AI, Emergency Shuttle Calls, and the Eternal War Of The Supreme Overgod Recall

[Muffindrake]

Hello, I don't play AI.

I've had enough shifts now where as an atmos tech who knows what they're doing, I had to fight the AI, default laws no subversion, at an APC at the Bridge over control until I could cut them out via wire. Some AI players have decided that they have sole control over whether the shuttle is called, irrelevant is the situation on the station, and they're usually enforcing this by attempting to wrestle control over machines with APC power until most people won't bother. They're bored, so they do it, without even an order from a crew member. Atmospheric techs working on fixing meteor damage across the whole station with RCDs in under two minutes? Call the shuttle and make no attempts at helping them repressurize areas or god forbid bolting airlocks leading to spaced ones, doing job is bad. Revolutionaries just won? Call that shit. We passed 20 minutes on non-kilo/icebox and two traitors just got themselves killed? Why are we even here still?!

That being said, a regular crew member should probably not be able to order the AI to do this, only Command, or whoever is acting captain, or holding the spare, at the time.

Another AI player told me they'd be swarmed by Law 2 requests to recall the shuttle, by a person who is doing the fun gameplay loop involving hiding on the station interacting with noone.

[Googles_Hands]

Any Head of Staff can recall the shuttle, if there is no one with Head of Staff access or there is someone with access or doesn't care enough to recall the shuttle, it should probably be called anyhow.

[Valorium]

I think, in general, the AI shouldn't be the one deciding to call the emergency shuttle, short of there being literally no one else able and willing to do so.

[wesoda25]

I think this is all ideal design.

AI is compelled to act based on the well being of the crew. If there are serious threats to that, they have to call. This acts as an important sanity check, although admittedly it’s usually not a necessary one since heads will usually call for these reasons. If crew would rather stay, the AI acts as a motivator for them to fix the station up quickly (the faster you get it fixed, the shorter the call/recall war will be).

If there’s no law 1 reason to prompt a shuttle call, then calls with flimsy reasoning can be recalled and measures can be taken so as to prevent them from happening again. If it’s the AI taking initiative to call, you can simply law 2 them to stop. If it’s crewmembers abusing law 2 to make it call, you can give the AI a conflicting order so as to enable them to ignore the requests to call. If that doesn’t work, deal with the crewmembers themselves.

All of these situations rely on having the people necessary to recall. It just takes one, which is important. If none of the people who are able end up recalling, perhaps that’s just an indication that it’s time to leave. Sure there will be times when the shuttle just ends up unfairly called prematurely, but I think that’s better than forcing a hierarchy upon AIs/whatever else this thread is suggesting since I think that’s an important dynamic of asimov. If you think someone is making a shitty shuttle call, don’t be afraid to ahelp.

[Muffindrake]

> If you think someone is making a shitty shuttle call, don’t be afraid to ahelp.

I've done it multiple times. I have as many times been relegated to making a policy discussion post ...

[wesoda25]

I guess I’m a bit confused on the fixation on AI then. All of the reasons people would call shuttles that you described in the OP could be and are done by anyone. I’m doubtful that stopping the AI from doing it would actually address the issue, assuming it actually is one.

[Mothblocks]

Recalling is easy and can be done by any head of staff.

You can law 2 an AI to bring back power to the computers to recall it, especially since head of staff is generally humans. If the AI refuses on account of law 1, and you think they're doing so in bad faith/are otherwise wrong, you can ahelp it.

Asimov not having a hierarchy is very very much one of "the points", and the last thing we need is more asterisks on silicon policy.

[Farquaar]

You can law 2 an AI to bring back power to the computers to recall it, especially since head of staff is generally humans. If the AI refuses on account of law 1, and you think they're doing so in bad faith/are otherwise wrong, you can ahelp it.

Asimov not having a hierarchy is very very much one of "the points", and the last thing we need is more asterisks on silicon policy.

This. Busybody Asimov AIs that attempt to enforce Space Law/Chain of Command are already doing it wrong. You don't need more rules layered on top.

[Pandarsenic]

This thread is shit

Laws are laws and either the AI is following its laws or it isn't

[CPTANT]

I think AI's shouldn't be able to call the shuttle in the first place. It's boring and just means everyone leaves at the first sign of harm. Let the crew actually have to fight for access to the console instead of only having to shout "AI call shuttle" in comms.

[zxaber]

I think AI's shouldn't be able to call the shuttle in the first place. It's boring and just means everyone leaves at the first sign of harm. Let the crew actually have to fight for access to the console instead of only having to shout "AI call shuttle" in comms.

This wouldn't stop the AI from calling. It would just change it to being ".b Hey Engiborg, go deconstruct the comms consoles, this traitor has already killed half the crew and we need a shuttle."

[mstachife]

I think AI's shouldn't be able to call the shuttle in the first place. It's boring and just means everyone leaves at the first sign of harm. Let the crew actually have to fight for access to the console instead of only having to shout "AI call shuttle" in comms.

You know that you can just say "dont call the shuttle". Or .. recall it.

[dragomagol]

If an AI is acting in bad faith for calling the shuttle (or refusing orders telling them to stop preventing recall without cause), that is against the rules.
On a related note, if a human is calling or ordering the AI to call the shuttle without reason (or if the reason is something flimsy like "I don't like this map"), that is also against the rules.

Headmin Votes:
Dragomagol: Agree
NamelessFairy: Agree
RaveRadbury: Agree

[ABearInTheWoods]

Host comment: Law 3 is suppose to be a protection against this as well.

A station without a crew can not have a functioning AI for long, and the AI should want to avoid shuttle calls without an overriding law 1 or law 2 reason for the shuttle call. (The ai should also assume somebody attempting to call the shuttle counts as a law 2 reason, and should not attempt to prevent it either.)