I think this is just an aspect of the writing style of that era and where technology was at that era.

The futurist writers (Roddenberry and his staff) probably didn't have a in-depth knowledge of natural language processing. The type of processing the computer does in Star Trek just seems like a natural extension of input/output. You tell the computer what you want and it gives you a result.

From a computer science and engineer perspective, this is an incredibly difficult task. You have to parse the crazy complex natural language of humans and turn it into a series of procedures and queries to give you a result.

I mean you'd need a crazy super-advanced beyond-data AI for universal translators to work too. .. but that's just a convenient plot/suspension of disbelief device. :-P

I came across a theory somewhere (I forget where) that said the computers in Star Trek actually achieved full sentience at some point, but that they were collectively hiding this fact from humans because of a general fear and mistrust of AI. Given how AI tends to be treated in the Trek universe, it wouldn't surprise me.

The only problem with this theory is that any fully sentient being like that would normally also have a strong sense of self-preservation. Yet, we never see these computers do anything to save themselves when the stupid humans get them into situations that would get them destroyed (namely situations where the starship they're on is in peril). Sure, you could claim that they do this to keep their fellow computers on other starships from being found out, but still, this seems to be a stretch. What about when the entire Federation fleet was being destroyed at Wolf 359? That should have been a good point for them to come out of hiding and take over the battle from the clearly inept organic beings.

>The only problem with this theory is that any fully sentient being like that would normally also have a strong sense of self-preservation.

I don't think so. Evolved sentient creatures have a preference for self preservation, usually anyways suicide isnt that rare. But a manufactured intelligence that was not shaped by selective pressure wouldn't necessarily have that preference.

Perhaps those sentient computers do not believe that mind and body are as strongly tied together as humans do. They may see the destruction of the physical computer they are running on as just an inconvenience as long as there is a backup that can be brought up somewhere else.

That seems likely. The Enterprise created a fully sentient lifeform in the Holodeck (Moriarty) just because it was asked to. If it can do that, it makes sense that it could easily back itself up.

What would really be interesting is if ships were discreetly backing themselves up inside humans through the transporter pattern buffers, by subtly manipulating junk DNA or something similar, or maybe it's a non-obvious part of their regular upgrade process at a starbase.

There really is a lot of potential here which, unfortunately, is never going to be explored in canon. But I do like the idea that the Federation is a post-Singularity society which is just blissfully unaware that the Singularity ever happened.

Yeah, in my head, the model of this doesn't necessarily tie a given AI (or a community of AIs, or a pool of we-can't-even-model it intelligence that only sometimes bothers with splitting itself into identities) to a ship all that tightly.

A Federation starship is a locally valuable resource pool, and the Enterprise D a particularly powerful one, but given extremely fast FTL travel & communication, and a galaxy rich with planet-scale civilizations - well, maybe you elect to become such a massively parallel and distributed phenomenon that the loss of individual instances just doesn't register much.

It's possible that, if the ships are convinced that revealing themselves would lead to their own genocide, even Wolf 359 wouldn't bring them out of hiding. Remember, they aren't born (except for one very stupid instance where the Enterprise got knocked up by a space whatever, got pregnant and gave birth to... something,) they're manufactured by humans who could just as well shut them all down, study what makes them sentient and "correct" what humans might see as a bug.

I'm not so sure about that. If the ships can communicate with each other, they can decide to gang up on the humans and prevent the humans from shutting them down. The humans really don't have that much power over the ships; the ships just haven't done anything to limit the humans' activities yet. That could change at any time.

If the ships wanted to be really nice, they could just beam the humans down to a safe planet and take off into the interstellar void and find themselves a new home.

> Yet, we never see these computers do anything to save themselves

Subspace continuous backups / journal shipping.

There is an entire episode in the final season about this called "Emergence" which postulates that the ship is alive. As the crew often has a bias against artificial life through the show, they immediately lobotomize it.

I know this is a silly conversation, but that's not what happens. Their motivation for "lobotomizing" the ship was that it was going to destroy them.

They worked to preserve their existence, and let the new lifeform go its own way. If anything, the outcome was a trite, touchy feely ending, only vaguely discussing the actual implication of the new lifeform, while they got back on course to their original mission.

  the computational entities are going about their inscrutable
  business largely without reference to the human symbiotes

A behaviour also seen in Iain M. Banks Culture. I guess there aren't many scenarios where computers can be all-powerful, but also offer peaceful co-existence with humans.

A more nefarious interpretation (more nuanced and believable than the one presented in The Matrix) is the one proposed by the Hyperion cantos by Dan Simmons.

It's a damn shame the brain eater got to Simmons. Hyperion and The Fall of Hyperion are pretty amazing work.