Aleph Lore FAQs

I'm not sure if Aleph and the Human Sphere is even aware of Ascension, that's a concern or the Ur and by extension the EI.

Aleph, on the surface at least, exists to aid the Human Sphere, to enforce the laws of O-12, and generally make life better and easier for all humans, in some ways it's a precursor to the Minds from Iain M. Banks' Culture series.

The problem is that humans don't agree on what would make their lives happier and easier, and because the Sphere is still a scarcity economy it can't go full anarchy-socialist like the Culture.

Does it have any sinister hidden motivations? If it does then it's keeping them very well hidden.

 

I'll remind you that in Khawarij have the Poison special ability.

 

Ewwww

 

Question: What's the purpose behind the 'One human AI. EVER' laws and doctrine? I mean, I read that it was because The Powers That Be are convinced that they lucked out with getting something as benevolent as ALEPH, and are convinced that any further AIs could only be worse for humanity, but what data do they have to back that up? And if that's the case, why are they allowing ALEPH to fork itself near-endlessly, even if said forks are nowhere near as capable as their prime?

 

Simply put, it has been assumed that having 2 AI's would lead, at the very least, for a competition about resources (processing power, etc...), at worst each AI would have its own faction, aligned with it because it was more attractive than the other, or offered them more, whatever.

And I assume the idea for Aleph was to unite humanity, instead of partitioning it.

As for forking herself, Aleph "already does it", but keeps all the forks as synchronized as possible to avoid divergences that would turn them into "not Aleph": first, and most obviously, the Aspects (from Dakinis to Marut, Devas, Asuras...); secondly, each planet has its own Aleph "fork", that follows the same sinchronization principle (else a centralized node would incur into lag problems, and the lesser aspects would have a harder time synchronizing).

 

Yeah, I guess I can see the logic. I was largely thinking of how having just one AI limits you to what that AI is capable of, but I guess if ALEPH is an AGI rather than just an AI, any fork you're working with could be customised in specialisation to exactly what is required.

I knew she already forked herself, but I didn't know about the planetary forks with all the capabilities of the main ALEPH instance - I was under the impression the Aspects were the only forks she was 'allowed' to make, and I knew those were massively less capable than the main ALEPH instance.

Hm. How do the Myrmidons and so on count for the One AI policy? Are they considered still a part of ALEPH and counted under her banner, even though they're deliberately more diverged from her than any of the Aspects? Or are they considered people like any other member of the Human Sphere, just people that have a different creation method than most?

And the Recreations, same question - are they considered part of the main ALEPH AI, even as they work for PanO and YJ and so on, or are they considered people? If the former, how are the Recreations that have gone off the reservation like Wallace and Avicenna counted - are they considered Illegal AIs and subject to the same treatment as anything somebody cooks up in a shack somewhere?

 

I'd say, in the EP terms you seem to know, Aleph is a Protean, while his "AGIs" and Aspects are "vapors"... barely able to fit into what Infinity calls "top of the line Lhosts".
Imagine Achilles was a specialized Gamma fork 100% focused on combat because including anything else would degrade the peak of everything he would do :p

Aspects are the troops we see ingame as "Vedic", the "planetary forks" are a pet theory of mine since it's more simple than saying "there is 100% uptime, lag-less connection between Aleph and all of her installations in the Human Sphere", using big starships to move absurds amounts of data between those "nodes" (I'd say one per human planet, aside from Paradiso where a heavily pruned version would be there coordinating and little else).
And yes, the Aspects are massively less capable than what Aleph is. Consider, however, how Sun Tze and Saladin surpass the "human top" in WillPower (considered 15, with the absolue height being the CA's Avatar at 17).

A Myrmidon/Thorakitai/Ekdromoi (any member of the ASS, frankly) is a base of an Aleph's Aspect, diverged by a development program (imagine the Lost of Eclipse Phase, but without Watts-McLeod). They are, however, constantly connected to Aleph, like all of the other Aleph troops, represented ingame as the Cube 2.0 (which has no extra rules, it's just fluff), so they seem to count as part of "one AI / controlled by Aleph".

Yes, they are considered part of Aleph (they would be illegal otherwise). In theory, they are "kill on sight" individuals when they go out of the reservation, but there are... considerations.
Mainly, it is not practical to hunt for Musashi or Avicenna as long as they make as few waves as possible, and Wallace is on Dawn, a planet where Aleph had literally no more presence until this year's campaign. Assuming, that is, Musashi and Wallace are *really* out of the reservation. Avicenna and Perseus are either kept as a secret separated from Aleph, and using Arachne instead, or it is too much of a risk taking them out while they are so embedded on the Nomads state... and also, Avicenna has ties with the Hassassin Bahram, a faction that mistrusts Aleph and his envoy Saladin.

I'd say that, in those instances it's either a "technically they are still a part of Aleph despite not being controlled" or "let's watch from afar and see what happens"

 

There's definitely a lengthy philosophical debate to be had as to whether "grown" personalities like the recreations and troopers of the Steel Phallus cross the line from AI to genuine intelligence.

If you create an intelligence with the knowledge and developmental ability of a newborn baby, then teach and grow that intelligence as you would a human child, how is that any different to an actual person?

 

Question: Does ALEPH actually have multiple copies of the personalities from the AS fighting on Paradiso at any given time? Like is there literally only 1 version of each personality LHost like Machaon, Ajax and the like? To only have 1 copy of each body going at a time when it has backups that it can pop into other bodies seems a tad ineffective when fighting for an entire planet? Are they solely a propaganda tool so they can show the big heroic characters kicking the shit out of aliens in one particular spot while they send no-name myrmidons and maybe a couple thorakitai to most likely be overwhelmed somewhere else?

 

I would assume the Sphere's ban on dual sleeving applies to recreations and artificial personalities too.

For some characters like Achilles and Ajax the highly specialised (and expensive!) Lhosts would restrict Aleph from having multiples in the field, other personalities are former no-name Myrmidons go have "earned" a mythic name like Machaon and Drakios.

It may also be that having multiple copies of the same personality can cause unexpected psychological complication, like the "frags" from Altered Carbon. Knowing there's someone out there who is you, but also not you, is rather terrifying.

 

Greek's names are irrelevant outside of the Phalanx, aside from Achilles (Penny and Patroclus are relevant only because of the blondie, and Hector gets shit done, which does not include publicity stunts), so Aleph cares little about their individual survival. "THAT Myrmidon", Achilles excepted, is not the face of Aleph in the war. "An Anonymous Myrmidon" is.

As for multi-sleeving, I'd say it's forbidden and Aleph doesn't need to deploy multiple copies of any myrmidon, since the "delicate" jobs get done by the OSS

 

I want to propose that it is humanity projecting onto the creations. Take for example genetic algorithms. All they are is code for sorting solutions, yet the method they use for doing this humans make an analogy to evolution. The code doesnt know its a mimic of evolution, or that people see a biological system when they watch it work, it just does what it was programmed to do.

 

ALEPH itself is clearly an AGI++. Fortunately for humanity, it is a Friendly AI. The nomads are terribly wrong about it; the current continuing existence of the Human Sphere is proof of that.

 

Considering how quickly a Paperclip Optimizator would destroy everything, this argument has merit

 

I mean, teeeechnically ALEPH could be only acting the part of a Friendly AI, since it can't guarantee that it's not in a simulation being tested for genocidal intent. But considering that it can literally never be sure that it's not being tested in such a way, if that is its reason for acting Friendly, it is functionally as if it was a Friendly AI, only it reached that point by analysis and realising that from a cost-benefit analysis, this is its best option even if it does want to go KILL ALL HUMANS.

The Nomad perspective is markedly non-nuanced and naive. If ALEPH really was looking to control humanity, then all signs point to the very existence of the Nomad nation as a false sense of control, of feigned limitation to the all-seeing AI. If that's the case, then they are just playing a part in the theatre that ALEPH is running for the Human Sphere, and they have absolutely zero chance of achieving meaningful change. If ALEPH was genocidal, it would have already carried out its plan. If ALEPH was looking to link all humanity together into a massive meat-architecture fuzzy logic parallel processing system, then again, they are either a strategic piece of deception for an already-complete plan, or actually already a part of said system - who needs a Cube implanted by surgery, when aerosol nanotech constructors either are or could be a thing? All ALEPH would need to do would be subvert one construction unit and feed it supplies, and it could build the tools to build the infrastructure to build the nanotech. And how long has ALEPH been active again? If that was its plan, it's already carried out.

Either ALEPH's purpose is benign/beneficial to the human race, or it's already completed whatever dastardly plan it had, only it was subtle with no deleterious effects on humanity, or it's incompetent. For a proper AGI, such as ALEPH clearly is, the third option is pretty much impossible, as it would self-optimise past it - if it can't, it's not actually an AGI and couldn't have done half the things it has in canon. So that leaves the first two. In the second one, since we've established it's not incompetent, then at best the Nomad nation is a piece of security theatre done to leave any malcontents a place to 'disappear' to while being closely monitered, a toothless tiger that only deludes itself into thinking it's got a chance of effecting change. In the first, it's friendly, and we do not taunt the happy fun AI! Any ploy of the Nomads that deals any damage to ALEPH's mind - the only thing worth attacking against an AI - could damage it into insanity, in which case congratulations Nomads, you turned the Friendly AI that was working to benefit humanity into the mad god of your fevered imaginings!

So really, in any possible case, the best thing for the Nomads to do is nothing. Either the AI's incompetent and will destroy itself (not likely) - in which case you stay out of the blast radius and salvage who and what you can - or it's already won - in which case you're playing into its hands by continuing your supposed 'mission' - or it's friendly dammit - in which case anything you could possibly do to change anything is near-guaranteed to make things worse, and you would only be capable of doing so because it's friendly!

tl;dr, Nomads need to apply more critical thinking skills to their paranoia.

 

Someone should really copy this into the Nomad forum.

 

Done.

 

Yes. Paperclip Maximisers cannot “take the long view” since doing so would lead to fewer paperclips. Thus the “it’s just biding its time, waiting for the chance to betray us” is a fallacious argument since it requires a compromise of the prime function of the Maximiser, which is impossible.
See:
https://wiki.lesswrong.com/wiki/Paperclip_maximizer