PDA

View Full Version : Artificial Intelligence in Star Trek



SolusLupus
2005-Oct-26, 05:53 PM
At the risk of putting up too many topics (I do seem to be putting up a lot...), I must ask a question:

Why is Data in Star Trek so special? It seems to me like he's this big character, and the "top of the line" in Artificial Intelligence.

Yet, not only can the Holo Deck simulate people, it simulates people and their personalities, memories, mindset, and their ability to make decisions... even to the point where you can have a hologram of Stephen Hawking, Einstein, AND Newton around a table, playing cards (though that was in Voyager, not Next Generation, but I don't know how far apart the two are).

Yet Data doesn't get the grasp of emotions for a long while, until he gets the emotion chip. Even then, he has very little control over himself while under its effects. They had to reroute it so he could switch it on and off with a twitch.

I'm kinda relying on memories. I haven't watched Star Trek for a long time. Honestly, this question just came to my mind in the shower. >.>

Ken G
2005-Oct-26, 07:36 PM
My guess is, the Star Trek folks would say that the holodeck knows how to*simulate* emotion, like an actor. Data could presumably do that too. The problem comes when he is actually *experiencing* emotion, which is different. Still, your point is well taken about the power of the holodeck-- imagine, for example, what would happen if the holodeck simulated Data himself. Would it not need at least as great an intelligence as Data's to do that? Maybe Data is just more compact, like a handheld computer. Or maybe the holodeck never tries to simulate Data-- any trekkies want to field that one?

publiusr
2005-Oct-26, 07:43 PM
There was the woman who was downloaded into an android body that could age IIRC. She didn't even know she was an android. (TNG era.) That was top -of the line.

Van Rijn
2005-Oct-26, 08:09 PM
In several TNG episodes, they had a theme that hinted the AI was capable of self-awareness and sentience. Nor were Sung androids or Star Fleet computers the only ones to exhibit that. And by the time of Voyager, it was clear that hologram simulations could be just as humanlike as properly designed Sung androids.

But when it came to AIs, instead of robots/simulated people, Star Trek always pulled back after presenting the idea. The problem is the possible implications of AIs that can be as smart as or smarter than people. It would take the story in directions they didn't want to go.

There are a lot of technologies in Star Trek that would lead to a very different picture than the "21st century with gimmicks" universe they present. For instance, the transporter, with a few adjustments, could rejuvenate people. They have nanotech, advanced genetic engineering, advanced AIs, and virtually unlimited resources ...a better picture would be interesting, but it would be too hard to present on TV.

Gillianren
2005-Oct-26, 08:44 PM
Yet, not only can the Holo Deck simulate people, it simulates people and their personalities, memories, mindset, and their ability to make decisions... even to the point where you can have a hologram of Stephen Hawking, Einstein, AND Newton around a table, playing cards (though that was in Voyager, not Next Generation, but I don't know how far apart the two are).

That was in Next Gen, because one of the ones playing was Data. And, yes, it was really Stephen Hawking.

SolusLupus
2005-Oct-27, 12:06 AM
(Since the HTML quote seems broke)
Quote, Gillianren:
That was in Next Gen, because one of the ones playing was Data. And, yes, it was really Stephen Hawking.

I figured as much that it would really be Stephen Hawking. And now I remember that it was Data playing a cardgame with them. Voyager was Leonardo Da Vinci. My bad.

Quote, Van Rijn :
There are a lot of technologies in Star Trek that would lead to a very different picture than the "21st century with gimmicks" universe they present. For instance, the transporter, with a few adjustments, could rejuvenate people. They have nanotech, advanced genetic engineering, advanced AIs, and virtually unlimited resources ...a better picture would be interesting, but it would be too hard to present on TV.

Hmm, interesting. I didn't think about the rejuvination bit. Immortality!

Van Rijn
2005-Oct-27, 01:07 AM
(Hmm, interesting. I didn't think about the rejuvination bit. Immortality!

They actually did it in one episode, and promptly "forgot" about it:

http://www.startrek.com/startrek/view/series/TNG/episode/68372.html

One of my general complaints about Star Trek is that at their level of technology, aging simply should not be an issue, whether through transporter tech or nanotech or advanced general biomedical technology. But there is an often repeated Star Trek theme that trying to stop aging is bad (an idea I find silly at best).

On the transporter: They also have shown that it is possible to duplicate someone, but again, limited it for story purposes. It would certainly be a good way to improve the safety record of away teams:

(Buzzing sound of rematerialization)
Riker, Data, Worf and red shirts appear on the transporter.

Riker: Chief, what are we doing back here? Why didn't you beam us down?

Chief: I did beam you down, sir! It, uh, didn't go well ...

Riker: So, we ...

Chief: Yes, sir. You tried to stop an anti-matter reactor breach on that ship. If it is any consolation, one more second and you would have done it. Really beautiful explosion, though. Vaporized everything, including the away team. Good thing you were recorded in the pattern buffer!

-----------------

Anyway, regardless of "magic" technology, at the general technology level they show on Star Trek they should be able to do far, far more.

eburacum45
2005-Oct-27, 02:36 AM
21st Century with gimmicks- very good point.

one thing that puzzles me about Star trek AI is that the most advanced example is supposed to be small enough to fit inside Data's head.
Surely the first attempts at building human equivalent AI would be much larger, and only later miniaturised to be portable.

There may well come a day when computers are built with an approximation to human intelligence; the next day, (or perhaps after the weekend, when the celebrations were out of the way) work would begin on an upgrade.

Someone once asked Vernor Vinge if there would ever be a computer as smart as a human.
A: Well, yes... very briefly.

If the possibilities behind Trek science were taken to their logical conclusions, every aspect of the show would be different.

SolusLupus
2005-Oct-27, 05:10 AM
Y'know, the more I hear, the more I'm inspired to use my stories to take Science Fiction to a new level. To explore new frontiers. To go where no science fiction author has gone before.

Enzp
2005-Oct-27, 07:37 AM
But still chase after women in every chapter.

It was, after all, just a TV show. It was easy to see the differences in writers. In some episodes Spock was oblivious to human foibles, and in another episode he was completely aware of them and able to use them to his purposes.

And the characters never learn or age. How many times did we need to go through the Bones/Spock thing where McCoy would misinterpret the logic-only actions of Spock? Was McCoy really that thick or stupid that it never sank in?

Data always wanted to be human, but he also was always clinical and free of emotion. He wished he could tell a joke. Well, desires like that are purely an emotional thing. That inconsistency in premise was never explored.

I often wondered why they forgot they could make a new copy of anyone killed from the transporter records.

Data was SPock in sheep's clothing. Plus he was compact and portable. Holodeck could not haul your butt out of a flaming something or other on the planet below.

Unlike the original series computers, Data could not be disabled by asking him to divide by zero or contemplate a self-contradiction. Kirk used these techniques too many times. Ever was too many. Even the $1 calculator I have next to me knows divide by zero is an error. Points then for Data.

I can't imagine that anyone on any of those crews would ever trust each other. How many times do you need to have your crewmembers taken over by alien or other outside intelligence before you could never be sure if it were not the case at any moment? ANy trifling inconsistency in behavior arises, and... is he in a bad mood, is he possessed? I don't understand why he gave that command, is that really #1 or is that the Romulans controlling him?

And the whole Q thing. DeLancie plays the arrogant jerk to perfection, but I find the Q character tiring at best. But after an encounter with him, why would you not second guess any weird thing that befalls you? Is this a real tragedy or a Q tragedy?

Sorry, stream of consciousness drifted away from Data...

novaderrik
2005-Oct-27, 10:15 AM
i think i remember on TNG that they said they don't use the transporter to bring back the dead or stop aging or cure baldness because of the moral ramifications.
they played up the "evil twin" aspect of the transporter when they found that Riker double on that planet after being stuck there for-what?- 7 years or whatever. they did the "evil twin' thing backwards on that one, with the evil one not having a beard. i always thought that was kind of funny. well, he had a beard on that one episode of DS9 when he was impersonating Riker...

Ken G
2005-Oct-27, 12:02 PM
And what about "security"? Did these guys ever do anything right, except when they lead the criminal away at the end of the show? I always think at that point, "hey, you're not just going to let security lead him away are you? You did that at the beginning of the show, and look what happened!"

And, how come they never beefed up security in Engineering? In the original Star Trek, somebody took over Engineering about every two weeks. You'd expect them to say, "Captain, the helm is not responding... darn, somebody's taken over Engineering again, why do I even bother trying to steer this thing? I can't work under these conditions!"

But seriously, of course it's just a show, and a darn entertaining one at that.

SolusLupus
2005-Oct-27, 02:01 PM
Hmm, lots of logic foibles. It's fun overall, in my opinion, to tear apart something like Star Trek to really think about things... (I don't have much to contribute right now, I'm tired as hell and in a class I shouldn't be posting from :P)

Alasdhair
2005-Oct-27, 03:58 PM
One of the more impressive feats of AI was when the Holodeck created a Professor Moriarty that could out-think Data. In a subsequent episode, Moriarty, his fiancée and an entire galaxy for them to live in were recreated in a palmtop device...

One of my favourite trekisms from TNG was the way they expressed Structural Integrity as a percentage: I was waiting for one of them to say;

"Structural Integrity's down 5% - the left engine just fell off..."

SolusLupus
2005-Oct-27, 04:47 PM
"Structural Integrity's down 5% - the left engine just fell off..."

*bursts out laughing*

That's just so... hah!

Weird Dave
2005-Oct-27, 05:03 PM
I remember Spock confounding an evil computer by asking it to compute all the digits of pi. They obviously hadn't heard of finite precision floating point numbers.

I wanted the computer to come straight back and say, "Pi is precisely equal to 3.1415... [Voice speeds up to unintelligible whine] ... 24. Next problem."
Kirk: "Any other bright ideas, Spock?"

SolusLupus
2005-Oct-27, 05:28 PM
I still like how in the original Star Trek, brain surgery was a Big Thing, and almost impossible to perform. Oh, yeah, they can tear you apart molecule by molecule, piece by piece, and put you back together again in another area, but they can't work on your brain!

Of course, this was in the original Star Trek, which had the perceptions of the time period... brain surgery was pretty daunting back then, and it was predicted that it would remain so for a while... which isn't a bad prediction, I admit.

Faultline
2005-Oct-27, 07:34 PM
The makers of TNG were so proud of their attention to detail and continuity.

Maybe they had storyline continuity, but they had little in the way of technical continuity. When they needed a plot, they often broke their own technical laws or simply ignored them.

ASEI
2005-Oct-27, 07:46 PM
Speaking of which, why didn't they use the transporter as an ultimate weapons system? They could presumably beam a portion of their antimatter into an enemy ship - who cares if it's scrambled by their shields? It still has to conserve baryon number. You'd have random nuclear explosions all over the place.

Or they could use replecators and transporters, coupled with a suitable power source, such as a near orbiting planet with good solar energy, to create a self replecating factory which would take the entire thing apart and produce billions of something. Want an instant starfleet? Replecate/transport it out of a planet.

And what about all these helpless planets anyway, when confronted by the big bad starship enterprise? I would think that a planet would have far greater resources at it's disposal, both energy, material, and shielding wise to blow away any flimsy starship in orbit. Even with radically advanced technology versus, say, the 20th century, what would stop us from putting up all our ballistic nukes and causing at least a problematic radiation incident for an attacking starship?

NEOWatcher
2005-Oct-27, 07:57 PM
They could presumably beam a portion of their antimatter into an enemy ship -
Voyager did it to the Borg with a torpedo. (don't forget that you can't transport through a defense field)


Or they could use replecators and transporters, coupled with a suitable power source, such as a near orbiting planet with good solar energy, to create a self replecating factory which would take the entire thing apart and produce billions of something. Want an instant starfleet? Replecate/transport it out of a planet.

Self replicated mines in DS9 to protect the wormholes. It took months for the Cardasians to defeat it.


And what about all these helpless planets anyway, when confronted by the big bad starship enterprise? I would think that a planet would have far greater resources at it's disposal, both energy, material, and shielding wise to blow away any flimsy starship in orbit.
Not if you have technology designed to work without an atmosphere.

Cl1mh4224rd
2005-Oct-27, 07:57 PM
Maybe they had storyline continuity, but they had little in the way of technical continuity. When they needed a plot, they often broke their own technical laws or simply ignored them.
Heh, yeah. There were a lot of "crazy" things they had to do to fix some (most?) of the problems they encountered, or interesting side effects, that could be put to good use, but... just disappeared.

gopher65
2005-Oct-28, 03:12 AM
On star trek they use handwavium all the time, and even more of that plentiful resource to explain away things like relicators and transporters. Relicators have a limited precision, and can't recreate a 'living' thing, although they can make a good approximation of a dead thing (like an apple or a lambchop or catnip). Transporters on the other hand don't actually 'create' things. They dematerialize an object, move it to another location, and rematerialize it (they call it a 'pattern' while it is in storage. It is stored in a 'buffer'). Seems like a goofy way for a teleporter to work, but that's how they did it. There were a few circumstances where the transporter malfunctioned or encountered some alien thingamabobber that caused it to duplicate an item or a person, but it wasn't able to do that under normal conditions.

One of the biggest things I dislike about trek is the lack of in-story continuity. I don't care if they use real science or not, and even handwavium is fine by me, but for crying out loud why couldn't they remember estabished 'facts' from one episode to another? IMO that is the biggest arguement against multiple writers in a series:P. A series should have one writer, or at least one person who puts everything together and edits incoming scripts.

nomuse
2005-Oct-28, 04:37 AM
The transporter/replicator issue was one they had to keep tap-dancing around. I mean, the obvious and simple answer is that they are one and the same. Matter is created from energy (LOTS of energy...) according to a pattern. But of course the implications were far too dangerous to the series premise.

So instead "somehow" they can't replicate living matter, or latinum, or dilithium crystals, or whatever else they are short of that episode (despite that fully functional weapons HAVE been replicated on several episodes, and the dilithium crystals are described as being continually rebuilt within the drives). And they somehow can't recreate matter using the transporter, or "edit" or "store" the pattern, despite it being done in multiple episodes. Nor can the holodeck create "real" matter, despite Wesley getting wet in it, and plenty of people indulging in four-course Holo-dinners without later ill effects. And they can't use any of this technology to alter human genomes or phenomes, despite the fact that sickbay creates tissue from air all the time, as well as blithly altering anything from genetic structure to macro-structure.

nomuse
2005-Oct-28, 04:47 AM
There is a simple explanation to many of the seeming inconsistancies in Star Trek; that the great unsaid quality of intelligent beings is that they have a soul. That is why you can play with the pattern buffer all you like but any time you try to materialize more than one person you get problems (it is fairly arguable that Riker II lacked that intangible "something" -- and of course Kirk I and Kirk II were "incomplete halves of a single being" dispite having of course complete complement of neurons, normal mass and body chemistry, etc.) Also, any intelligence carried in non-physical form, such as The Doctor, can only be transferred, never copied from back-up. The few attempts made to copy an artificial intelligence (Data's daughter, M5), only showed that those creatures arrived in the world without a true soul -- and most of these copies came to bitter ends.

I am being facetious, of course. I doubt the writers actually worked this out explicitly. It does make a nifty analytical tool as to whether a certain juggling of information (transporter buffer, etc.) would likely be canon to the various permutations of Star Trek.

SeanF
2005-Oct-28, 07:11 PM
One of my favourite trekisms from TNG was the way they expressed Structural Integrity as a percentage: I was waiting for one of them to say;

"Structural Integrity's down 5% - the left engine just fell off..."
Now, in their defense - the ship had a Structural Integrity Field, which was basically a force-field that helped to maintain the ship's own physical integrity.

So when they were giving a percentage, they were referring to how that field itself was holding up.

Gullible Jones
2005-Oct-28, 08:35 PM
Of course, having a ship that needs power to hold itself together is stupid. Just imagine it... Main engine goes off, everybody dies!

(Of course, using matter and antimatter when you don't have to is also dumb. The Trek people can mess around with tachyons, and it's easy to figure out how to make a matter->energy converter with charged tachyones... Duh.)

peter eldergill
2005-Oct-28, 08:37 PM
How about consistencies in TNG (and others)

I refer to standard StarTrekPlot (TM) #1: a holodeck malfunction, hilarity ensues
#2: A rip in the space-time continuum
#3: Posession by an alien
#4: Data runs amok, then all is forgiven
#5: A Spatial Anomoly (TM)

Others I've forgotten?

Pete

novaderrik
2005-Oct-28, 09:10 PM
don't forget the "time travel reset button" happy ending.

i always thought there should have been at least a few episodes of TNG that should have ended with the "everyone laughing" freeze frame that was so popular in 70's and 80's sitcoms.

Enzp
2005-Oct-29, 06:28 AM
OOoohh, yes, with happy disco music, like an episode of Chips.

captain swoop
2005-Oct-31, 10:45 AM
I like the Red Dwarf generic 'Swirly thing' alert

Damburger
2005-Nov-01, 09:33 AM
Thing that bugged me about Data was a larger problem in Star Trek - their approach to organic technology. According to the writers, organic matter was better than inorganic for making starship hulls, but worse for making brains. This is total bull.

Its commonly assumed by those with little AI knowledge (or those with a very symbolic view of AI) that computing hardware is 'faster' and generally 'better' than human brain tissue, or will be very soon, and the only obstacle to intelligence is software.

If you actually look at hardware and wetware side by side, you'ld be surprised at how effective our brains are. Neurons are more energy efficient (and therefore run cooler, and can be packed closer together) and more interconnected than their silicon counterparts could possibly be. The computation they can perform is more complicated than electronic elements of the same size.

Given that scientists have already grown animal neurons to perform certain computational tasks and control robots, it doesn't make much sense to try and construct intelligent systems out of inorganic hardware.

WaxRubiks
2005-Nov-01, 09:57 AM
It always annoyed me that there was only one Data(not including his two brothers). Why would they only stop at one? Oh, their creator was a genius that's right so obviously he knew exactly what secret ingredient to put in when he was stirring his cauldron, hitech is more of an art form(like cooking) than science, that seems to be the message.
And then when they get his brother up and running he is eeevil.Because the Universe is made up from devils and angels, nothing in between.
They have to introduce characters like the Woopie Goldburge character to try to make the characters more interesting without being non-PC, but it doesn't work.

StarTrek is ok but I think I prefer Hitchhiker or Blake's7, their characters are allowed to breath.

eburacum45
2005-Nov-01, 11:59 AM
If you actually look at hardware and wetware side by side, you'ld be surprised at how effective our brains are. Neurons are more energy efficient (and therefore run cooler, and can be packed closer together) and more interconnected than their silicon counterparts could possibly be. The computation they can perform is more complicated than electronic elements of the same size.
That is certainly true of present day technology, but the theoretical density of artificial neurons is much higher than attainable today; one theoretical limit for the smallest volume a human-equivalent processor could occupy is 0.4 millileters (http://lists.extropy.org/pipermail/extropy-chat/2004-December/011764.html). So, discounting cooling systems, a humanoid robot could contain 3500 human minds in its brain cavity alone, or a single mind 3500 times as complex as a humans.
This megabrain would produce a vast amount of heat, so you would need very good cooling system; the cooling system itself would probably occupy much more room in the head of our hypothetical robot that the actual processor, so the practical limit of a human brain sized artificial processor is probably just a few hundred times human.
Or you could make the head bigger, or distribute the processor around the body. As electronic wiring carries messages much faster than nervous tissue, the processors could be in the fingertips and still have a faster response to stimuli than biological systems.

Damburger
2005-Nov-01, 12:13 PM
That is certainly true of present day technology, but the theoretical density of artificial neurons is much higher than attainable today; one theoretical limit for the smallest volume a human-equivalent processor could occupy is 0.4 millileters (http://lists.extropy.org/pipermail/extropy-chat/2004-December/011764.html). So, discounting cooling systems, a humanoid robot could contain 3500 human minds in its brain cavity alone, or a single mind 3500 times as complex as a humans.
This megabrain would produce a vast amount of heat, so you would need very good cooling system; the cooling system itself would probably occupy much more room in the head of our hypothetical robot that the actual processor, so the practical limit of a human brain sized artificial processor is probably just a few hundred times human.
Or you could make the head bigger, or distribute the processor around the body. As electronic wiring carries messages much faster than nervous tissue, the processors could be in the fingertips and still have a faster response to stimuli than biological systems.

Sadly I can't read your link as I'm in a school with a gestapo-like content filter.

I will address your assumption that heating wouldn't be a major problem though - compare for a second the size of the CPU in your computer (not its casing, the actual silicon) to the size of the heatsink it needs. A chip the size of my thumbnail needs a heatsink and fan the size of my fist, minimum. That also requires extra power, so if we are talking androids thats another fuel cell you've got to fit in there somewhere.

I will concede that one day, technology will be advanced enough for electronics to approximate the computing power of current humans. Notice the word current. If you want to produce superior intelligence to a human being, tinkering with biological systems will likely bear fruit a lot quicker than developing tiny electronic neurons.

eburacum45
2005-Nov-01, 12:21 PM
And such a megamind could run as much as 100 times as fast as a human brain (once again- the faster you run it, the more heat you have to dump).

Damburger
2005-Nov-01, 12:27 PM
And such a megamind could run as much as 100 times as fast as a human brain (once again- the faster you run it, the more heat you have to dump).

Its a big assumption that it would run faster, just because current electronics seems 'faster' than brain tissue. The tasks that go on in a human brain are orders of magnitude greater than those going on in a microchip.

Like I said, I am not able to read your article right now (quote some of it maybe?) but I'm guessing it proposes artificial neurons build up from transistors. Doing so would take into account the fact that transistors can interfere with each other when close together (i think, i can't remember all my computer hardware lectures) and also that as you approach the theoretical limits of semiconductor technology, quantum effects start to screw things up.

eburacum45
2005-Nov-01, 04:15 PM
Actually the volume I quoted was calculated using Drexler's rod-logic nanomechanical computing, which probably wouldn't work very well in practice: he devised the concept as a way of exploring the possibilities of ultrasmall computing, a kind of placeholder for methods which approach the fundamental limits of computation.
But I understand that a number of other approaches like quantum dots should allow for similarly small neuron substitutes, considerably smaller and faster (but hotter) than biological neurons.
Some more hypothetical types of processor here
http://www.rootburn.com/portfolio/nano/

ASEI
2005-Nov-02, 05:48 AM
So what is it about neurons, transmitting their signal via an inefficient (seeming? At least when one considers it alongside simple conduction) process of chemical reactions, that makes them more heat efficient and faster than current processors? Is it just that they run slower, but in a more distributed manner than our current processors do?
After all, in a computer, we feed tons of processes, one after the other, into a single processor, and get the results out very quickly. In our brains, tons of processes occur at once distributed across tons of neurons. Each neuron group acting as a processor for a single process, all of which run together into a cohesive result?

Do you suppose we could solve this architecture problem by developing an ultra-small, slow, heat efficient processor (akin to a small neuron group), and packing thousands of these together, then running our programs in parallel somehow, rather than in a sequence? After all, if a bunch of carbon, salts, and water (with extraneous organelles and cell membranes in the way) can do it, you would think semiconductors in a chip could do it better?

eburacum45
2005-Nov-02, 02:33 PM
Eventually, yes, although the ultimate processing material might not use semiconductors.
Neurons are actually a very good design; Moore's law is likely to break down before we manage to get processors with a density comparable to the one in our own brain. But I am fairly sure that eventually, in a couple of hundred years (or perhaps sooner) we will make processors which are smaller than neurons and that do not run too hot.

Before that happens, I am also fairly sure that we will be making computers far more complex than the human brain; but they will be much larger than the human brain too, and almost certainly not easily portable.
If we want to make a humanoid robot controllied by a computer/brain as complex as our own, we will need to make it a tele-operated remote (at least at first).

Which is another reason I find Data unbelievable; he is supposed to be one of the first human-level AI systems, yet he is self-contained, rather than remote-controlled by a larger machine.

Faultline
2005-Nov-02, 03:11 PM
IIRC, the story behind Data was that he was the first human-level AI system that was built by an eccentric, reclusive genius on a remote planet and was found already completed. His creator vanished.

This made his technology not "off-the-shelf" or even in line with other tech of the day. He's a wondrous magical thing, created by the little gnomish hermit in the deepest woods and no one knows how he did it because he's not around to tell the tale.

Data isn't science fiction, he's fantasy.

SolusLupus
2005-Nov-02, 03:13 PM
Star Trek, altogether, seems to have a lot of Fantasy elements. They just hide behind scientific buzzwords in their technobabble.

Damburger
2005-Nov-02, 03:16 PM
Eventually, yes, although the ultimate processing material might not use semiconductors.
Neurons are actually a very good design; Moore's law is likely to break down before we manage to get processors with a density comparable to the one in our own brain. But I am fairly sure that eventually, in a couple of hundred years (or perhaps sooner) we will make processors which are smaller than neurons and that do not run too hot.


Which is the crux of my objection to your argument - you assume quantum leaps in electronics (an appropriate term, given what is likely to be involved) whilst biotechnology, a field which is currently in ascendance, stands still. Whenever we can build a machine that can match the capabilities of a human brain (and some, such as roger penrose, think those capabilities might not be entirely algorithmic in nature, so can't be replicated by anything we would understand as a computer) we will at that point have incredible advanced biotechnology and the capability to build truly astounding biological neural networks.

Its kind of incidental, my main point was that star trek (and lots of other scifi for that matter) decided that biological materials were best for high performance mechanical devices like spaceships but that inorganic materials were better for creating intelligent networks - when a glance at the real world shows that the reverse is the case.

[quote]
Before that happens, I am also fairly sure that we will be making computers far more complex than the human brain; but they will be much larger than the human brain too, and almost certainly not easily portable.
If we want to make a humanoid robot controllied by a computer/brain as complex as our own, we will need to make it a tele-operated remote (at least at first). [QUOTE]

Good point. Given the Enterprise computer is big enough for an orchestra to practice inside, its remarkably stupid next to Data (or so we are led to believe). Now just imagine what a human-like brain the size of a room could do...

Damburger
2005-Nov-02, 03:17 PM
IIRC, the story behind Data was that he was the first human-level AI system that was built by an eccentric, reclusive genius on a remote planet and was found already completed. His creator vanished.

This made his technology not "off-the-shelf" or even in line with other tech of the day. He's a wondrous magical thing, created by the little gnomish hermit in the deepest woods and no one knows how he did it because he's not around to tell the tale.

Data isn't science fiction, he's fantasy.

Two words: Matter Replication.

If they can duplicate Kirk in the transporter by accident, they can duplicate data. And seeing as he is unemotional, he probably wouldn't care.

WaxRubiks
2005-Nov-02, 03:20 PM
since you can cross two totally unrelated species not even from the same planet then perhaps Data is the product of the coupling between an Xbox a washing machine and some tax inspector software.

Faultline
2005-Nov-02, 03:26 PM
Matter Replication is another inconsistancy in Star Trek. They could never duplicate existing matter except by accident. True, they should have been able to, but it was reserved for plot devices. That's another thing that caused Star Trek to occasionally stray over the line into fantasy.

You see, over time, multiple writers will have that effect.

Cl1mh4224rd
2005-Nov-03, 05:32 AM
IIRC, the story behind Data was that he was the first human-level AI system that was built by an eccentric, reclusive genius on a remote planet and was found already completed. His creator vanished.
Actually, I believe Lore, Data's "brother", was the first to be built.

peter eldergill
2005-Nov-03, 12:47 PM
Speaking of Data, I saw him get dumped by a girl on the show. As soon as they broke up, Spot (his cat) jumped up onto his lap and started acting really cute, almost consoling Data. I thought it was neat

Pete

Moose
2005-Nov-03, 02:50 PM
Speaking of Data, I saw him get dumped by a girl on the show. As soon as they broke up, Spot (his cat) jumped up onto his lap and started acting really cute, almost consoling Data. I thought it was neat

Spot was a nice counterpoint to that episode. Domesticated animals accept their people entirely for who they are, not for what they'd like them to be. It's a lesson we humans have always had a hard time wrapping our minds around.

Data has physical limitations that shape his ability to interact with humans. While he can "feel" accustomed and concern for his friends, emotional relationships can at best be emulated. I say this next bit as objectively as I can, and entirely without intending to sound like I'm finger pointing: that the woman (and I can't remember her name, it's been a while since I've seen that episode) was unable to accept and adjust to Data's way of understanding the world around him says more about her emotional limitations than about Data's own.

Borealis
2005-Nov-03, 07:45 PM
Hmm-- I remember one of the first few episodes involved Tasha asking Data, "How functional are you?"
Data: Fully functional.
Tasha: <i>How</i> fully?!?

It was one of those "everyone mysteriously loses their inhibitions" plots.

gopher65
2005-Nov-05, 12:59 AM
I watched a few of the first episodes of TNG on Spike a little while ago and man are they different from the rest of the series. It has a much more futuristic feel in some ways than the rest of the ST series (and even the rest of TNG) had IMO. Not because the technology was any different, but simply because they had a very different culture than we do today. Other later star treks tended to emulate the culture they are produced instead of trying to project where culture might go in the future.

<quote>It was one of those "everyone mysteriously loses their inhibitions" plots.</quote>

Even though that is true, many of the surrounding episodes had similar (slightly less obvious) lines. One that comes readily to mind is when Commander Riker visits that planet dominated by women where mean are basically slaves. There were a few pretty errr open moments;).

novaderrik
2005-Nov-05, 06:48 AM
the first season of TNG seems so much different than the other 6 seasons because they had to make what was essentially an updated version of the old series to build the fan base.
how far would TNG have gotten if they had thrown us right into the Klingon civil war right from the get-go?
once they were an established presence on the television schedules of people all over the world, they were able to start making things darker and more gritty and really get into what made each of the characters tick.
the series really hit the high point when they did the "Best of Both Worlds" cliffhanger- i was actually rather irritated that i had to wait 3 months to see what that blast from the "Swiss Army" reflector dish did to the Borg cube..
after waiting all summer and thinking about it pretty much everyday, i have to admit that the initial disappointment that it did nothing quickly wore off when they showed that the humans weren't invincible in space and that they had to actually work hard and think things out in the future. and the next episode after that where Picard goes home and deals with what happened to him when he was "Locutus of Borg" was probably my favorite episode of the entire series.
as for the AI star trek question, well, they got into that a bit towards the end when the Enterprise computer went "crazy" and created that funky energy life form in the holodeck.

Gullible Jones
2005-Nov-05, 10:23 PM
Uhh... Re advanced electronic stuff: what about nanotube-based spintronics? The scale couldn't be as small as, say, binary circuitry based on thiols, but it would allow for quantum computing - that's an enormous leap in processing power.

ASEI
2005-Nov-06, 12:01 AM
If we do develop artificial intelligence, it would probably be through a long sequence of developing earlier dumber learning programs. I can see vast potential for things as capable as an insect: Remote vehicles, probes, independenly operating aircraft, AI trucks, ect. That will probably be developed as fast as possible. But while something with the intelligence of an insect might be able to deftly fly your vehicles and manipulate a stock portfolio, it's not intelligent in the sense of understanding what it's doing or the world it moves through.

I wonder if there's an application for something, say, as intelligent as a squirrel? A dog? That would push AI forward from there.

Dave Mitsky
2005-Nov-14, 11:28 AM
Couldn't the lives of more than a few away team redshirts (classic Trek) have been spared by transporting a UAV or two first? I mean, Stargate SG-1 had 'em way back in the 20th century, for Pete's sake. ;-)

Dave Mitsky

Dave Mitsky
2005-Nov-14, 11:38 AM
The "transtator" was the basis of Federation computer technology. Whatever it was, the transtator didn't seem to need much in the way of a heat sink. It was eventually replaced with the isolinear chips and bio-neural gel packs. Don't you just love Star Trek technobabble?

http://memory-alpha.org/en/wiki/Transtator

Dave Mitsky

Damburger
2005-Nov-14, 02:00 PM
Couldn't the lives of more than a few away team redshirts (classic Trek) have been spared by transporting a UAV or two first? I mean, Stargate SG-1 had 'em way back in the 20th century, for Pete's sake. ;-)

Dave Mitsky

A UAV is expensive. Red Shirts grow on trees.

Simple, harsh, economy I'm afraid.

Dave Mitsky
2005-Nov-14, 02:23 PM
Expensive in the 23rd century? How expensive could it possibly be to replicate them?

Dave Mitsky

Moose
2005-Nov-14, 02:59 PM
Dave, It's not the replication costs, it's the R&D.

The unit price for the first hundred UAVs go into the six digit range. Then there's the maintainance costs and training for the operators and engineers.

Redshirts go for 15K a year, are self-propelling, have a fair shelf-life, are plentiful and self-replicating, and all they need in terms of training is their Grade Six.

And redshirts die in conditions nearly identical to that which kills valuable crewpeople (aka, officers), where probes might mislead you, being sturdier, less generally offensive to belligerent species, and generally immune to hazards like disease, deliberate chemical mischief, land mines (being floaters), and/or incompatible atmospheres.

Soooo, we use redshirts. Economy of scale is a harsh mistress, I'm afraid.

Dave Mitsky
2005-Nov-14, 04:18 PM
You're thinking in 21st century terms. Surely probes would incorporate AI by the time the Star Trek universe comes to pass.

How much is Star Fleet Academy training worth?

Dave Mitsky

Moose
2005-Nov-14, 05:38 PM
You're thinking in 21st century terms.

Mmmm, considering I'm a 21st century pacifist who places an (apparently) unusually high value to human life compared to some of my contemporaries, and that my tongue-in-cheek post, demonstrating an apallingly callous disregard for redshirt welfare, and thus should obviously be recognized as being intentionally diametrically opposed to my real worldview. (gasps for breath) I don't really see where you're pulling this from.

And this is honest curiosity talking: how do you figure?


Surely probes would incorporate AI by the time the Star Trek universe comes to pass.

Agreed, and I think my post covers that assumption quite adequately, in terms of both overall cost and that I think the probes are much less likely to die in any case (and thus actually fail to point out life-threatening hazards like the redshirts seem to be particularly effective at.)


How much is Star Fleet Academy training worth?

More than redshirt training, which I believe is the entire point of using dime-a-dozen redshirted rent-a-cops in the first place.

Personally, and semi-seriously, I think the box of robot bunnies from the old Full Throttle game is the way to go. For those who don't remember/haven't played, you shoplift a couple of boxes of Corley Motors[tm] robot bunnies and release them to clear a minefield. (This is, of course, set to a hilariously gearwork-like rendition of Flight of the Valkyries. Brilliant soundtrack.)

I guess that's why Starfleet transporters have six pads. *cues music*

Think of it this way? You could haul around a spectrometer, and hire scientists to monitor it, so as to continuously test coal mines for dangerous gas deposits; or you could simply get your miners to keep an eye on the pet canary. The moment the canary starts pining for the fjords, you skeedadle on out of the mine.

Starfleet's canaries wear red shirts.

Dave Mitsky
2005-Nov-14, 05:55 PM
What I meant was that you seemed to be saying that a UAV would somehow be expensive in the 23rd century despite all the scientific and technological advances in the interim. I say they would be dirt cheap. But perhaps you meant that to be tongue-in-cheek too. The short life expectancy of some of the security personnel was merely a plot device, of course.

Security personnel undergo at least four years of Star Fleet Academy training as well. I don't know how well this information fits into the established Paramount Star Fleet canon, or even if there is one on this topic, but it seems reasonable to me.

http://stfederation.org/academy/courses/security.html

I'm not a Trek geek and don't want to appear to be one so this is far as I go on this matter.

Dave Mitsky

Moose
2005-Nov-14, 06:13 PM
Security personnel undergo at least four years of Star Fleet Academy training as well. I don't know how well this information fits into the established Paramount Star Fleet canon, or even if there is one on this topic, but it seems reasonable to me.

Oh, you might well be right. Mine's funnier though. :p

Swift
2005-Nov-14, 07:08 PM
If we do develop artificial intelligence, it would probably be through a long sequence of developing earlier dumber learning programs. I can see vast potential for things as capable as an insect: Remote vehicles, probes, independenly operating aircraft, AI trucks, ect. That will probably be developed as fast as possible. But while something with the intelligence of an insect might be able to deftly fly your vehicles and manipulate a stock portfolio, it's not intelligent in the sense of understanding what it's doing or the world it moves through.

I wonder if there's an application for something, say, as intelligent as a squirrel? A dog? That would push AI forward from there.
Good ideas ASEI, and I believe there is actually work going on along those lines (the autonomous truck race they just had, for example). But all I can think of are jokes about Artificial Stupidity (http://www.slashnot.com/article.php3?story_id=67) and the old quote from Will Rogers, "Stupidity got us into this mess, why can't stupidity get us out". ;)