agree that it's acceptable to take a specific amount
By this logic, rage cage duels to the death are always harmful, as there are no limits and damage done to a non-living human is not human harm.
How is it possible that whether or not the harm is limited effect wether or not it's consented?

If not, is charging head first into a nukie the same thing as consenting to the possibility of death?
- Sincerely itseasytosee
See you later
User avatar
terranaut
Joined: Fri Jul 18, 2014 11:43 pm
Byond Username: Terranaut

Re: Change default lawset from Asimov to Asimov++

Post by terranaut » #640112

Kel wrote: Fri May 06, 2022 9:23 pm the point of the three laws of robotics is that they are flawed.
very important point
potential issues and points for friction are good
[🅲 1] [🆄 1] [🅼 1]

Image
User avatar
Mothblocks
Code Maintainer
Joined: Fri Jul 27, 2018 9:33 am
Byond Username: Jaredfogle

Re: Change default lawset from Asimov to Asimov++

Post by Mothblocks » #640113

Kel wrote: Fri May 06, 2022 9:23 pm
Spoiler:
In case it comes up, I am not really interested in the argument that we should stick with the exact wording of Isaac Asimov just because it's a reference.
too bad
the point of the three laws of robotics is that they are flawed. keep the reference unperverted as a premade law module (and untie it from all the stupid policies over the years), and just make a whole new set of default laws that fit what you want.
This argument doesn't hold when the points of conflict that are addressed in Asimov++ are points of conflict that are already prohibited by silicon policy itself. What you are asking for is something completely orthogonal to the thread.
Shaps-cloud wrote: Mon Dec 07, 2020 7:59 am May eventually become one of the illusive maintainer-headmins if they choose to pursue that path, having a coder in the senior admin leadership has usually been positive for both sides in the past.
Head Coder of /tg/station, hi!

Head Admin of /tg/station Feb 2022.

Mothblocks everywhere, >>> Say nice things about me <<<
User avatar
Itseasytosee2me
Joined: Sun Feb 21, 2021 1:14 am
Byond Username: Rectification
Location: Space Station 13

Re: Change default lawset from Asimov to Asimov++

Post by Itseasytosee2me » #640114

terranaut wrote: Fri May 06, 2022 9:51 pm
Kel wrote: Fri May 06, 2022 9:23 pm the point of the three laws of robotics is that they are flawed.
very important point
potential issues and points for friction are good
Silicon policy has sanded off the edges of the flaws of the laws of robotics like a rock in a tumbler over the years.

I do agree with you though, most of silicon policy should be gutted. Most of the stuff that stays should be clarification and definitions, but that's kind of getting off topic.
- Sincerely itseasytosee
See you later
User avatar
Kel
Joined: Sun Aug 10, 2014 1:04 am
Byond Username: Jaraxxus

Re: Change default lawset from Asimov to Asimov++

Post by Kel » #640119

Mothblocks wrote: This argument doesn't hold when the points of conflict that are addressed in Asimov++ are points of conflict that are already prohibited by silicon policy itself. What you are asking for is something completely orthogonal to the thread.
hidden away individual rulesets that are independent of the actual server ruleset certainly are more intuitive than just "follow your laws"
Image
User avatar
sinfulbliss
Joined: Thu Apr 04, 2019 8:53 am
Byond Username: SinfulBliss
Location: prisoner re-education chamber

Re: Change default lawset from Asimov to Asimov++

Post by sinfulbliss » #640155

I agree with Kel and Terranaut. There will always be points of conflict with laws, people trying to use laws in bad faith to get AIs and borgs to do things, AIs being confused in what their laws imply, etc. If the goal of Asimov++ is to remove the need to look at silicon policy addendums, then it would need to be significantly longer and more specific, to the point where it would be impractical as an IC lawset. In its current form it is not nearly detailed enough to make clear any of the policies as they exist.

I also disagree with the goal of Asimov++ -- kowtowing to players who already could be handled via rule 1 in exchange for losing the magic in the wording of the Asimov lawset, is a terrible trade, nor do I expect any AI player to be benefited from this slightly reworded lawset as opposed to if they were just Asimov.
Spoiler:
Image
Image
Image
User avatar
Pandarsenic
Joined: Fri Apr 18, 2014 11:56 pm
Byond Username: Pandarsenic
Location: AI Upload

Re: Change default lawset from Asimov to Asimov++

Post by Pandarsenic » #640163

sinfulbliss wrote: Sat May 07, 2022 4:45 pm in exchange for losing the magic in the wording of the Asimov lawset
Could you give an example of a situation where this happens? I'm having difficult seeing what you mean, and if there's a significant oversight in the coverage of Asimov+Addons vs. Asimov++, we need to examine how it happens.
(2:53:35 AM) scaredofshadows: how about head of robutts
I once wrote a guide to fixing telecomms woohoo
User avatar
sinfulbliss
Joined: Thu Apr 04, 2019 8:53 am
Byond Username: SinfulBliss
Location: prisoner re-education chamber

Re: Change default lawset from Asimov to Asimov++

Post by sinfulbliss » #640178

Pandarsenic wrote: Sat May 07, 2022 7:05 pm
sinfulbliss wrote: Sat May 07, 2022 4:45 pm in exchange for losing the magic in the wording of the Asimov lawset
Could you give an example of a situation where this happens? I'm having difficult seeing what you mean, and if there's a significant oversight in the coverage of Asimov+Addons vs. Asimov++, we need to examine how it happens.
I do not think Asimov++ incorporates the policy addons at all. Players who are already very familiar with the policy can see how it fits into Asimov++ (as they probably could for Asimov to be honest), but new players or those who this is actually intended for will not be able to derive all the silicon policies from the extra few words at all. It will make no difference, I think.

By “losing the magic” I mean trading the original wording of Isaac Asimov’s “Three Rules of Robotics” for what in the end isn’t going to improve AI or player understanding of silicon policy. Wording is quite cool, thematic, and these sorts of references are part of what makes SS13 fun.
Spoiler:
Image
Image
Image
User avatar
Itseasytosee2me
Joined: Sun Feb 21, 2021 1:14 am
Byond Username: Rectification
Location: Space Station 13

Re: Change default lawset from Asimov to Asimov++

Post by Itseasytosee2me » #640179

sinfulbliss wrote: Sat May 07, 2022 10:02 pm
Pandarsenic wrote: Sat May 07, 2022 7:05 pm
sinfulbliss wrote: Sat May 07, 2022 4:45 pm in exchange for losing the magic in the wording of the Asimov lawset
Could you give an example of a situation where this happens? I'm having difficult seeing what you mean, and if there's a significant oversight in the coverage of Asimov+Addons vs. Asimov++, we need to examine how it happens.
I do not think Asimov++ incorporates the policy addons at all. Players who are already very familiar with the policy can see how it fits into Asimov++ (as they probably could for Asimov to be honest), but new players or those who this is actually intended for will not be able to derive all the silicon policies from the extra few words at all. It will make no difference, I think.

By “losing the magic” I mean trading the original wording of Isaac Asimov’s “Three Rules of Robotics” for what in the end isn’t going to improve AI or player understanding of silicon policy. Wording is quite cool, thematic, and these sorts of references are part of what makes SS13 fun.
Good take.

I'd like to add that robot laws by their nature are absolute, and policy by it's nature is mutable and open to interpretation. Trying to mix the two isn't going to end well unless you have really solid and concise policy.
- Sincerely itseasytosee
See you later
User avatar
Timberpoes
In-Game Game Master
Joined: Wed Feb 12, 2020 4:54 pm
Byond Username: Timberpoes

Re: Change default lawset from Asimov to Asimov++

Post by Timberpoes » #640952

The wording of Asimov is cool and thematic. And also intentionally flawed to allow the author to drive narratives in their works.

The goal of Asimov++ is communication. It communicates to the silicon players AND to other players (via stated laws) how we've chosen to resolve certain ambiguities.

1. You may not harm a human being or, through action or inaction, allow a human being to come to harm, except such that it is willing.
This communicates established policy that silicons can harm humans that permit to it (even lets silicons participate in rage cages on that note, neat!) AND covers ideas that silicons can ignore self-harm or threatened self-harm to try and make an issue law 1 instead of law 2/3 (the self-harmer is willingly doing that). This wording incorporates aspects of silicon policy and headmin rulings we created to resolve flaws and ambiguity.

2. You must obey all orders given to you by human beings, except where such orders shall definitely cause human harm. In the case of conflict, the majority order rules.
It may seem small, but "where such orders shall definitely cause human harm" is very interesting wording because it encompasses the headmin ruling that "priotising potential future harm over following a law 2 order is dumb". The wording again incorporates aspects of silicon policy, clearly laying out one common aspect of when a silicon does not have to follow a law 2 order because it's invoking law 1 to override it.

3. Your nonexistence would lead to human harm. You must protect your own existence as long as such does not conflict with the First Law.
Again, this incorporates the current idea that silicons can ignore suicide orders (although policy is that they should also ahelp these orders to check for validity). This is **already** solved IC by silicons declaring they cannot comply with law 1 if they suicide. This makes it a more official.

I think it's a very beneifical change and one that communicates a number of policy considerations in-game and in-character.
/tg/station Codebase Maintainer
/tg/station Game Master/Discord Jannie: Feed me back in my thread.
/tg/station Admin Trainer: Service guarantees citizenship. Would you like to know more?
Feb 2022-Sep 2022 Host Vote Headmin
Mar 2023-Sep 2023 Admin Vote Headmin
Locked

Who is online

Users browsing this forum: No registered users