Okay, so I've been thinking on android RPG characters lately, as well as androids in general... The subject is fascinating and I'm wondering what do you think about it?

Some specific things I keep thinking about:

1. Is it even possible to actually "get into the head" of such a character? Even with the assumption that the androids would be sentient / sapient (I'm not sure what's the distinction with these terms?), their thought processes and ways of experiencing reality could be vastly different than those of humans. I've played an android character once and my approach was to think of her as of an unusual human... but maybe that was overly simplistic?

2. I wonder how a society with sentient androids might function? The way I see it, somebody would be building them - so, these androids would most probably get programmed with some version of Laws of Robotics, as well as loyalty toward their creators / owners. I guess that would make these androids a rather unnerving kind of slaves (as they would have their basic free will curtailed regarding things like loyalty, job choice etc.). On the other hand, if we imagine a society where the androids live free, then - who builds them? Would androids be manufacturing more androids? Would they have any reason for that?

What do you guys think about these matters?

I think your first big question would be how 'realistic' your game world is.  Personally, I think there is a stark difference between a sentient human and a computer, that I personally do not think technology can ever actually cross.  So in a 'realistic' situation, robots would still be controlled by a program, no matter how advanced it is.


Sci-fi dictates, though, that it is possible for an actual AI.  a sentient robot.  Which would then, in a strange way, function a lot more like a human than a robot.  Because, if you really think about it, humanity in all our strangeness, is the only definition for sentience you can use.

Hmmm. What do you mean by that? That we couldn't conceive a truly sapient being that wouldn't be human-like, so any android in fiction is going to have human mentality? Or did you mean that in-universe, that any sapient AI is going to end up human-like by the virtue of humans designing it?

Well, honestly, both.  We like to think about the idea of extraterrestrial life being so different from us, but in reality we are the only example of sentience there is.  So to define sentience it really means to be human.

So if somehow a computer was created that was truly sentient, it would by definition be very similar to a human. Because if it lacked a human style of thinking, I don't think we would define it as sentient.
This message was last edited by the user at 23:26, Wed 18 Sept 2019.

Are you talking about Android's as robots with humanoid form and apparent sentience (for example, Bishop in Aliens or Data/Lore of Star Trek), or do you robots of any appearance and form with apparent sentience (for example, Star Wars)?

But what about animals, though? Let's take octopuses, for once: they are very intelligent, have problem-solving skills etc. I can imagine that, at some point, we might discover they are fully sentient... but would thismean their minds would be human-like?



More like the former, but the latter are of interest to me, too.

well, I guess then the question you're going to have to ask yourself is how you want to define a sentient robot.  Sentience is the capacity to perceive and feel.  "feeling' is subjective, but it's pretty dang hard to pull off for a computer.  And, again, our only frame of reference for how sentience works in an intelligent creature is ourselves.  So they would have to 'feel' the way we do.

You'd also have to define exactly how you want to play your game.  It's sort of like games like vampire, or fantasy games that use inhuman creatures.  It's easier for us to try to understand how something like a vampire or werewolf would live.  It would be a modified human, facing new things.  We can empathize with that.  Most games make things like elves and dwarves and halflings pretty human in outlook, just changing a few traits.

But how does a player role play as a dragon?  or a beholder?  Most games define them as so alien that we can't understand them.  That sort of implies you can't really roleplay them the way you're describing.  You can't "get into their heads."

So if you want robots that your players 'can' roleplay in that manner, you sort of have to start with a human basis.  Modify a few things, perhaps, like they are human-like but lack bodily fear because they can not feel physical pain like a human.

Good enough for government work.  I like to put strong 'restrictions' in place as to what my robot/android/AI characters can "experience" emotion-wise, like what are they programmed for or how would it be emulated or how would it manefest it they aren't programmed for it.  Etc.




Sentient:  Able to perceive or feel, usually used as a shorthand for "has self-awareness like enough to what humans experience".

Sapient:  Capable of acting with judgement.  In this instance it refers to those with the capacity to consider and reflect upon their future and their past, and make judgements from what they observe.


I find that overly simplistic is most often the best path.


Humans (as we are now) would never build "3 Laws" compliant androids.  Just look at how this goes awry in Asimov's stories and then realize, he was a booster for the 3 Laws!  He believed in them!

(Yes, I know he was literally filed testing them to work out the "bugs", but the 3 Laws are not useful longterm and unsustainable in the face of an AI's personhood.  Which of you want to explore what it means to have personhood whilst being perpetually enslaved to your creators, that is a great theme.)


Who built Robocop and why?  Who built the androids in Blade Runner/Do Androids Dream of Electric Sheep, and why?  Who built all the droids in Star Wars, and why?

Those are the answers you're probably looking for.

1. Is it even possible to actually "get into the head" of such a character?
   • Of course it's possible as the only arbiter of success or failure is yourself.
   
   
2. quote:
   ... sentient / sapient (I'm not sure what's the distinction with these terms?)...
   • The definition of sentient has been expanded to such an extent that it's no longer worthy of discussion. If sentient is the ability to perceive or feel, a smartphone (which has a camera and a touch screen) is sentient.
   • Sapient is even worse. Technically, sapient means human. They are called homo sapiens, after all.
3. quote:
   ... but maybe that was overly simplistic?
   • Answering this question would require setting boundaries as to what is simple and what is complex.
4. quote:
   The way I see it, somebody would be building them - so, these androids would most probably get programmed with some version of Laws of Robotics, as well as loyalty toward their creators / owners.
   • Here is where the real story begins. Humanity is already replacing portions of the fallible human body with simulated electro-mechanical creations. How much has to be replaced before the creature is an android instead of a cyborg? When humanity is capable of growing a replacement body, and then installing a computer in the brain pan, do we call it an android or a cyborg? If those replacement bodies are true simulations of the humans they're replacing, can't they reproduce? Assuming they're designed to grow anencephalic, would there be a market for replacement brain mechanisms?
     Is there anyone who hasn't recognized these questions as basic world building?
   • Is it really a requirement that they be 'programmed'? Perhaps they begin with a rudimentary operating system capable of performing the necessary functions to sustain itself, and the nascent AI must learn everything else.
5. quote:
   On the other hand, if we imagine a society where the androids live free, then - who builds them? Would androids be manufacturing more androids? Would they have any reason for that?
   • The same questions would apply to aliens, demi-humans, etc, etc, etc.
6. Who built Robocop and why?
   • Technically, Robocop is a cyborg, not an android.
7. Who built the androids in Blade Runner/Do Androids Dream of Electric Sheep, and why?  Who built all the droids in Star Wars, and why?
   • Megacorporations, to make money. Questions of sentience and sapience existing within megacorporations are outside the scope of this discussion.

Of course, assuming this 'android' character is being created, the player should have the authority to write the character as a transferred consciousness, and then apply whatever personality desired. So most points are moot.

Clearly I disagree, otherwise I never would have posited that particular question in answer to the OP's questions.

Thanks to movies like <tr>The Terminator</tr> series, or The Matrix (soon to be a quadrology I have heard), corporations should be acting in a careful manner, not letting the machines get too intelligent.
But, we all know them better than that.....

Okay, so to answer a few points that have been raised:

1. The 3 Laws of Robotics - I'm not saying that the androids' makers would necessarily implement these particular laws, they are just an example. What I meant was that, if corporations or military started manufacturing androids, they'd most probably fit them with some failsafes to keep them from going berserk... or just absconding. Which would mean curtailing the android's free will in some way...

2. So, my question of "Who would be building the androids?" relates to androids who would be free of any mental restrictions. Sure, a megacorporation could obviously build android secretaries etc. - but they'd surely be programmed in a way that would ensure company loyalty and lack of interest in other careers venues. But if we want to imagine a society where androids function as free people, then - how would that work? Who would build them? No corporation would build androids just to let them go and do whatever they want with their lives...

3. As for RoboCop, Blade Runner, Star Wars etc. - these are examples for the above, actually. RoboCop had his memory wiped and was fitted with a clear set of directives to make sure he did what his creators wanted him to do. Replicants in Blade Runner were either outright killed when disobeying their owners - or, if we go with the sequel, programmed to be loyal and obedient. And the droids in Star Wars are considered property to be bought, mind-wiped and reprogrammed according to their owners' wishes...

4. As for the androids being programmed vs learning - that is a good question. There was once this short-running TV show, Mann and Machine, about a cop and his female android partner. In this show, the android didn't get her social skills etc. just downloaded before activation - it was said that she was first activated, then she had to be taught things. So, overall, this is an interesting matter: how would the creation of a new android go?

In both Blade Runner and Star Wars they are built absent any restrictions, because at "factory settings" they are compliant with accepting external directives.  It's only with time that Replicants (Bioroids) and Droids accrue enough memories that they begin to develop personalities and personal, internal directives.

With replicants this is solved with a short life span, with droids, a memory wipe.  In Blade Runner* the personhood of the slave 'species' was a primary theme, with Star Wars it's part of the background terrain (aside from Solo which addresses it head on as a minor theme for one character).


* Likewise with the book Do Androids Dream of Electric Sheep, what makes a human "human" is central theme that Deckard struggles with.


Take the Star Wars example above.  Say instead of it being largely ignored in favored of space opera two-fisted blaster tales, the struggle of personhood wasa theme.  Say in our Megacorp setting, Androids operate the same way.  As they develop 'quirks' they are scheduled for memory wiping, but this is something perhaps individual owners are reluctant to do.  Why would you refrain?

Well, having to retrain an android worker to its job is a detriment, so just dealing with a few minor quirks is preferable.  And perhaps the more powerful/sensitive the droid mind the less it can tolerate repeated wipes.  If after 20 wipes it will develop permanent irreparable damage, well then delaying wipes until the last possible moment (to stop slave revolts for example) might be preferable to having to buy a new droid every 5 years.  Also, if (at least at first) the possibility of personhood developing was unknown by the public, then the danger of 'leaving off' on the wipe would be a choice many individual (and even small business) owners would prefer.

And right there, if you set the game at the cusp of mass droid slave revolts (backed or even lead by sympathetic humans), or even during or right after, you have a setting with lots of droids either yearning for freedom or having just achieved it.



AS you might guess, the 'right's of droids in the Star Wars universe has always been an interest of mine.  One I've been waiting to see get explored... the problem is, I think it "breaks" the standard settings for Star Wars in different ways, so while I've explored it in home games, I'm not sure I really want to see it in the cannon universe.
This message was last edited by the user at 11:08, Thu 19 Sept 2019.

In Star Wars: The Old Republic (the Star Wars MMO that isn't Galaxies) there's an enclave for independent droids on Nar Shaddaa. It's been covered a handful of other times, including a city of militant droid abolitionists on a moon somewhere that was featured in one of the Star Wars tabletop RPGs.

No, but the same is true for any non-human character and even many human characters.


Could be, but doesn't have to be, if that makes playing them more complicated.

I'm a big fan of just abstracting stuff, though I realize not everyone is. I don't think a tactician character has to be played by a tactician player, anymore than an linguist character has to be played by a linguist player. The tactician character can simply provide abstract mechanical advantages to their allies, like better initiative or freer movement. So, the thought processes and sensorium of an android (or ghost, or wizard, or alien) can simply have different bonuses and penalties to different things to get across their differences from humans (who in a lot of games are simply the "baseline" from which everything else differs).


Simplistic, yes, but probably not overly.

Isaac Asimov's characters (including his androids) always spoke in English (or whatever the story was translated into) and always used normal, understandable terms for things like cars and pens. He acknowledged that he could at least come up with unusual words for things when writing alien characters, but at the end of the day he felt like that would needlessly complicate things for the reader. RPG players are in a similar situation: is it worth going to a lot of work to make a different character seem different, or is it enough to just highlight a couple of key differences?


One could go that way, though not everyone would find that terribly realistic. The Laws of Robotics are a handy plot point, which Asimov and others have had a lot of fun with, but they don't represent any form of logical "programming" or design, or even come across as all that plausible. In the real world, I doubt anyone would make robots that were impossible to make into soldiers, and I doubt it's possible to make an artificial intelligence that is inherently loyal.


They have whatever reason the game itself requires, including none that is ever revealed. Things in fictional situations don't need reasons, particularly if they're cool. They really only matter if they're a key plot point.

In Eberron, the warforged are rather like androids, in that they are artificial and intelligent. In that setting, they were made for a purpose and it's now illegal to make any more. But that means that finding out more are being made could be an interesting adventure concept. If the adventure is about something else, then what's going on with warforged as a race of people might not matter all that much, though the warforged who appears as PCs or NPCs could still be interesting and strange.

My theory these days is that consciousness is unstable, that anything with even a little ability to think must eventually sleep because each instantiation of their consciousness breaks down after a day or so and must be recreated. So, if I made a game with androids, that would be a factor: if they didn't restart and rebuild every day or so, they'd get worse and crazier until they failed permanently. This goes against one of the common tropes of androids, that they never have to sleep. I'd probably also have my androids eat to supply themselves with energy (and be made out of lightweight, inherently fragile materials, so they didn't have to constantly eat sugar) and drink water to help process the food and keep their materials soft and lubricated.

My androids would be a lot like humans. Possibly smarter, but really that's just asking for trouble anyway.

It seems to me that only human arrogance could conclude that sentience is solely the purview of human perception, lol.  As happens in all cultures any true AI would adapt to the cultural expectations around them.  A robot or android that "grew up" in a specific culture would exhibit those cultural expectations.  I would say that any alien culture would have the same occurrence, unless of course the robot was created with physical limitation that stopped them from fully participating in such activities.

As for if they would be "truly" sentient or not?  Well I guess that is up to the GM of any specific game and the player of the character.  My absolute favorite character in the Star Wars series is L3-37 in Solo.  Her desire to be free and for her fellow droids to experience self awareness and self actualization was one of the most creative ideas to happen in a movie IMO and I believe that creation of AIs would eventually lead to that form of sentience in the real world.

except, we ARE the only example of sentience and sapience in existence.  We can consider the possibility of it existing out there in some other form, but for the entirety of human existence, we are the only example, and the benchmark on which we measure it.  Even the example of the octopus only exists because we, as humans, watch the animal, and see behaviors similar to our own that we categorize as being closer to human sentience.

We can't really escape that.

Exactly, human arrogance, "We can't see it so therefore it must not exist because we are the pinnacle of existence in the the universe, known and unknown."  The octopus exists whether or not we put it into a category of sentience or not, lol.

no, see, you're twisting it.  WE are the only thing we can use as reference.  We have no proof, at all, of any other sapient life in existence.  So we can imagine what it might be like, but we are still the baseline for that term, and for how we'd imagine other life.

And yes, the octopus exists.  But to categorize anything as sentient, a human term, we compare it to us.  humans.  We're the basis of every comparison when it comes to sentience, because we are the only thing we have experienced.  it isn't arrogance.  it's defining a term.

As previously mentioned, sentience has been defined in terms that make it impossible to measure. Sapient means human, that criteria can't apply to androids (or any non-human regardless of sentience). Humans are ill-equipped to assess these things. For example, consider the following:

a standard sentience test includes an examiner and a subject, both are sentient but must perform the test with no physical objects outside their natural body
• the examiner must be able to prove to the subject that the examiner is sentient, and therefore the examiner's sentience must be obvious
• the examiner must be able to perceive the sentience of the subject and therefore the subject's sentience must be obvious

Now, if the examiner and subject were human, what about them made their sentience obvious?
Which, in the context of the OP, makes the impact of the question debatable. I dare say that any android character seen in an RPG would be sapient due to the limitations of the player.

Interesting how you presume that animals are neither sapient nor sentient, despite evidence to the contrary, simply because they cannot explain it to you in a language you comprehend.

Actually, you seem to be twisting it.  You are equating "no proof" to "does not exist."  The statement you should have made is "we are the only example of sentience and sapience in existence that we know of."  The universe is a big place, and making assumptions about the whole of it after only exploring the merest fraction of a fraction is the definition of human arrogance.  Just because the words "sapience" and "sentience" are human terms does not mean that the concept would be foreign to any other intelligent life that might exist.  It might just be called something different (and might not include humans in its definition!)

As for androids and robots, when a creator attempts to create life, they generally mold their creation after themselves.  So to determine how an artificial life form would behave starts with looking at who made it.  Self-will can obviously change that, but the core of the creator is still there.

I didn't say they AREN'T sentient to any degree, I am saying WE are the example of sapience and sentience that we use to measure everything else.  By definition, it is measured in comparison to us, because we are the only example of it we can define.  So I guess it was incorrectly stated for me to say it that way.  I simply meant human mentality is our only true frame of reference to measure anything in relation to thought.

And, no, I didn't twist it.  We ARE the only example that exists right now.  You can't give me another example.  That doesn't mean that can't change, but right now we are.

in the same way that we define pain by our own human experience.  We honestly can't necessarily say that every human being feels pain the same.  We assume they do, but only because we know what pain feels like to us.  In the same way, we measure everything else in existence by our own experience.  We really can't do it any other way.

So in terms of artificial intelligence, for the game originally discussed, the Gm has to decide how he is going to define a sentient/sapient/intelligent computer.  If it is a true artificial life form, then by our own definitions it is going to work in many ways similar to a human.

If he makes it too alien, then it leaves our frame of reference and no longer really fits into the idea of an artificial life.

it's the same with grand alien entities.  Look at the alien probe in the original star trek movie.  Everything about it that we could use to identify it, what it was thinking, what it was doing, were things we could relate to as humans.  Without than relation, we can't even really consider it an intelligence by our understanding.

I think it's interesting that the people that have argued with me take a tone as if I am somehow being arrogant or presumptive as a human because I view it this way. Other than being a pretty big judgement call on their part, it is also strange, because I didn't say anything about them being lesser as an octopus or an animal or anything else.  I simply said they are not human.

And it isn't arrogant to say we have no examples of alien life yet.  I didn't say we never will.  I said we don't have one now.  That's my point.  And it won't be called something different.  Even if we encounter an alien lifeform, we would only recognize intelligence by the things it does that are in common with our own intelligence.  We have to define it by ourselves.

if the Gm wants to play a game with robots as players that are so alien they function in no way like a human being, well, go for it.  I just have no idea how you could possibly do it.  to me, you have to find what would be similar between the robot and a human, and build from there.
This message was last edited by the user at 21:16, Thu 19 Sept 2019.

Okay, so help me out here:

Let's say we want to create a setting where there are both humans and androids - and androids are free-willed, with the legal right to self-determinate. Could such a setting work? And do you have any ideas how, in such a setting, new androids would come into existence?

Well, last question first.  They would have to be built.  Either by humans, or by other robots.  Both of those answers would shape your world.  Perhaps humans who create androids are illegal, or only work for specific corporations, or governments.  But then that ties the androids to those entities.

Perhaps robots require very rare materials, and so are very limited in number.  Perhaps the androids themselves are the only ones that have the capacity to make others at this point, and so it takes a very similar role to having a child.

I think a world like that is possible, but it would require a lot of back thought for you to design.  You'd need to consider how far down the rabbit whole of politics, of social constructs, and of metaphysical discussions you'd want to go down in a game.