Fianna wrote: ↑Wed Jun 01, 2022 5:34 am
Look at it this way, if you were a parent with access to a holodeck, would you ever bring your kid into a program like the one in a fistful of datas? A program where there are characters that are absolutely ready to assault, kidnap, and murder your child?
Bear in mind, Alexander's parents were Klingons.
On the other hand, it wasn't made by either of them. It was made, at least in part, by Barclay. And while he may not always have the best judgment, I also can't imagine someone who is so full of anxiety being interested in programs with sword fights or gun play that feature simulated opponents actually trying to kill him, restrained only by a line of code somewhere. He of all people would know that the whole thing should be fun and lighthearted.
I would also assume the majority of the program draws on existing templates and modules in the system, so as to make it easy to put something together without building it from scratch. And those would also be assets created in the federation, not by Klingons. But that's also putting more thought into it than writers ever did.
Especially when there is no good reason to not have the NPCs in every program (be they fictional characters, competitors, instructors, or sex toys) play the role of someone who exists to serve, and who prioritizes the safety, well being, and entertainment or education of the user.
I think it's misguided to view each NPC as a distinct entity. There is
one artificial intelligence running the whole program; the AI might generate many holographic characters, but it's the same computer mind behind all of them. And like most computers, even in Star Trek's future, it's very limited in its ability to infer what people want, and it takes commands literally. If you tell it to turn the safeties off, then it will do precisely that, no more and no less; if that means its holographic avatars might hurt or kill you, well, that must be what you want to happen, otherwise you would have left the safeties on.
While I think that would be more realistic, the show very much portrays holographic characters as operating independently. That's why the EMH and Moriarty can be self aware, while other holograms and the main computer are not. It's also presumably why commands to the computer can cease functioning, even while conversations with characters continue normally.
But even if they aren't separate programs or AIs, and they're actually all one big AI, that doesn't change the fact that the the holodeck is simulating believable interactive characters which will pursue their individual motivations. The program will play out and the characters will play their role no matter what, even if that means murdering real people. And the only thing stopping that is the safety protocol that does things like make the completely unnecessary bullets turn intangible before they can hit someone, instead of just never having a real bullet to begin with.
And if the computer is unreliable in interpreting what people want, that is all the more reason to add more layers of protection. If one wrong word will cause it to lock someone in a deathtrap or simulate an explosion that could blow a large hole in the ship, you should probably have as many levels of protection as you possibly can. Because the system you describe is one that combines the reckless instability of Jurassic Park,* with the user friendly experience of a monkey's paw.
___
*
Let's have our convoy of tourists travel in self driving electric rail cars, and stop them next to the t-rex enclosure secured only by flimsy electric fences. Sure, even a momentary power outage will put the giant predators on the honor system, but what are the odds of that ever happening? 
