The AI for the Enterprise D was run from three computer cores on the ship (two in the saucer, one in the engineering section), made up of Isolinear Chips subjected to a small warp field (iow, it makes computes using light faster than the speed of light). Subspace communication bandwidth is too limited (and potentially affected by latency, since it had to travel through repeaters throughout the galaxy) to provide realtime cloud computing as we experience it.

There are some cannon exceptions to this (such as in Nemesis where the subspace communication interruption affected the star charts), but even then the functionality of the ship was not impacted.

The Star Trek ships are very analogous to our own ocean-bound ships, where satellite communication is possible almost anywhere, but they don't rely on it.

So, yes, the AI is completely confined to the ship.

What about when someone was (permanently) transferred between ships?

Was there ever an indication that their AI-level data was transferred along with their personnel file? For example did the replicator know what food to offer them on day one?

If so, then it's seems reasonable to assume that the Enterprise's AI data was backed up at Federation HQ during routine maintenance, and that the "IT department" at Federation knew exactly what you liked to do on the Holodeck.

> did the replicator know what food to offer them

Through specific indication from the user. Recall the constant utterance of "tea, earl grey, hot"?

Ultimately, I imagine the user's information (documents, etc) was passed directly between ships, or through (as you say) Federation HQ.

> Enterprise's AI data was backed up

Ultimately, I think this is where AI will differ from ML. An AI won't have data that isn't a part of the AI - i.e. you couldn't separate out information specific to Picard from the rest of the AI code. An AI might be able to "scribble" down some notes about interacting with Picard and pass them off to another ship's AI, but the second AI would never treat Picard quite the same way as the first, even with those notes.

This stems from my belief that how ML interprets data is different from how an AI would. If you were to copy all of the data used to build a ML model and apply it again, you'd end up with the same ML model. An AI, on the other hand, if built twice twice from the same data would end creating two separate AIs.

Thank you for that very thorough answer!

I never got the feeling that there was a lot of mistrust of other federation people.

for example, the Star Trek universe didnt seem like a universe where you had to shop around for a trustworthy mechanic, who wouldnt overcharge or over diagnose. (e.g. headlight fluid)

Maybe the implicit trust of other people was integral to the AI being successful in that universe.

It's definitely a per-ship thing, because there are several episodes where they have to down/upload updated data to/from the Federation (at least in TNG canon).

I also think the comparison isn't perfect, because Federation vessels (in my mind) are similar to today's Navy vessels. All onboard systems are connected to other onboard, but opsec demands the ship systems not be influenced by external actors.

And a big point is that on a Navy-analogous vessel, it's reasonable to assume most activity on the ship is being monitored, for the safety of the ship and crew. The fact that the ship knows where everyone aboard is or that they can pull up the full Starfleet record for anyone who's ever been in Starfleet is not surprising, this is the military* and records and accountability are a big deal. But there's nothing to indicate that Federation civilians are monitored to that extent, and I'd argue enough episodes are strong on fundamental individual rights, that it's hard to imagine Federation life for civilians being a surveillance state.

* Don't start that argument. Seriously.

> But there's nothing to indicate that Federation civilians are monitored to that extent

The clearest example of extensive off-starship monitoring (within the Federation) that I can think of in TNG is a civilian (though, to be fair, a civilian in a role analogous to a "defense contractor"), Dr. Leah Brahms.

> I'd argue enough episodes are strong on fundamental individual rights, that it's hard to imagine Federation life for civilians being a surveillance state.

Actually, I'd say that its quite plausible that the Federation is a "benevolent surveillance state", that is, one with pervasive monitoring but a very low incidence of "serious" abuse (that is, the kind that substantially limits practical liberty -- casual intrusions on privacy may be more common.)

While the Federation seems keen on "fundamental individual rights", it doesn't seem to exactly mirror, say, some modern views on what those rights are -- and not just in terms of privacy.

I saw you give the Brahms example in another comment before I posted, so I am not surprised to see you bring it up. But I think you answered your own point. She worked heavily on the Galaxy-class starship's warp drive, which would be a relatively classified project that they would be very unhappy if the Romulans or any other hostile parties were intercepting data on.

And arguably, if she was working at Utopia Planitia Fleet Yards on the Galaxy class project, she presumably worked on a Starfleet orbital facility (technically, a number of facilities) over a span of years, where certainly enough data would be collected to make a poor replica of her personality, as in the show. I don't see anything suggesting specifically outside of basic biographical data, that she was being monitored in her civilian life.

La Forge, having interacted with that hologram extensively, and having surely read Starfleet's records... apparently didn't know she was married.

Even then, IIRC, it was limited to the ship(s). If you didn't want the ship to follow you, you could simply avoid entering the ship.

We're starting to lose that ability now.