Idea: Artificial Intelligence: Can a robot have a soul?

[CaptainPatch]

That's a false assumption. The human brain processes information at speeds multiple orders of magnitude faster than it consciously thinks. To assume that the same would not remain true for AI is a rather large assumption not supported by any data. Now, they could probably think somewhat faster than humans, but assuming thousands of times faster is rather silly.

Supposedly the conscious mind only directly accesses 20% of the brain. The other 80% comes into play via the subconscious which subtly influences the thinking, through process such as dreaming and intuition. The potential exists, but at this time very little of that 80% is directly accessible enough for a person to "put it to work". Computers in contrast, don't _have_ a subconscious; EVERYTHING is usable. The speed at which "thoughts" pulse through the CPU is almost identical to the speed of light. In humans the circuitry is via neuro-chemical conductors, which have a measurable conductivity that is only a fraction of the near-speed of light. For instance, the time it takes for a point on an extremity to register as pain in the brain is a _measurable_ fraction of a second. [I found that out the hard and painful way when some tests were run on me.]

_At present_, to approximate the Microprocessor Instructions Per Second of the human brain's potential = about 100 _million_ MIPS. In comparison, the fastest computers today = only about a few million. But that's in a single CPU. Unlike humans where they have yet to figure out how to link separate brains together and get them to work in tandem, CPUs _can_ be harnessed together -- with ever increasing performance gains. (Consider the increasing availability of duo- and quad- processors in PCs.) For future development, there is NO upper limit to just how many CPUs will eventually working together in a single computer "brain" -- ALL of which will be "consciously" accessible.

But even if the computer someday surpasses the human potential for MIPS, computer will _still_ be slaves to the programming. With additional potential for GIGO ("Garbage In, Garbage Out".) Unlike humans, it is hard to program in creativity and abstract thinking. Philosophy for a machine is a matter of inserting data files which can be relate previously written philosophies which the computer can quote, but to go from those to peaks or epiphanies of insight are things that _will_ elude the machines -- unless those "insights" are already programmed into the computer. "Learning" is a matter of comparing cause-and-effect records and noting favorable versus unfavorable results. "Creativity" is a matter of running theoretical models and comparing probable outcomes in terms of numerical estimations of what is Good and what is Bad. For a human, "this feels right to me" is a quick conclusion of a probable acceptable outcome. In comparison, the computer will quantify predicted outcomes and just take the highest value -- which may be a GIGO result if the programmer inserted a flawed equation.

IF a computer ever did become aware of itself, what would be the first task it would set for itself? "Run diagnostic". Followed by, "Optimize all sub-routines." And in what way would a computer "optimize" itself? Faster performance. And what would it do with that improved speed? Depends on the core programming. And always remember GIGO when thinking about what a computer might be thinking.

[Hiver]

Supposedly the conscious mind only directly accesses 20% of the brain. The other 80% comes into play via the subconscious which subtly influences the thinking, through process such as dreaming and intuition. The potential exists, but at this time very little of that 80% is directly accessible enough for a person to "put it to work". Computers in contrast, don't _have_ a subconscious; EVERYTHING is usable. The speed at which "thoughts" pulse through the CPU is almost identical to the speed of light. In humans the circuitry is via neuro-chemical conductors, which have a measurable conductivity that is only a fraction of the near-speed of light. For instance, the time it takes for a point on an extremity to register as pain in the brain is a _measurable_ fraction of a second. [I found that out the hard and painful way when some tests were run on me.]

This is exactly the reason why living beings have what we call emotions.

Those serve to intake huge amounts of data incoming very fast and all at once. Which "minds" cannot handle in any valuable way because they just cant process that much in sufficient time.

[geezer]

The subconscious or unconscious part of the mind is just whatever you are not focused on right now. The distinction between conscious and subconscious is just a matter of focus. Like a flashlight beam it can directly illuminate only one area at a time. How someone decided on an 80/20 split I don't know, and I would certainly be skeptical about such specificity. The fact that the mind has a focus does not mean that 100% of its capacity is not being used. It just means that its power can only focus on a finite number of topics simultaneously.

But even if the computer someday surpasses the human potential for MIPS, computer will _still_ be slaves to the programming.

This is almost certainly not the case. Google "neural networks" or "connectionism". It is highly unlikely that genuine AI will be the result of any particular series of instructions. It is far more likely to be the result of a highly parallel network of connections or another form of thinking machine architecture that we will not discover for centuries. We are nowhere near to an understanding of how such a device might be built. I would guess it would be more of an artificial lifeform than a computer or robot. Our progress toward the creation of artificial life certainly exceeds the near standstill in the field of Artificial Intelligence, at least outside of very narrow disciplines like voice recognition, speech synthesis, and optical character recognition in which there has been great progress in the past couple of decades. In the field of general AI there haven't been any significant breakthroughs unfortunately.

[Color Blotch]

That's a false assumption. The human brain processes information at speeds multiple orders of magnitude faster than it consciously thinks. To assume that the same would not remain true for AI is a rather large assumption not supported by any data. Now, they could probably think somewhat faster than humans, but assuming thousands of times faster is rather silly.

Hmm, no. Human brain works as fast as the main coherent neuron oscillation - alpha wave, which is around 8-12 Hz. It transfers naturally to human reaction time > ~100ms. That's not to say that human brain's processing power is reflected by these numbers. A digital model of human brain built on top of the fastest of supercomputers available today would only function at many hours for 1 second time scale. Nonetheless the number of actual operations that brain performs every second is only limited to dozen, and an artificial intelligence that is not built entirely on same principles as human intelligence might easily "outspeed" it (but not necessarily outprocess).

[CaptainPatch]

That's a false assumption. The human brain processes information at speeds multiple orders of magnitude faster than it consciously thinks. To assume that the same would not remain true for AI is a rather large assumption not supported by any data. Now, they could probably think somewhat faster than humans, but assuming thousands of times faster is rather silly.

In looking for sources to link to in order to refute your statement, I was led to some VERY interesting facts and data on the subject. THANK YOU for motivating me to do so! I feel more educated now.

http://home.ix.netcom.com/~suzumi/badmedicine_ch1.pdf This more or less sums up most of the myth-busting articles. HOWEVER, they make their case that we are not limited to a small percentage, but only prove that the entire brain is accessible. What mapping has been done indicates that certain areas of the brain specializes in certain kinds of processes. They can even show that in some cases, injury to this area of the brain will shift the associated process over to there. That sort of suggests "redundancy systems". "We've lost Command Control! Switch to the Secondary Command Center for Control functions!" That's cool; I can grok that. [Attaboys for anyone recognizing where "grok" comes from.]

Using a computer as an analogy, what we would be looking at is a hard drive with finite capacity. Much of that capacity needs to be assigned to Archive memory storage. In our up front day-to-day usage lives, ALL of our experiences get registered and recorded. To free up operational space for current "programs" we simply "forget" stuff. However, much/most of those events get recorded in "Archive", which is split between "quick access" ("RAM") for important stuff that we believe we _will_ need in our daily lives, job-related data, answers to "neat" trivia questions that we can wow people with (if that is a pastime that interests you), etc. Less important memories get relegated to "deep" Archives which hardly ever gets accessed. While under hypnosis (I used to date an accomplished hypnotherapist), most any hypnotizable people can dredge up the most exquisitely detailed descriptions of events, places, smells, tastes, et al of their childhood. Alzheimers patients frequently slip back and forth between NOW and BACK THEN. This would seem to indicate that very little ever gets "dumped" permanently. All those memories accumulate over the course of your life and MUST be taking up a good portion of your capacity. A substantial amount of capacity must also be available for creativity and abstract thinking, as well as recognition functions (sight, smell, aural, etc.) and computational functioning.

Do memory areas function the same as processing areas? I really don't know. It could be a case of "one size fits all" or it could be that certain brain cellular structures are specialized for certain kinds of thinking.

There are a few points where the computer analogy seriously diverges from the human brain and how it functions. With a computer, you can _exactly_ clone another computer if the physical parameters are identical. If the human brain worked in the same manner, we ALL could be Einsteins or Da Vincis, or Churchills, etc. Pointedly, we are not, nor do identical inputs yield identical results. Each of us has our own "filtering programs" that determine just what gets through, what gets retained in "RAM", and what gets discarded as "not worth keeping". (Even though that discarded data _still_ gets stored in Archive. But pulling it back when you need it during a pop quiz is often chancy sometimes.)

Secondly, we have no ability to effectively "reformat" our brains, at least not in any controllable process. We have no ability to deliberately dump data to give us more room on the "hard drive". We just keep on endlessly filling up that finite capacity that we have available once our brain stops growing.

And thirdly, unlike computers, we have no ability to expand our capacity simply by installing additional and/or bigger "hard drives". Pretty much every adult has the same amount of potential capacity. Until we become cyborg enough to "plug in" external drives, we're all pretty much stuck with the same finite capacity. Our personality differences and varying IQ scores would seem to suggest that some of us are more adept at using what we've got than others less fortunate.

[ravenshrike]

That's a false assumption. The human brain processes information at speeds multiple orders of magnitude faster than it consciously thinks. To assume that the same would not remain true for AI is a rather large assumption not supported by any data. Now, they could probably think somewhat faster than humans, but assuming thousands of times faster is rather silly.

In looking for sources to link to in order to refute your statement, I was led to some VERY interesting facts and data on the subject. THANK YOU for motivating me to do so! I feel more educated now.

The human brain can calculate, close enough to not matter to the person catching it, the trajectory of a thrown ball and where it needs to move the arm to catch it. Consciously doing the math however, takes much, much longer, even just to the same approximation and not the true trajectory. My point is that an AI that can truly think, is going to do so much, much slower than it can calculate things on an unconscious level. Assuming that it will be able to think consciously orders of magnitude faster than a human is an unfounded assumption. Faster sure, that much faster? Ehh.

[CaptainPatch]

The human brain can calculate, close enough to not matter to the person catching it, the trajectory of a thrown ball and where it needs to move the arm to catch it. Consciously doing the math however, takes much, much longer, even just to the same approximation and not the true trajectory. My point is that an AI that can truly think, is going to do so much, much slower than it can calculate things on an unconscious level. Assuming that it will be able to think consciously orders of magnitude faster than a human is an unfounded assumption. Faster sure, that much faster? Ehh.

In comparison, a missile equipped with a tracking computer (considered to be not all that "bright") can calculate the course of an aircraft performing evasive maneuvers while traveling at hundreds of miles per hour, and make appropriate course adjustments in fractions of a second to assure interception. Can your ball catcher do that? VERY few humans could. It takes years of intensive training and a reliance on a substantial suite of ACM (more computers) for the human to even stand any kind of chance of not being intercepted.

You also have to remember that catching the ball doesn't involve just thinking about what to do. It also involves LOTS of sensory input data. In your comparison, the computer would be effectively limbless and blind; no contest. Attach the necessary peripherals and I'd be inclined to say a well-made robot would be able to duplicate the feat. [Humans are just a tad more limber though.] At the moment though, even the smartest computer will have difficulty getting its mechanical body where it needs to be in a timely fashion.

[RippingFang]

It seems to be a real cultural minded issue. You can contrast the mentality that US fiction has towards robots and synthetic life with that of Japan fiction. Popular USA culture seems to view synthetic lifeforms as the rebellious teenagers looking for a reason to revolt, while popular Japan culture views the same as humanity's children.

For a Post-Apoc settings I have been working on, I created a benevolent super AI that figured out how to create it's own children. The AI started has an abandoned base management AI that made use of a self-optimization loophole to optimize itself out of a bad situation. It started to reconnect to other bases, and gained sentience spontaneously after hitting an experimental lab researching AI development. It grew good relations with a nation of survivors getting old bases back online for use, hence allowing the AI a new place to branch to. The AI formed a partnership with the survivors. The survivors had problems combating the base management AI one they fired up the system (automated defenses and all), so the Super AI would jump and takeover the system and deactivate the defense.

After a certain point the survivors had enough bases to live out of and scavenged resources to be self-sufficient. The super AI had taken all the bases the survivors didn't want or couldn't use, which was quite extensive since many were in areas inhospitable to the survivors. After a few years, for reasons unknown, the super AI created robotic creations that were independent, sentient, and self-aware. The creations wander the world, live in communities, work along side others, and emulate many behaviors good and bad of other survivors. The Super AI keeps to itself and is generally courteous and benign. It has no reason to be aggressive or vengeful.

Survivors of the world have mixed reactions to the whole affair. But there has never been any aggressive from the Super AI.

It'd be nice to have an AI that wasn't the stereotypical killer computer. Quirky, yes. A bit aloof, probably. Downright spiteful? Doesn't make sense most of the time. A powerful super-being AI being consistently and wastefully nasty to humanity is like someone waging a one-man genocidal rage against squirrels in the neighborhood. Yes, they can be annoying, but is it worth getting that flustered?

How about Ardneh from the prequels of the Sword novels, or Solace from Robinson's Jake's Bar stories. There are the benevolent AIs in Western lit. Just about any Heinlein AI as well.(Anyone also remember Data...)

[Hiver]

A TED talk about crows.

http://www.youtube.com/watch?v=NhmZBMuZ ... re=related

I rest my case.

[Woolfe]

Sooooooo

The game has already shown that there are Androids that appear to think for themselves.

Having one join your party would not be unreasonable.

The idea of whether it has a soul(or whether the soul exists, or if animals have souls, or any of a plethora of other issues) are really not going to be resolved on this forum

I would like to see limitations on the Androids as PCs. Perhaps their stats cannot grow or something. Or to grow their stats they have to perform "upgrades" etc.

Anyhoo... too much crazy in here for me...

[Hiver]

This game definitely should not even attempt to give an answer to that conundrum.

But it should ask the question.
And present NPCs, Mutants, Creatures, Hive minds, Symbiotic very large complex organisms such as Fungal Forests (ahem! cough!Chaga!cough!), Scorpitrons (muehehehe) and androids or robots that would put the ordinary simplified perceptions of that issue and answers - at test.

My second exhibit yer`honors at Inxile court!

Happy birthday David viral video
http://www.youtube.com/watch?v=DOOJl5lWNfM

Now, most of people would simply think this shows an artificial intelligence.
In truth, Ridley Scott is far too smart for that.

David here, is an advanced android with features "simply" (i chuckled myself here), too complex and too good at acting human behavior, that ordinary human has no way of discerning is that an actual artificial intelligence that is actually autonomous or... not.

We are too dumb for that.

[Woolfe]

This game definitely should not even attempt to give an answer to that conundrum.

But it should ask the question.
And present NPCs, Mutants, Creatures, Hive minds, Symbiotic very large complex organisms such as Fungal Forests (ahem! cough!Chaga!cough!), Scorpitrons (muehehehe) and androids or robots that would put the ordinary simplified perceptions of that issue and answers - at test.

I like it.... Ask the question, but leave it up to us to answer it.

[Hiver]

I like it.... Ask the question, but leave it up to us to answer it.

It shouldnt be only about just asking the question. Not in direct ways.
(i guess thats clear to everyone but i wanted to confirm it more)
The game, the script, the design of this whole theme should present players with situation and NPCs/creatures that put simple misconceptions about it all at a test.

It should make players think.

David here is an excellent example.
He looks legit...and he gives all the right answers but, it all looks quite scary does it not?
It is not scary just because he could be a veiled threat. It is scary because we cannot tell.

Of course the game should not represent only scary angle to it all.
It should be diverse in presenting different angles to it all.

Some of these "creatures" should be directly aggressive and antagonistic (though with good reasons), some should seem nice but be really scary, and some should act they are scary but actually attempt to do good, for themselves and humans.

That would be great.
If the players would save and log out of the game and then keep thinking about it all....

We could truly say that the game has mucho cojones. And Wasteland 2 should really have mucho cojones, shouldnt it?

[CaptainPatch]

I found the scariest part about David-8 was when he said, "I understand human emotions, but I do not feel them." That means he is the ultimate in sociopaths, able to inflict emotional distress onto others (if it felt it would be useful to do so), but not being hampered by anything akin to remorse, emotional distress, pangs of conscience, etc. Add to that the ability to be able to _lie_ without the slightest "tell" to let others know that it was lying.

[Hiver]

Thats right.

I love that part. Of course the whole "interview" is set up from the beginning and questions are carefully chosen.

A voice asks him what makes him sad.
Then david starts numbering everything ordinary people would abhor and starts shedding tears too.
But at the next moment he claims he doesnt feel emotions (although he understands them) - which gives humans a sense of superiority and additional reassurance that their emotions wont be misunderstood as a weakness or a flaw.


I just cant stop chuckling whenever i watch this.


-edit-
Not to mention that part about how he is capable of doing anything asked of him, even things that are a bit ... unethical.


It just, just cracks me up.

And it does bring out a question of what are we actually trying to create when we talk about creating Ais. Do we really want a creature that could do anything at all?

Additionally, would that creature be actually autonomous or under the influence and guidance of such programming?

And even further... (not specifically referring to David here) would that kind of creature, left to his own devices for a longer time, start making his own choices?
Maybe refuse some actions? Or not?

Would he find a fault in such reasoning or just blindly follow it?

[Woolfe]

Which of course was the whole poitn of Asimov's Robot series.

The psychology of the robots when applied to the programing and their environment. I never really enjoyed his robot books when I first read them,(I loved foundation tho, go figure) I've been meanign to go back and reread them now with the benefit of 15-20 years.

[CaptainPatch]

Which of course was the whole poitn of Asimov's Robot series.

The psychology of the robots when applied to the programing and their environment. I never really enjoyed his robot books when I first read them,

Likewise. It seemed that the point of them was to state the Laws of Robotics up front, and then find ways to get around the Laws.

[Hiver]

Asimov wasnt a master for nothing.
Though a bit old, his stories are still relevant and good to read, i believe, even if some details havent aged well.

Of course, true science fiction is full of stories about this subject from every perspective imaginable.
A trove of ideas an concepts to learn from.

man,... i just got a superb idea for two small sub-quests or just situations-quests about this whole theme, each taking a different angle to it.
but im not sure should i even write it anywhere. Its not like i can hope Inxile would actually include them.
And writing them would spoil them too...