Robots

[kafka47]

At several points in Traveller cannon, Robots get mentioned. When they are they appear as some sort of Asimovian Frankenstein or at least, a slave race. Or something far greater, that is being an equal partner to sentient life, although, in Cleon's words, never a citizen...

Yet, looking over the official construction rules and what has been mentioned about robots. They seem little more than automated vaccum cleaners. Can we expect to see more of Robots in T5 and T20?

 

[robject]

It's very possible. Actually, I'd like to open this thread up to people to get their opinions on how they think Traveller should handle robots.

What role do you see robots playing in the Imperium? Are there several categories? Is there a taxonomy? Are there rules to differentiate them?

I wanna know.

 

[mike wightman]

Robots could play a much larger role within the Imperium, and are probably there as another "background" tech.
Traveller canon is quite clear that the 3rd Imperium has a great social prejudice against the use of robots, so IMHO they would be hidden in plain sight - i.e. they will still be there, just not necissarily as "robots".
They will be disguised as automated manufacturing machines, autodocs, autopilot upgrades.

As for types and taxonomy, they are given in LBB8. The DGP supplement 101 Robots isn't a bad guide either

 

[veltyen]

(Referring to T20)

The main problem is that computers are dumb. Getting towards something able to replicate a human intellect requires an enormous computer, with high TL programming (for low AI). Even then you are stuck with a level 0 character, who gains XP like they are brain-damaged.

Anything less then an AI is just a machine. Not exactly what I would call a slave race, sure they might be relatively helpful, but not near as useful as another competent person. A voice controlled toaster is a gimic, rather then a citizen.

So, until you start looking at TL17+ there really isn't anything matching human ability, and there certainly isn't anything even able to stretch to being called an android. The cloest you can get is a human mimic, that if well programmed could be able to pass as human for *ahem* state events.

There are still places where a machine intelligence could be useful. For autonomous logic you have the advantage that loyalty can generally be assured if no-one has access to re-program the machine after initial programming. While the system that the intelligence inhabits has to be large, it is only large on a human scale, compared to starships and buildings they are relatively small. Automated factories, starship systems (such as Zen/Slave from Blakes7 and Holly from Red Dwarf), even systems as simple as those controlling telephony and communication could have extremely intelligent (for a machine) and capable systems.

On the military front using robotic spacefighters, grav tanks and robotic battlepods has certain advantages, especially if they could be remote controlled as well. This would allow you to use the superior human abilities when available, while still being capable of completing simple missions when communication is down or to go on suicide missions without effective loss of expertise.

So robots walking arround and being just more people is certainly unlikely.

However truly stupid robots also have their place. A partially robotocised infantry force, with many human NCO's and simple humanoid robots able to mimic and follow simple commands is a definite possibility, and quite useful if they could share equipment. Larger robots (ED209:Robocop) could be somewhat more independant, and have duties where eternal vigilance and no bladder would come in handy (simple patrols, guard stations).

Where there are a lot of people arround robots should be rare. People are quite cheap, and get all uppity (read riot) if a lot of them are unemployed. In industries where humans can break easily such as smelting and mining there may also be a large ammount of automation, though if the labor is cheap enough sentients would still be used.

The one big advatage that machine intelligences have is recall. A dumb lawyer who can cite ALL precendents would probably be better then a smart lawyer out of their normal juristiction. A corporate voicemail system able to remember everything, and be the entire corporations personal secretary.

IMTU there is both a robotic terrorist organisation (SELF- sentient engine liberation front - gratuitously stolen from UFO) and a large number of rogue high AI's, who tend to be very careful about upsetting the powers that be. There are also several AI run and controlled religions.

Enough rambling

 

[Flynn]

QLI has the manuscript for TA: Robots of Charted Space, which contains a lot of information on robots in the Gateway Era, as well as 31 robot write-ups using the T20 design system. They've been looking for art and such for about a year now for the supplement, so I think it's probably on the back burner with everything else that's going on right now.

Once it's released, its companion piece, Robot Adventures, should come out soon after, which contains something like sixteen patron encounters, four mercenary tickets and two amber zones, as well as a few more robot designs. In the second manuscript, I tried to explore a lot of different scenarios that involve robots in the OTU. I believe there's a lot of diversity in the roles that robots can fulfill for plots and adventures, and made a strong effort to demonstrate that in both manuscripts.

Hope this helps,
Flynn

 

[RobertFisher]

Originally posted by robject:
What role do you see robots playing in the Imperium?

For me, it would be much like robots in our own society. Robots are used a lot, but you almost never see or recognize them in your daily life.

I personally feel that true-AI is nigh impossible, so I have a tendency to consider it something to be left to the magic TLs. You can have some psuedo-AIs, though, that are good enough to fool a lay person as long as there are some limits to the interaction.

Are there several categories? Is there a taxonomy? Are there rules to differentiate them?

All questions I haven't had a need or desire to think about too much yet. (e.g. I've only ever skimmed Bk8 so far.)

 

[mike wightman]

The robot brains in LBB8 can be of a maximum Int equal to TL-3.
Throw enough money at your robot design and you hit this Int ceiling:
</font><blockquote>code:</font><hr /><pre style="font-size:x-small; font-family: monospace;"> TL Int
8 5
9 6
10 7
11 8
12 9
etc.</pre>[/QUOTE]The point being that in CT at TL 10 and above the robot brain is of human level Int.
Now add the fact that a robot can begin with a skill level of 4 and you have to wonder why it's only the Hivers that built robotic combat vehicles and starships
A robotic crew could begin with level 4 in all relevant skills (and learn by experience at higher TL's) - and take up less space in the ship/vehicle thanks to the life support saving (no staterooms???)

 

[far-trader]

Yeah, I always found that Int equal to TL-3 too generous given the Imperial stance on robots. As I recall I had an easy fix, a little pencil work and that TL-3 becomes TL-8

Changing your robot design Int ceiling to:
</font><blockquote>code:</font><hr /><pre style="font-size:x-small; font-family: monospace;"> TL Int
8 0
9 1
10 2
11 3
12 4
13 5
14 6
15 7
16 8</pre>[/QUOTE]So now the Imperium isn't building truly human "brains" till the upper TL and pretty recently. It's cutting edge research material, like Aybee and the AI on the the Kinunir.

I'd also limit initial skill levels* to equal to or less than Int. (* total of all skill levels was the way I had it iirc, I think Aybee exceeded this by learning)

So no practical robotic crew or ships until the later TLs though perhaps some limited applications earlier.

But that's just some old mtu hacks. And of course it applies only to the Imperium, other polities had different priorities and were more advanced.

 

[aramis]

Naw, y'all have it wrong. It's TL÷3...
TL 3-5 Int 1
TL 6-8 Int 2
TL 9-11 Int 3
TL 12-14 Int 4
TL 15-17 Int 5
TL 18-20 Int 6
TL 21-23 Int 7
TL 24 Int 8

 

[far-trader]

Originally posted by Aramis:
Naw, y'all have it wrong. It's TL÷3...
TL 3-5 Int 1
TL 6-8 Int 2
TL 9-11 Int 3
TL 12-14 Int 4
TL 15-17 Int 5
TL 18-20 Int 6
TL 21-23 Int 7
TL 24 Int 8

Ya know I was laughing at first (at the idea of TL3 robots) but on reflection that's not a bad progression at all. At least for my perception of robotic intelligence in the Imperium.

 

[robject]

Originally posted by Sigg Oddra:
Robots could play a much larger role within the Imperium, and are probably there as another "background" tech.
Traveller canon is quite clear that the 3rd Imperium has a great social prejudice against the use of robots, so IMHO they would be hidden in plain sight - i.e. they will still be there, just not necissarily as "robots".
They will be disguised as automated manufacturing machines, autodocs, autopilot upgrades.

As for types and taxonomy, they are given in LBB8. The DGP supplement 101 Robots isn't a bad guide either

Sigg and Robert are correct. Robots run in the background, ubiquitous but effectively invisible.

And as far as AI goes, my own opinion is that there has to be a strong divide between computers, which are fancy calculation engines that must be programmed, and artificial intelligence, which is a massively associative, parallel system which must be taught. Computers are precise and inflexible. Intelligence is imprecise and flexible.

 

[aramis]

I've toyed with AI's, inspired by various works...

All in all, Humanism overtakes me. I actually use the TL-3 limit in my games. But, for my new homegrown TU I've been starting to tinker with, AI is one of the things that will be right out.

Now, as for Int, since Int is the ability to cope with the unexpected AS WELL AS processing data AND INTERPRETING data, I'l use the TL/3 for "Responsive Intellect" and the TL-3 as "Computational Intellect".

Geeze, that's looking like a GURPS-style split stat... ick!!!

 

[mike wightman]

Why not cap the robot's intelligence based on the law level of the planet, within the Imperium only.

Something like max int = TL - LL (where LL = 3 or local law level, which ever is higher)

 

[mike wightman]

I've just been reading TNE's Vampire Fleets.

I'd forgotten that there is a robot design system that ties in with FF&S in the book.

The robot brain table gives a more moderate value for robot Int:
</font><blockquote>code:</font><hr /><pre style="font-size:x-small; font-family: monospace;"> TL Int CTInt
8 1 2
9 2 3
B 4 5
C 3/5 4/6
D 4/6 5/7
H 6/7 7/8
J 8/9 9/A </pre>[/QUOTE]At TLC and above the number before the slash is the Int of the linear-parallel processor based brain (larger and cheaper), while the number after the slash is the Int of a brain with synaptic processors (smaller but more expensive).

 

[Fritz_Brown]

Since I'm banging on robots again, I will cast a Thread Resurrect! here, and say, that you need 15 parallel and 8 synaptic CPUs to max out a TL15 robot on Int. This is expensive!!! (Especially at 50kCr per CPU - 8 of these costs more than everything - including fusion plants - except a FGMP-15! )

How in the world do you use FF&S for robots? It appears you can't get a "body" less than 14m^3!

 

[aramis]

Since FF&S will make sub-1Td craft... just don't look under starships... look under vehicles.

 

[Fritz_Brown]

Which FF&S, Aramis? FF&S1 has the smallest vehicle hull as 1dT/14kl. Even my "wheel"chair is only 1000l. (That's under "Ground Vehicles", BTW.) I could interpolate, but there's no smaller size to work with.

 

[darktalon]

Surface area should correlate to the two-thirds power of the volume, and since material volume is related to surface area in FFS1 it should have that progression. However, I'm at a loss as to exactly how the numbers in the book were calculated. Best I could manage was a power curve which I derived in Excel:

y = 0.329 x^0.648

in which y is the MV from the table on p.11 and x is the volume in dtons. The results it gives approximately model those in the book, but nowhere near perfectly - but it might be a starting point if you want to figure MV for something smaller.

Repeating the analysis with the exponent at the 2/3 geometry tells us it should be:

y = 0.274 x^(2/3)

but this gives even less agreement with the numbers in the book for small things (spot-on for your 1,000,000 dton superdreadnought though, and not bad for anything down to about 60 dtons). Just use either of those as you fancy.

The second, incidentally, is pretty close to what you get if you derive the expression for material volume from first principles (considering that just about anything big enough to be designed using FFS is big enough for the material volume = surface area x 1 cm approximation to apply).

 

[darktalon]

As an addendum, the only way I could manage to get the 1-3 dton material volumes to fit was to use them as the only data points. This gave me:

y = 0.388 x^0.489

Use this if you want to be consistent with the book regarding small things.

 

[Fritz_Brown]

Thanks, darktalon.

Incidentally, I was trying to get the MV values by calculating a sphere 1cm radius larger or smaller than the one required for the given volume. I then tried a box 1cm larger on each side. You don't get the answers in the table by doing the two volumes, then subtracting the "empty" volume of the smaller object.