[Space] Adapting "Honorverse" Concepts to GURPS

[Bengt]

Originally Posted by Ulzgoroth

Admittedly having ~0 improvement in AI over that time is not really plausible, but that's 'safe-tech' for you.

I think it makes a setting better to have some techs stall. Obviously if you are making a tech book for a generic rpg it's a good idea to suggest improvements for everything you can think of (and have room for). But when you put your future setting together it gets more interesting if you pick and chose a bit.

[Ulzgoroth]

Originally Posted by Gedrin

There are several references to the history of space navies. Saganami is the great naval hero, prior to Harrington, of the RMN. Several other ships and classes are named after people other than royalty. I'd guess the honored people are naval paragons or something similar. Haven and Anderman do the same. I imagine SLN does as well, but they go for about a few hundred years without a serious competitor, which might make heroes harder to find.

Saganami is possibly the earliest RMN hero, but he's not the only one mentioned.

But there's over a thousand years of naval history from which we know barely any names.

Originally Posted by Bengt

I think it makes a setting better to have some techs stall. Obviously if you are making a tech book for a generic rpg it's a good idea to suggest improvements for everything you can think of (and have room for). But when you put your future setting together it gets more interesting if you pick and chose a bit.

It doesn't seem plausible.

[Bengt]

Originally Posted by Ulzgoroth

It doesn't seem plausible.

So you see proper AI as a given development? I don't.

[ericthered]

To be fair to both sides, I don't see proper ai as a given, but we are getting to the point where it can be reasonably faked with good voice recognition and a big enough database.

[Fred Brackin]

Originally Posted by ericthered

To be fair to both sides, I don't see proper ai as a given, but we are getting to the point where it can be reasonably faked with good voice recognition and a big enough database.

That's sort of mentioned in A Rising Thunder.

It's stated that there's a lot of dedicated apps that can do a decent imitation of a human within their area of specialization. I'd simply assume that there's probably a shortage of people (or at least important people) whoa re so bored they try and natter on about the weather with artificial taxi drivers.

[Gedrin]

I've noticed that computing/networking seems less advanced than one might expect in the Honorverse. It might be a result of communications lag time and the diaspora. The technical problem of slower than light communications making much of the setting we see detached from conventional internet, and the diaspora producing numerous variations in protocols, dispersed populations, and self contained networks. It's possible they've just not figured out how to not make berserkers, but, as with genengineering and nano, it's likely there's a universal aversion to tech without a clear off switch.

IRL I'm pretty sure it's just DW's desire to avoid the subject.

[Ulzgoroth]

Originally Posted by Bengt

So you see proper AI as a given development? I don't.

Depends what you want 'proper AI' to be. (The scale running from 'we've had it for over a decade' to 'only almost certain'.)

I think AI better and cheaper than any I've seen evidence of in the Honorverse is possibly present day, and if not will be with us in a couple decades at the outside. They have a really enormous dependency on centralized, human-directed fire control.

Originally Posted by Gedrin

I've noticed that computing/networking seems less advanced than one might expect in the Honorverse. It might be a result of communications lag time and the diaspora. The technical problem of slower than light communications making much of the setting we see detached from conventional internet, and the diaspora producing numerous variations in protocols, dispersed populations, and self contained networks. It's possible they've just not figured out how to not make berserkers, but, as with genengineering and nano, it's likely there's a universal aversion to tech without a clear off switch.

IRL I'm pretty sure it's just DW's desire to avoid the subject.

Major worlds have populations more than big enough to make internal networking a huge thing, one would think. Interworld communications have considerable unavoidable friction.

The books don't spend much time in civilian contexts, though. And, of course, On Basilisk Station was published in 1993, so there's a certain 'reality marches on' aspect.

[doulos05]

Originally Posted by Ulzgoroth

I think AI better and cheaper than any I've seen evidence of in the Honorverse is possibly present day, and if not will be with us in a couple decades at the outside. They have a really enormous dependency on centralized, human-directed fire control.

Yes, but look at what they use it for. The use human directed fire-control to make intuitive judgments about which targets to prioritize. They don't seem to direct fire for each individual missile, it seems instead they lock in their priority targets and warhead selection and let the computer sort of which bay fires at which target at which time. In each of the situations I can recall where the fire control teams feature, they are executing pre-configured firing patterns (programs) and designating targets for said programs (variables). In other words, the fire control teams are ordering expert AI systems to allocate weapons fire given the constraints of targets they have selected.

Also note that whenever a fire control team is forced to adapt/react to the tactical situation without computer assistance (like the religious fanatics on that Sultan class over Grayson), they regularly under-perform expectations (by a large margin) indicating that the computers are doing far more heavy lifting than you're giving them credit for. Before you say, "They weren't a trained crew." The book explicitly states that their lack of training meant they were doing by hand all the ECM and ECCM decision making that the computers were supposed to do.

[Ulzgoroth]

The book I just read (Echoes of Honor) reminds me that Honorverse reaction thrusters are canonically insanely powerful. A secondary propulsion system that's basically never used for anything except docking maneuvers and seems to occupy no significant portion of the ship can push major warships at 150 g!

Also, fuel for fusion reactors is an endurance issue for warships that may be solved by using fission reactors instead.

Originally Posted by doulos05

Yes, but look at what they use it for. The use human directed fire-control to make intuitive judgments about which targets to prioritize. They don't seem to direct fire for each individual missile, it seems instead they lock in their priority targets and warhead selection and let the computer sort of which bay fires at which target at which time. In each of the situations I can recall where the fire control teams feature, they are executing pre-configured firing patterns (programs) and designating targets for said programs (variables). In other words, the fire control teams are ordering expert AI systems to allocate weapons fire given the constraints of targets they have selected.

I can't think of any sign anywhere in the books that the computer has any control whatsoever over weapon allocation, except in point defense roles. Basically, I don't think anything in that paragraph is correct.

And if it were, it would still be unimpressive in the extreme. The place you need a better AI isn't in the tac officer's console doing their job for them. It's in the missiles, making them less horrendously stupid and reliant on telemetry from the launching ship.

Which Manticore eventually implements...in the form of special oversize missiles dedicated to controlling groups of other missiles.

Originally Posted by doulos05

Also note that whenever a fire control team is forced to adapt/react to the tactical situation without computer assistance (like the religious fanatics on that Sultan class over Grayson), they regularly under-perform expectations (by a large margin) indicating that the computers are doing far more heavy lifting than you're giving them credit for. Before you say, "They weren't a trained crew." The book explicitly states that their lack of training meant they were doing by hand all the ECM and ECCM decision making that the computers were supposed to do.

Actually, you seem to have gotten that backwards. Their lack of training meant that they didn't know how to manage their ECM by hand, so they had to let it run on a (critically inadequate) fully automatic mode. Proper ECM use appears to involve the tactical department constantly re-tuning their ECM scripts.

I am in no respect saying that the computers aren't absolutely essential. Operating without computer support is something which does not happen, because if you are without computer support, your ship doesn't operate! (Very pointedly observed in Echoes of Honor in more than one place.) What I am saying is that the computers are incapable of or incompetent at many things that computers shouldn't be.

[Bengt]

Originally Posted by Ulzgoroth

Depends what you want 'proper AI' to be. (The scale running from 'we've had it for over a decade' to 'only almost certain'.)

Yeah, "proper" wasn't very clear. In this context when I hear AI I think of what is basically a character, e.g. EDI in Mass Effect 2.

I think AI better and cheaper than any I've seen evidence of in the Honorverse is possibly present day, and if not will be with us in a couple decades at the outside. They have a really enormous dependency on centralized, human-directed fire control.

Contemporary "AI" are not necessarily very noticeable if they are mostly advisory. In the particular case of live fire I would find it plausible if they were strictly advisory. YMMV.