AI on Ships

[GCU Grey Area]

With the new game I have found some thing fun and that is
once can get shares more than one kind of ship.
As swoppoing them is not really a good Idea (ill let you work out why swopping them all for one kind of ship is bad) I sell one set and
then update the ship i have shares left in. (yes you could buy more but whats the fun in that.)

What I like to do is put an AI in the ship, say a Tech 14 one if I can get it that has Intelligence and skills (pilot 1, Nav 1 and Space ship Eng 1 say)
that gives the GM a nice hook for any game.

"Computer says no" or "Dave do you REALLY wont me to do that"

Also you can make it so the majority share holder of the craft has the last say on what the AI will do.

But its also good for a party, if your pilot gets shot up then well
"Hello ship get us the hell out of here!" is a cool thing

and why give it skills of use, why not give it master systems on other fun things like Streetwise hehe.

"Hold on I think that spaceship is following us"

or Diplomacy

"hello cave men this is your god talking!, yes that right the big thing in the sky shooting fire and death at you!"

all good fun.

so what you think AIs fun or not?

 

[rust]

so what you think AIs fun or not?

The starships of my settings usually have AIs with personality simulations,
although these were originally designed with GURPS, not Traveller.

AIs are a nice referee tool, they turn the ships into NPCs with their own
interests and motives (including not being shot at - I once had a ship sur-
render although the characters wanted to fight ...). The only problem to
keep in mind is that an AI can replace the characters' skills, leaving them
not much to do while on board of the sentient ship if this is not handled
with care.

Oh, and one should also keep in mind that an AI needs the ability to modi-
fy its own software in order to be able to learn, and that an AI is probably
the most impressive "hacker" one can imagine. This means that the cha-
racters' control over the AI will be rather limited, no matter what they try
to keep the AI from going its own way - sometimes literally, leaving the
players stranded if their "AI Relations" skill failed them.

 

[AndrewW]

The Twilight Sector setting has AI's that have equal rights with humans.

 

[BP]

The ship can always choose not to use all its skills! :roll:

There is always the older ship AI that has spent too much time in the radiation soup of Jovians...

 

[rust]

There is always the older ship AI that has spent too much time in the radiation soup of Jovians...

If it is a harmless one it only keeps constantly begging for upgrades, free-
zing the characters' bank accounts to keep them from spending the mo-
ney on anything else, selling the characters' speculative cargo to buy a
little more memory space ...

In one of my settings I had a ship AI (named Mac ...) who was such a nui-
sance that the characters finally abandoned the ship and bought another
one, with the result that a now very angry Mac attempted to ambush the
new ship in order to destroy it and the characters. In the end the ships'
AIs came to an agreement, the characters were spaced in escape pods,
and the two ships left together as friends.

So, yes, AIs can indeed be a lot of fun. :twisted:

 

[alex_greene]

"Hi, there! I'm Eddie, your shipboard computer!"

Dear Gods, no ...

 

[DFW]

IMTU @TL15 there is only pseudo-AI. Within its specialty the computer looks/sounds like it is intelligent.

 

[hdan]

I wasn't originally intending to use AIs in my game, though I do have my computers talk to the players, especially when it furthers the "drama":

While docked to a larger ship in a decaying orbit, attempting to rescue the shp's crew:

Computer: "Atmospheric re-entry in 3 minutes. Initiating automatic undocking procedures...."
Pilot: "Override! Guys, get back on the ship!"

The computer does not actually have an intelligence per-se, but it does sometimes act with the sort of "malice" that complex devices sometimes do when you need them to behave.

 

[Captain Jonah]

OTU and outside of hiver space full sentient AIs are a no no. The only one mentioned was Abey one o one who was kept hidden and pretended to be human.

Personality overlays, very detailed response software, logical/situational programing etc will give you a ships computer that can prep for take off, plot courses, and fly you there. But, and this is a BIG BUT, they do not self program, they are not (at tech 15) self aware, they are not virus.

High tech computers functioning as crew or aids will be able to do a great deal but while they may be complex and capable they should be of lesser skill levels in many fields. Computer astrogator yes since the living astrogators are extinct. Computer pilot, yes if you want to go from A to B in a straight line. DO anything complex, combat manoeuvres etc use a human.

Outside of hiver space the three laws of robotics are not a control when computers can reprograme themselves or learn. The logic to break the three laws or its equal in the 3I will be overcome very fast if needed by fully sentient software. Ship is under attack, must take off, crew on ground will die, crew on ship will be saved, 10 crew on ground, 100 on ship, action to save 100 higher logical priority than failure to save 10. Inside Hiver space who knows what they use to control the AIs

At which point your ships sentient AI just logic'd killing 10 of your crew and having done so once can do it again, do you flush your AIs every week and reboot them to stop this happening.

An AI Singularity will not be nice, at the tech levels of the OTU it bought down all of known space.

Twilight has high level AI's as fully sentient with sentient rights and subject to sentient punishments but they still skirt the danger.

Fear the AI wars, when your enemy can turn your own machines against you the end will be messy. :twisted:

 

[alex_greene]

When all your ships look sleek and angular with a red light strip oscillating back and forth to a continuous "whew-whew" sound effect and the computer insists on calling the Captain "Michael Knight," you'll remember my warnings ...

 

[rust]

... do you flush your AIs every week and reboot them to stop this happening.

I had characters who attempted this. They failed to realize that the AI had
long ago deleted the star charts of some remote sectors from the astroga-
tion computer and had installed its core program there. Each time the cha-
racters threatened to reboot it, it begged for mercy, and after each reboot
it "behaved" until the next time it considered it necessary to "override" the
characters' decisions. The characters never found out that the AI had out-
witted them and watched their antics with mild amusement.

 

[AndrewW]

When all your ships look sleek and angular with a red light strip oscillating back and forth to a continuous "whew-whew" sound effect and the computer insists on calling the Captain "Michael Knight," you'll remember my warnings ...

But what if he starts calling him Michael Long¿

 

[alex_greene]

When all your ships look sleek and angular with a red light strip oscillating back and forth to a continuous "whew-whew" sound effect and the computer insists on calling the Captain "Michael Knight," you'll remember my warnings ...

But what if he starts calling him Michael Long¿

Or even Garth Knight? :shock:

 

[rinku]

When all your ships look sleek and angular with a red light strip oscillating back and forth to a continuous "whew-whew" sound effect and the computer insists on calling the Captain "Michael Knight," you'll remember my warnings ...

The real concern is when it forgets its cover and acknowledges an order with "by your command"...

 

[Jhary]

My Player used Exesive Robots and rohnes. Then they decided to Upgrade theire Ship now the ship has not only a AI it has also a Punch of drohne to look inside an treat the Players if they don´t do what th Ship Want. The Ship name ist Apple and the name of the AI is Tux we used the german Robot Suplement.

 

[zero]

I'd use a Limited AI, much like in Cthulhutech, where a computer is programmed to act like an AI, but it is only responding to the owner and its commands, not acting on its own without prior programming.

In Cthulhutech theyre quite useful and the computer can speak to you with a programmed personality and be a kind of pet, but the thing isnt alive, you're telling it to do something and it does so. Theyre mainly used in that setting as bonus actions when piloting a mecha, for example while youre driving it the LAI will fire at targets you decide, or the thing will move the mecha at a speed and direction you dictate whilst you use the weapons (my favourite thing to do, LAI arent good shots :wink: ).

 

[AndrewW]

At least that one did not take the order "push all energy into the drives" as "switch of all other systems". On the plus side 2300AD has inertialess drives. On the negative side dropping shields and point defence in the middle of a battle is bad for the armor belt...

Not to mention cutting power to life support is bad for the biological entities on board if they aren't in sealed suits with life support.

 

[Cyphr]

When all your ships look sleek and angular with a red light strip oscillating back and forth to a continuous "whew-whew" sound effect and the computer insists on calling the Captain "Michael Knight," you'll remember my warnings ...

The real concern is when it forgets its cover and acknowledges an order with "by your command"...

Be really concerned when the ship is more responsive to leggy, blonde fem fatales, asian pilots, and cynical, older priests.

 

[BP]

One of my favorite AI personalities is Multi-unit Electronic System (MESS) from DOOM STAR (by Richard S. Meyers).

It is a neurotic computer that generally runs on the principle of self-preservation - its first lines in the story are "Permission to fire, permission to fire, permission to fire!", on the premise that sure-fire safety lies in the destruction of anything moving before it can become a possible threat. Fortunately, it 'generally' requires verbal authority to carry out its wishes...

 

[rinku]

One of the driving forces behind intelligent agents and AIs in a communication lag setting (such as 3I) would be to address telepresence and decision making issues. If it takes 2 months for a report to get to HQ, and another 2 months for a reply, followed by a further 2 months for feedback about the implementation of the reply to get back to HQ, you REALLY need any help you can get to shorten the decision making process. (And anyone wishing to criticise the Vilani empire - which was limited to J-1, remember - should bear this issue in mind. Bureaucracy is a solution to a problem in its inital form).

For example, a sector wide company that needs consistent middle management decisions might need a virtual vice-president program that can act as an advisor for local branch mangagers, based on the CEO's personality and standing orders. This would be periodically updated from the core source. Actual decision making would normally still rest with the local manager, but as these programs become more sophisticated the trend would be for them to become defacto bosses. Sort of autopilots for business.

Virtual experts, celebrities, loved ones, pets and superiors could all be facilitated in this fashion. Boyfriend 2.0 indeed...