M-SPACE Starship Design

[g33k]

9 hours ago, clarence said:

... But sometimes, especially when reading tech news on robotics and AI, I feel today we might be underestimating what technology will bring in the coming century. I mean, landing a rocket autonomously on a small barge out on the ocean - that should be impossible in my view and we can already do that. And AIs spotting cancer that human doctors can't see; where will that type of functionality take us?

Yeah.  Medical-diagnostic "expert systems" are getting better and better.

I don't know how much you notice "smart" advertising, and various "helpful" technologies; our household uses quite a bit of Google stuff, and it's both useful and rather alarming how much the "Google Ecosystem" seems to know about us...  Despite having various privacy-settings "set."

The "internet of things" (where ordinary devices can be remotely accessed/queried/operated) meets "smart technology" (aka expert-systems / AI):  I envision the day where you can buy (or rent) a pre-set-up "lifestyle" the same way you can buy a car or a home... it includes a car, AND a home, and the fridge is pre-stocked (with the same choices (food, personal-care-products, etc) you made in your prior "lifestyle," (unless you override them)).  When you go on vacation, your "lifestyle" automatically moves food to the freezer or the garbage, adjusts heat/AC/etc; if you move on to a new "lifestyle" it cleans itself, discards food, etc, & prepares for the new lifestyle of the new resident.

You probably recall "Google Glass" and the associated fuss.  Privacy & security concerns were raised (one guy hacked Google Glass to add facial-recognition and profiling software, able to search people he talked to & see their profiles as he chatted).  Did you know that Sony has a patent on a camera built-in to a contact-lens...?   http://www.freepatentsonline.com/20160097940.pdf    And "bionic eyes" are being tested:  http://www.the-scientist.com/?articles.view/articleNo/41052/title/The-Bionic-Eye/ .   

And you have GOT to know that various militaries are looking at "autonomous smart car" tech and "drone" tech, with the sort of target-id / target-acquisition that's already "helping" soldiers find and hit unseen targets, put onto something like an Apache AAH...

Etc...

 

[Joerg]

1 hour ago, clarence said:

I find it very fascinating how early RPGs struggeled to get computers and computer use right.

I don't think that early RPGs are to blame - rather the then prevalent fiction writers' ideas about future technologies in space opera series.

1 hour ago, clarence said:

The way Traveller handled it feels very dated. But sometimes, especially when reading tech news on robotics and AI, I feel today we might be underestimating what technology will bring in the coming century. I mean, landing a rocket autonomously on a small barge out on the ocean - that should be impossible in my view and we can already do that. And AIs spotting cancer that human doctors can't see; where will that type of functionality take us?

Kubrick's 2001 with HAL as pinnacle of cybernetics defined the computer technology used in Traveller, I think.

TV-Tropes discusses this with regard to the agency that gives superheroes their powers (unless you have a metaplot device like in X-Men or Wild Cards) - it is usually the newest badly understood, somewhat threatening topic of science. Mr. Hide used chemicals to become the Victorian Hulk, while Marvel's incarnation gets exposed to radiation/radioactivity.
Any semi-hard to hard SF is especially prone to find a "mule" in its Foundation-like forecasts, a sudden development of an unexpected technology or more often a non-development of a long anticipated technology that renders the technological base it uses quaint, a slightly more modern and certainly less charming variation of Steampunk.

[soltakss]

I like the idea of just giving AIs skills. If you have a MediDoc then you want it to be as good as a Doctor, or maybe a Surgeon, or perhaps a really good Surgeon, you don't want it to be able to land a spaceship. So, you have a collection of small AIs, each specialising in a certain area, effectively running as NPCs.

Most people use computers to do stuff. Not many people program computers to do more stuff and even they work to various limitations when developing new things.

What do people use computers for?

1. Taking out the drudgery from work
2. Speeding things up/Automating things
3. Finding things out quickly
4. Running Simulations
5. Entertainment

So, an AI in a SciFi game can do all these things.

A Medi-Doc can diagnose and treat (2), has access to a vast store of medical knowledge (3), can work out the effects of treatment (4) and means that the actual doctor can be freed up to do other things while the Medi-Doc is working (1).

An Engineering AI checks that all the systems are up and running (1), runs a lot of the systems itself (2), checks to see if it can fix any faults (3), checks to see what would happen if it changed part A for part B (4) and so on.

Computers are not magical things that can do anything you want, they are tools that can do certain jobs. 

[heathd666]

i can see AI's already in the M-Space book, kind of, in the form of apps when building your ship, so mechanically speaking clarence thought of almost everything already

[Joerg]

I wouldn't call specialist systems like an engineering app or a medibot app AIs - these things don't need consciousness to operate within their area of specialisation. I doubt that they are AI-capable (like e.g. the Voyager holodoc or other holograms, beginning with the Moriarty one from Next Generation), either. It's rather cruel to place an emerging AI in a single purpose slot.

The webcomic "Freefall" (http://freefall.purrsia.com/) has an interesting take on emerging AI in robots and synthetic life forms while playing with the three Asimov laws, well worth a binge read (available under http://tangent128.name/depot/toys/freefall/freefall-flytable.html).

In RuneQuest speak, such expert systems could be fixed INT entities, or similar to bound spirits. Ship-operated drones (beginning with a space version of roomba) would be fixed INT appliances, too.

[clarence]

I use skills to portray 'narrow' AI (specialised for just a few tasks) but for self-conscious systems full stats work better. Telling the two types apart may not always be easy though. I believe most sci-fi settings can harbour narrow AIs but that fully consious computers only suit some campaigns. 

As heathd666 mentioned, Apps in starships are expert systems with limited intelligence. They could as well have been defined as skills, but I'm always wary of making objects too PC-like (and I never enjoyed the way HERO system builds starships). 

For the modular droid rules I'm working on, skills are used though. A skill Module (called microModule) is added and contains 3-5 skills. For example, Language is needed for them to communicate with humans and Psychology if they are meant to socialise. The other slots can be filled with whatever's needed. 

[g33k]

6 hours ago, soltakss said:

I like the idea of just giving AIs skills...

...

Computers are not magical things that can do anything you want, they are tools that can do certain jobs. 

I sometimes use the term "AI" in a sloppy, pop-culture sense.  What you're describing is properly in the realm of an "expert system" not an "AI" -- something with a comprehensive suite of information within a specific, clearly-bounded domain of knowledge.  A "medical" expert system, etc.  It can (ideally) fully-emulate "intelligence" (and pass any degree of "Turing Test") within its domain.

A true "AI" shows intelligence -- it's "artificial" not biological, but it's intelligence -- which includes the ability to generalize across disparate domains of knowledge, and extend into new ones.

Within a future-tech sci-fi setting, I expect lots and lots of "expert systems," but the issue of "AI" is kind of a key question -- are they inherently limited to a humanocentric scale, and suitable for "protagonist" roles (PC in a RPG, main character in a story, etc)?  Do they become "more than human" -- smarter, more informed, etc?  Do they proceed directly to Transcendence?  Do they scale everywhere in between?  Or ... ?

[soltakss]

6 hours ago, Joerg said:

It's rather cruel to place an emerging AI in a single purpose slot.

What is the difference between an emerging AI acting as a Doctor or a Pilot, as opposed to a human acting as a Doctor or Pilot? They both do the same job day in, day out.

[soltakss]

1 hour ago, g33k said:

I sometimes use the term "AI" in a sloppy, pop-culture sense.  What you're describing is properly in the realm of an "expert system" not an "AI" -- something with a comprehensive suite of information within a specific, clearly-bounded domain of knowledge.  A "medical" expert system, etc.  It can (ideally) fully-emulate "intelligence" (and pass any degree of "Turing Test") within its domain.

Probably.

For a set of SciFi rules, do we really need different levels of rules for different types of computer?

 

A true "AI" shows intelligence -- it's "artificial" not biological, but it's intelligence -- which includes the ability to generalize across disparate domains of knowledge, and extend into new ones.

For a starship, would you put a true AI anywhere near it? It seems to me that a set of expert systems, perhaps controlled by another expert system, would be a far better fit for a starship than a true AI.

Reminds me of the Dilbert cartoon where he creates a robot, who says something like "I'm better than you, stronger than you, more intelligent than you and I can live forever! Ha, ha,ha,ha! I'm bored ..."

 

Within a future-tech sci-fi setting, I expect lots and lots of "expert systems," but the issue of "AI" is kind of a key question -- are they inherently limited to a humanocentric scale, and suitable for "protagonist" roles (PC in a RPG, main character in a story, etc)?  Do they become "more than human" -- smarter, more informed, etc?  Do they proceed directly to Transcendence?  Do they scale everywhere in between?  Or ... ?

 

Honestly, I can see them doing all of the above.

An AI in a stationary fixture could have a robotic drone that it can use to move around with, thus making it like a PC. It could also be planning strange things, using its huge intelligence, but cannot actually do anything due to its limitations.

Asimov had a story where a Supercomputer, in effect, was built, then slowly moved into Hyperspace, as it was faster there, it processed so much information and became omniscient, then when it approached the heat-death of the universe it looked for a solution, the solution eventually being "Let there be Light".

 

[g33k]

2 minutes ago, soltakss said:

For a set of SciFi rules, do we really need different levels of rules for different types of computer?

Well, an "expert system" is basically just a "fix this problem" tool.  Want to elide the healing-cycle from game-play?  AutoDoc & done.  Anything where you need a problem solved because the setting demands it, but don't want to make it a PC's activity, apply the "expert system" handwave...  

Also for setting-fluff / colortext -- e.g.:  "autopilot for all vehicles, because... sure, someone can learn to operate a vehicle manually, but WHY, when biologicals have slower reflexes, worse multitasking, fewer channels of data-input, and are demonstrably and measurably WORSE drivers/pilots/etc in all circumstances?"

But if you want a playable PC, you obviously don't want "Expert Systems" limitations!

And if you want PlotDevice levels of capability (e.g. Terminator's SkyNet), you want even fewer limits...

So if you want all 3 present in your game, I'd suggest you need "Expert System" rules, maybe just use ordinary "PC" rules + "Robot/Droid" rules for that scale, and minimal rules with lots of handwavium for PlotDevice AI devices...

YMMV

[Joerg]

1 hour ago, soltakss said:

What is the difference between an emerging AI acting as a Doctor or a Pilot, as opposed to a human acting as a Doctor or Pilot? They both do the same job day in, day out.

The difference is that a human has a life, doing something besides his job.

[soltakss]

On 3/15/2017 at 8:28 PM, Joerg said:

The difference is that a human has a life, doing something besides his job.

Perhaps you can explain that to my wife, she has the opposite view when it comes to me.

[Joerg]

On 15.3.2017 at 9:28 PM, Joerg said:

The difference is that a human has a life, doing something besides his job.

 

29 minutes ago, soltakss said:

Perhaps you can explain that to my wife, she has the opposite view when it comes to me.

Ah, but you are already doing something besides your job - listening to your wife's complaints, and all the other fun stuff (like putting out the garbage) resulting from married cohabitation. Nothing like that for an emergent AI stuck in a routine job.

(I noticed that when one is told to "Get a life" that often means little more than "do your chores and stop bothering me"...)

Edited March 17, 2017 by Joerg
grammar

[Alex Greene]

The discussions on here are just as fun as they are on any Traveller board.

Suddenly, I feel right at home here.

[g33k]

So ... let's drag this back on-topic...  

Could someone do "sample builds" for the really iconic ships from sci-fi?

• The Enterprise (both Kirk's & Picard's)
• The Millenium Falcon
• Serenity
• And let's not forget some Bad Guy ships... including the Death Star!  Also Boba Fett's little ship.  Klingon ships, and Romulan ones.

And you can stop sweating in fear:  not gonna ask for the Tardis (also not bothering with Discovery One).

[clarence]

Ok, here's the Millennium Falcon, straight from the rules:

 

[Baragei]

While we're on the subject of the Death Star. I suppose that one would classify as a capital ship? Apart from Force-guided photon topedoes, how do you damage capital ships? Not kill them, but damage them. Am I missing something in the rules, because I can't find any other way than to chip away at their insane number of hit points?

[clarence]

All you need is a big ship with big cannons. Check out the rules for weapon upgrades to increase damage for a canon. Very simplified, the more Modules a canon has, the more damage it can do.

There are no rules for construction weaknesses though (like vents, where tiny starfighters can fire one missile to destroy the whole ship). Those are left to the art of storytelling. 

Edited March 18, 2017 by clarence

[Alex Greene]

4 minutes ago, clarence said:

All you need is a big ship with big cannons. Check out the rules for weapon upgrades to increase damage for a canon. Very simplified, the more Modules a canon has, the more damage it can do.

There are no rules for construction weaknesses though (like vents, where tiny starfighters can fire one missile to destroy the whole ship). Those are left to the art of storytelling. 

"Okay, Fred, it'd be a Herculean task for everybody else, but the Force is strong in you, so it's Very Easy. Just roll the dice, will ya? Any roll but 00 is a success. 00 is a fumble."

*rolls 00*

"The Force must be taking a holiday today, Fred ... WATCH THAT WA - oh, never mind."

Edited March 18, 2017 by Alex Greene

[clarence]

Let's just hope Fred wasn't out of Luck Points that day...

I put together some quick stats for Serenity in a thread at RPG.net recently:

https://forum.rpg.net/showthread.php?799151-BRP-Space-M-SPACE/page2&highlight=m-space

Not sure how correct they are, but reasonably close I think.

Edited March 18, 2017 by clarence

[...]

So I read of passing reference to the need of refueling the hyperspace engines, except no reference to what the fuel is? Myself, I'm thinking of D-T Fusion being "standard" in my setting, I was just wondering if there was something else in mind. Plus I am wondering how maybe to make a ship such as the Nostalgia for Infinity from Revelation Space, it's big, real big, except not much crew.

[heathd666]

@dragoner im not familiar with DT fusion reactor im guessing its from Revelation Space but i am familiar with the Epstein drive from The Expanse books which sounds like its familiar. here is an interesting page for it.   http://expanse.wikia.com/wiki/Epstein_Drive            in the books they mention use of fuel pellets for it. hopefully this may help with your inquery. M-Space i think is more of a toolkit for sci fi games atleast thats my take on it. sort of like the big golden book from chaosium. meaning that you add whatever flavor you want for your soup i mean setting lol. you want fusion drives that run off some sort of fuel whether it be pellets or liquid fuel or whatever make it happen  probably not what you were looking for but its my 2 cents worth

[clarence]

Fuel is left open intentionally, to allow the GM to fit it to his campaign. There is a simple system in REFLUX though: the hyperdrive needs recharging, either at a starport or through the efficient meta-material solar panels all ships are equipped with.

I'm not very familiar with the Nostalgia for Infinity. Can you tell me more about it?

[...]

heathd666, D-T Fusion, or Deuterium-Tritium Fusion, is:

The most promising of the hydrogen fusion reactions which make up the deuterium cycle is the fusion of deuterium and tritium.

http://hyperphysics.phy-astr.gsu.edu/hbase/NucEne/fusion.html#c2

I love the Expanse though, it's great.

clarence, the Nostalgia For Infinity is the main ship from Revelation Space by Alastair Reynolds, a very good space opera novel, here is a wiki entry: http://revelationspace.wikia.com/wiki/Nostalgia_for_Infinity

Thank you for clarifying what fuel is, I will look at Reflux to see how it is handled there. Some of my questions are just for pre-designing spacecraft, as fuel should be included in the modules maybe, at least for a homebrew universe/setting.

[clarence]

I've read parts of the first book but I have very vague memories of it. With such a complex ship, I really can't say much without knowing more. Below are some initial thoughts.

The size versus crew size should work out fine as long as there are AIs doing the work. The Size Rating (SR) table would need to be expanded a bit (if 160 000 passengers is a reliable figure):

SR 19: 262 144 Modules

SR 20: 524 288 Modules

SR 21: 1 048 576 Modules

That’s a lot of Modules but with the way Capital Ships are constructed, it’s not really a problem. Also, I would probably divide the passenger areas equally between cubicles and open space.

Ablative armour is easy to model simply by lowering the armour value with each hit. How well ice will perform depends on the weapons used in the setting though.

Wayward AIs and body swapping are not covered in the rules at all. My only recommendation is to run AIs as regular NPCs and give them a few unique ‘powers’ and restrictions.