Would CASIE implants be common in THS?

[Tyneras]

Originally Posted by vicky_molokh

Why wouldn't we outsource other things that require raw power in some area, and remain to manage the things that provide this raw power?

Well, considering we would likely invent managing AI first (probably for really efficient office buildings or something similarly mundane) the management part may get outsourced pretty fast, or maybe first. Then what do you have? Nothing! Ha! The management has outsourced themselves.

I apologize for going off-topic with this bit (although that seems to be the thing with THS threads, they go off-topic or die early on).

However, I do think that a CASIE implant would be great justification for something like Shy (Mitigator, CASIE implant, -60% (Electrical, -20%); -48%). My numbers may be off, running from meat-memory here.

[vicky_molokh]

Originally Posted by Tyneras

Well, considering we would likely invent managing AI first (probably for really efficient office buildings or something similarly mundane) the management part may get outsourced pretty fast, or maybe first. Then what do you have? Nothing! Ha! The management has outsourced themselves.

That's kinda what I was thinking in some other setups. In settings where AI IQ depends heavily on the size of the computer available (or even worse, depends linearly on processing power!), the CABAL approach is very good. (I.e. having a single big AI doing high-complexity work, with cheaper humans or dumb bots performing its orders.)

But in THS, where AI IQs above 12 are problematic, racking up a single skill, especially skill that enjoys good scaling from sensory ability, is cheaper and more efficient. Besides, it is canonical that THS has 'bot-bossing' as a job.

[wellspring]

Originally Posted by vicky_molokh

Um, they're so essential that we developed computers, engines and the like to do these things better than us. In fact, ability to perform these actions is so essential that if we fall back onto the human level of ability (be it ability to solve complex math problems or push around carts weighting tonnes), TL will drop.

You're not addressing problem-solving skills so I'm not sure what you're trying to say here. Or are you agreeing with me?

Originally Posted by vicky_molokh

I wasn't actually thinking of training simulations - I was thinking of something closer to book learning. Specifically, the kind of stuff Lightman & Co did in the series - watch thousands of videos of people in different emotional states, telling truth and lies etc., and looking for universal patterns. In the age of THS, I'm pretty sure gathering such videos isn't very hard, especially for a company that gains profits from selling NAIs trained on those videos. And the amount of material throughout the 150 or so years of video technology is definitely more than enough for a full-time AI.

Not only do different people lie in different ways, but the way in which we employ deception evolves over time due to the ebb and flow of different influence strategies and the counter-acting responses. This is called the Persuasion Knowledge Model. Almost by definition, an attempt to use a model of persuasion tactics fitted to historical data will produce inferior results. Even if you model individual-level parameters, you're making an out-of-range forecast.

(More generally, book learning isn't a bottomless pit of expertise in any field. You'd have a similar problem trying to raise a martial arts skill by reading martial arts books and watching martial arts movies but never actually spending time in the dojo. Learn-by-reading in some fields is simply deeply inferior to learn-by-doing.)

And again we get to the problem of comparative advantage. We both agree that LAIs and NAIs can spend CP on social skills (Though I argue that it requires training time spent with a human or experienced SAI). LAIs have a penalty to social skills, and advantages that greatly help them with mathematical skills. So why not use the same techniques to get LAIs and NAIs to replace scientists and engineers? What does a human (or for that matter, an SAI) bring to the table to justify the considerable expense of training and retaining them?

If there's no ceiling on book learning, then NAIs can replace everyone else in any field (as Tyneras rightly points out). They'll replace techies first (where they have an advantage), before the other fields. If NAIs are going to be doing your socializing for you, it'll only be after all the other fields are conquered. Set a ceiling (as I do) for book learning, and you stop earlier-- but either way, social skills are last to be replaced.

Now Xplo tries to claim that this won't happen because "they lack initiative and property rights and such". But Vicky's argument is that social skills are completely modeled by the bonus and penalty structure in the Basic Set, and that the system is mechanically complete. If you accept that, then "initiative" is just a series of skill penalties that are part of the template that can be overcome with brute force skill-mongering. Without a ceiling on book learning for applied fields, there's no need for humans anywhere (possibly true in the long run especially with SAIs, which are human-equivalent).

If you don't believe this (as I don't), you're arguing that the rules are representing the setting, and that the "learn by watching instructional vids" approach isn't a bottomless well of knowledge. That the Basic Set rules have to be applied with a grain of salt when it comes to the limits of things like reviewing archival footage as the basis for developing social skills. In that case, an AI might assist a human as I described in my post above, but wouldn't replace a human's social acumen.

[vicky_molokh]

Originally Posted by wellspring

You're not addressing problem-solving skills so I'm not sure what you're trying to say here. Or are you agreeing with me?

I'm saying that the ability to perform certain actions is essential to people, but this is not a reason not to outsource these abilities to tools. To the contrary, an essential action will be outsourced to a tool that performs it better than the person who uses the tool (or does it reasonably good without the human wasting time on maintaining the innate ability).

Originally Posted by wellspring

Not only do different people lie in different ways, but the way in which we employ deception evolves over time due to the ebb and flow of different influence strategies and the counter-acting responses. This is called the Persuasion Knowledge Model. Almost by definition, an attempt to use a model of persuasion tactics fitted to historical data will produce inferior results. Even if you model individual-level parameters, you're making an out-of-range forecast.

That would require Body Language and Detect Lies to be /TL skills. That would be very weird. The whole point of Body Language is the study of non-conditional reflexes associated with different mental states. Non-conditional reflexes, unlike conditional ones, do not depend on upbringing like that.

Originally Posted by wellspring

(More generally, book learning isn't a bottomless pit of expertise in any field. You'd have a similar problem trying to raise a martial arts skill by reading martial arts books and watching martial arts movies but never actually spending time in the dojo. Learn-by-reading in some fields is simply deeply inferior to learn-by-doing.)

And I'm not trying to teach the AI to do martial arts - merely analyse them.

Originally Posted by wellspring

And again we get to the problem of comparative advantage. We both agree that LAIs and NAIs can spend CP on social skills (Though I argue that it requires training time spent with a human or experienced SAI). LAIs have a penalty to social skills, and advantages that greatly help them with mathematical skills. So why not use the same techniques to get LAIs and NAIs to replace scientists and engineers? What does a human (or for that matter, an SAI) bring to the table to justify the considerable expense of training and retaining them?

I'm not saying that they are irreplaceable. In fact, I think it was DryaUnda to be the first one to come up with the realization that a LAI specialized in Engineering can replace a large group of human techies of various specializations.

Originally Posted by wellspring

If there's no ceiling on book learning, then NAIs can replace everyone else in any field (as Tyneras rightly points out). They'll replace techies first (where they have an advantage), before the other fields. If NAIs are going to be doing your socializing for you, it'll only be after all the other fields are conquered. Set a ceiling (as I do) for book learning, and you stop earlier-- but either way, social skills are last to be replaced.

What I'm saying is that by current discoveries, Body Language seems closer to semi-technical skills like Physiology than to the fuzzy ones. And by extension, Detect Lies might be too.

Originally Posted by wellspring

Now Xplo tries to claim that this won't happen because "they lack initiative and property rights and such". But Vicky's argument is that social skills are completely modeled by the bonus and penalty structure in the Basic Set, and that the system is mechanically complete. If you accept that, then "initiative" is just a series of skill penalties that are part of the template that can be overcome with brute force skill-mongering. Without a ceiling on book learning for applied fields, there's no need for humans anywhere (possibly true in the long run especially with SAIs, which are human-equivalent).

That's a good question. What exactly does Slave Mentality (as an example of Lack of Initiative) do that isn't explicitly spelled out in the book?

[jason taylor]

Originally Posted by Jürgen Hubert

While I can see a use for such AIs, I really don't see why they need to be implanted. There seems to be a general reluctance to implant stuff that can function just as well as a wearable - possibly justified that most people expect to live through many, many generations of hardware (unlike in the Cyberpunk genre, where many people don't expect to live long anyway).

That was one objection I had to implants, IMTU aside from the in-verse prejudice against some technologies. While an implant might be useful for a command program it was hard for me to see how it would be all that more useful for a starship captain then a cyberhelm fitted over the head to translate information into a comfortable form(by making it appear as if space was surrounding and the captain was riding on top of the starship as if it was a horse).

I do use implants for those captains who prefer it. I also use them as security modules. They can contain passwords to forbidden computer files for instance. Compulsory implantation is also convenient substitute for prison for some crimes; implants never teach the convict to be a cleverer criminal the way being cloistered with other criminals might and they don't subject him to physical bullying as other criminals would, although Social Stigma(convict) is a necessary implication.