AI in Startrek

Flectarn · Post by **Flectarn** » Sun Jun 29, 2008 1:27 am

So, I've been wondering recently why we don't see much in the way of robots/androids in the Federation and major surrounding powers besides Data. After all, they extremely advanced computing, to the point that the ship realy does seem almost like a character. Moreover it seems like their constantly making sentient beings by accident in the holodeck (Moriarti, The first time Picard runs the Dixon program, The Doctor). And the federation certainly seems interested in having the technology, so just what is it that prevents them from obtaining it?

Cocytus · Post by **Cocytus** » Sun Jun 29, 2008 3:02 am

The most likely objection to creating more is a moral one, the very one debated in "Measure of a Man," which establishes conclusively that Data is not the property of Starfleet, but has the right to make his own decisions. Creating more Datas, with superior strength and endurance, easily trained and easily replaced, would prejudice Starfleet towards using them for the dirtiest, most dangerous missions, and thus a race of Datas would be enslaved. I'm also inclined to say it would be prohibitively expensive. Soong managed to create several (Data, Lore, B4, and a few tries Julianna Tainer mentions in "Inheritance") with only the resources of the Omicron Theta colony, but then again, Soong was a genius, and the process of fabrication would probably be more complicated and time-consuming for normal engineers. But mainly the objection is a moral one.

As for why other powers don't, I suppose they just haven't had a Soong come along yet. Certainly the Romulans and the Cardassians would jump at the chance to have armies of androids. The Klingons probably wouldn't, however.

Flectarn · Post by **Flectarn** » Sun Jun 29, 2008 3:57 am

The problem with the moral objection route is that it seems like they should have been able create AI, even if it was less advanced then Data ultimate was. Measure of a Man shows that Starfleet was going to default to a position holding that he was property. I'm not really as concerned with creation of more Soong androids, as i am with basic bots with, for instance a Moriarty type self awareness, where the computer literally created a sentient intelligence because it was asked to.

Post by **Mike DiCenso** » Sun Jun 29, 2008 4:30 am

After a manner of speaking, Starfleet did intentially create AIs. The EMH doctor and other holograms like him are effectively "hardlight" androids. As with Data, many of the same moral dilemmas have surfaced. Effectively we have seen where Starfleet and the Federation have gone borderline into the AI slavery route of the SW Republic/Empire with their 'droids by having holograms shown doing dangerous, tedious, and nasty work.
-Mike

Cocytus · Post by **Cocytus** » Sun Jun 29, 2008 4:45 am

Anything with a demonstrably sentient intelligence, be it humanoid or not, will be refused on those same moral grounds, as were the Exocomps in "Quality of Life." And, as you noted, the computer is capable of such effective self-regulation that helpers with only basic commands, incapable of independent though, probably aren't necessary.

Until the Doctor, of course, no hologram had ever demonstrated sentience. They simply hadn't been allowed the time to do so. Moriarty was an extraordinary fluke, and a dangerous fluke at that. There was Vic Fontaine, but O'Brien admitted he was a special case. I'm willing to bet Starfleet conducted a thorough reappraisal of its position on holographic sentience after the "Author Author" court case.

SailorSaturn13 · Post by **SailorSaturn13** » Wed Jul 02, 2008 2:41 am

Actually, it's easy. Most of sentient AIs turned out to be malevolent and uncontrollable like M-5. Data was the first AI to be human-friendly - and even he had an evil copy. And the last thing Federation needs is to create a Cylon problem - in addition to Borg and the augments. Thus the holograms only have basic intellignce mechanisms and that's it. Holodoc was actually designed to run 1500 hours max and to be replaced or at least reset aflter that.It was only due to Torres's decision to merge holodoc with another hologram to give him place to "grow" that allowed it to become fully sentient.
Romulans likely have similar problem: how to make an AI sentient and capable of true intelligence and yet make rebellion impossible? Most likely (as Lem nodded) there simply is no way - "Either a human, with free will, including ability to go bad, or a puppet".

Post by **Mike DiCenso** » Wed Jul 02, 2008 3:25 am

The episode "The Offspring" where Data creates a daughter android shows that it actually is fairly complex creating a Soong-type android. While the body of the android is apparently not the limiting factor, the creation of the positron neural nets are.
-Mike

Praeothmin · Post by **Praeothmin** » Wed Jul 02, 2008 1:53 pm

SailorSaturn13 wrote:Thus the holograms only have basic intellignce mechanisms and that's it. Holodoc was actually designed to run 1500 hours max and to be replaced or at least reset aflter that.It was only due to Torres's decision to merge holodoc with another hologram to give him place to "grow" that allowed it to become fully sentient.

Except that the EMH on the Prometheus seemed to have a personnality as develloped as that of Voyager's EMH, and he wasn't designed by Torres...
In fact, the Holographic miners seemed to also have develloped a personnality and were seemingly unhappy with being relegated to mining operations.

Roondar · Post by **Roondar** » Wed Jul 02, 2008 2:16 pm

If the AI's created by the computer of ST are near-sentient (or at least acting like it) that has some serious repercusions on computing power.

The ability to project a bunch of 'sentient' holograms from a single computer is impressive to say the least.

Who is like God arbour · Post by **Who is like God arbour** » Wed Jul 02, 2008 4:49 pm

Praeothmin wrote:
SailorSaturn13 wrote:Thus the holograms only have basic intellignce mechanisms and that's it. Holodoc was actually designed to run 1500 hours max and to be replaced or at least reset aflter that.It was only due to Torres's decision to merge holodoc with another hologram to give him place to "grow" that allowed it to become fully sentient.
Except that the EMH on the Prometheus seemed to have a personnality as develloped as that of Voyager's EMH, and he wasn't designed by Torres...
In fact, the Holographic miners seemed to also have develloped a personnality and were seemingly unhappy with being relegated to mining operations.

The question is, if they realy have personality or is their program simulating personality?

We for example know from Dr. Zimmerman, that there were several marks of EMH's before the LMH was developed. One complain about the EMH mark I was, that is has lacked empathy. The logical conclusion would be to write a few subroutines to simulate such empathy. That would be simpler than to program real empathy. Such subroutines could have implemented in EMH's mark II, III or IV and maybe even implemented in the EMH mark I afterwards (but not the one EMH mark I, that was lost in the delta quadrant).

Praeothmin · Post by **Praeothmin** » Wed Jul 02, 2008 8:55 pm

Well, then the programmers did one hell of a job... :)

I do agree it's possible that they programmed these "emotions", but for a collection of Forcefields and programs, displaying fear should not be part of that programming, and yet, in ST: FC, when the Borg come in Sickbay, the EMH does display what we could call fear.

On the Prometheus, the EMH mark whatever did seem apprehensive at the idea of taking out the Romulans by themselves.

I don't think these things would be programmed on EMH holograms, since they have no use and would take up spce better used for additionnal medical knowledge, for example.

Who is like God arbour · Post by **Who is like God arbour** » Thu Jul 03, 2008 4:56 am

Praeothmin wrote:Well, then the programmers did one hell of a job... :)

I do agree it's possible that they programmed these "emotions", but for a collection of Forcefields and programs, displaying fear should not be part of that programming, and yet, in ST: FC, when the Borg come in Sickbay, the EMH does display what we could call fear.

On the Prometheus, the EMH mark whatever did seem apprehensive at the idea of taking out the Romulans by themselves.

I don't think these things would be programmed on EMH holograms, since they have no use and would take up spce better used for additionnal medical knowledge, for example.

Or it could be part of their programming not because they would have use fot it themself but to be able to recongnize and understand the fear of their patients.