1. For prospective new members, a word of warning: don't use common names like Dennis, Simon, or Kenny if you decide to create an account. Spammers have used them all before you and gotten those names flagged in the anti-spam databases. Your account registration will be rejected because of it.
    Dismiss Notice
  2. Since it has happened MULTIPLE times now, I want to be very clear about this. You do not get to abandon an account and create a new one. You do not get to pass an account to someone else and create a new one. If you do so anyway, you will be banned for creating sockpuppets.
    Dismiss Notice
  3. If you wish to change your username, please ask via conversation to tehelgee instead of asking via my profile. I'd like to not clutter it up with such requests.
    Dismiss Notice
  4. Due to the actions of particularly persistent spammers and trolls, we will be banning disposable email addresses from today onward.
    Dismiss Notice

Alice Long Book 2 Speculation + Discussion

Discussion in 'General' started by Roark, Nov 1, 2018.

  1. Roark

    Roark Usually Lurking

    Joined:
    Jul 6, 2017
    Messages:
    462
    Likes Received:
    1,165
    So, ShaperV released a bit of news about his sequel to Perilous Waif (if you haven't read it: do) on his blog a few months ago. It's been percolating in the back of my head, along with everything else about this fantastic setting, and I've come up with a few questions I'm dying to have answered.

    Naturally, satisfying my curiosity is hardly any reason to demand the attention (let alone answers!) of the author himself, so instead I figured I'd post my questions here, where I can be appropriately ignored, blessed, or (as I'm loudly hoping will happen) joined in ignorance by likeminded fellows who are interested in speculating with me. I would have done this in the AWOYO thread, but I'm trying not to Necro.

    There will, naturally, be spoilers for the entire first book from this point on. I'm assuming you've read it.



    Item #1: The Lurkers
    Strange Loop Sleuth may be my favorite character, despite his brief appearance. Still, his very existence as well as the information he shares with Alice shape the rest of the book, and presumably lay the groundwork for the rest of the series.
    -He reveals the existence of SuperAI lurking about
    -He introduces Alice to the "Key Deliberation"
    -He suggests that Humanity approaches a period of significant instability
    -He proposes supporting a friendly (to SuperAI) nation through such a period
    -He uses the phrase, "a brighter future", a phrase that looks Mirai-ish. (yes I'm aware mirai=future)

    Which raises for me these questions:
    Would the Mirai Kingdom have qualified as a candidate for support to SLS if they were still around?
    What is the source/cause of this instability? Does it reach extinction level?
    What brighter future? Did the Mirai have something specific in mind? Does SLS?

    Item #2: The Mirai Kingdom
    Basically, their "Holdout colony" option succeeded, in a very roundabout sort of way and depending on how you define success. Alice Long escaped her native kingdom's fate, with complete access to said kingdom's technology. This also has a lot of implications for the story.
    -Alice lives under threat of discovery/ Inner core expeditionary force being sent to wipe her out
    -Alice has access to some of the highest known technology in the universe
    -Alice has strong incentive to use that technology given Item 1: The Lurkers

    Which makes me wonder:
    -What is Alice going to do with that technology?

    And despite all of this we don't really know very much about the Mirai Kingdom. Such as:
    -What was their form of government really? (An emperor and his family, but how did they maintain popular support?)
    -Who founded them? And why? (That name appears to have been around the whole time. were they always shooting for Alice?)
    -Did they know about Item#1: the Lurkers in its entirety? Is that why they pushed so hard?
    -What brighter future? how?
    -Were they actually evil?

    Item #3: ShaperV's plans
    ShaperV has clearly put a lot of thought into this setting. As with his Daniel Black books, I've been assuming that he has a general idea of where he's going with it. That assumption took a bit of a hit with the blog post: "...My plans for Alice have undergone substantial revisions, because I finally have a plot for a real sequel to Perilous Waif...".

    I mean, he clearly had plans to revise, but apparently not a plot? Although the way it's phrased, he had a plot for further down the line? I'm also nervous about where he intends to go with things in the long term though, so I wanted to speculate with some other people to see what they feel would be interesting to see happen next.

    For example:
    -Is Alice going to get involved in any empire building later?
    -Is Alice going to get dragged into the Akio mess any further?
    -Is Alice going to have to fight off Inner Core expeditionary forces?
    -Will Alice share out the supposedly-beyond-all-others Mirai mental enhancements at any point? (candidates: Dika, Kavin and SleepingDragons, Naoko and SquareDeal crew.)
    -How will Alice be involved in the Key Deliberation (something SLS thought she'd play a significant role in even before she recovered the entire Mirai tech base)?
    -Will Alice's Mother find her or vice-versa?
    -What role will Alice's Mother play in everything?

    Don't get me wrong, I'm delighted to hear that Book 2 is on track (and I love the naming scheme), but I'm worried that Alice is going to be stupid for reasons of plot (although ShaperV has a good record about such things). And beyond that I'm curious to see what shape the conflict and solutions take in the ensuing books.




    TL;DR I can't wait for Merciful Troubleshooter, and Thrall isn't for another month. So here's a bunch of questions and thoughts that probably don't matter to anyone else, but please join me in speculating.
     
    Last edited: Nov 1, 2018
    alethiophile, WyldCard4 and wasprider like this.
  2. DannyVanMeagher

    DannyVanMeagher Your first time is always over so quickly, isn't it?

    Joined:
    Sep 1, 2019
    Messages:
    1
    Likes Received:
    1
    I was searching for any release date updates on Merciful Troubleshooter and ran into this forum and thread. I'm looking forward to reading the Naruto fanfic by SV that was mentioned elsewhere here. I thought I'd respond to a couple of the questions you posed and also mention some other stuff I'd been thinking about since reading PW.

    First, I was unaware that "Mirai" means "future" in Japanese. Thanks! :) That's a good bit of trivia/possible foreshadowing to learn.

    Guessing, but from the talk of all the Mirai supersoldiers, I got the impression that the upgrade technology was widely spread throughout the Mirai Empire. While I'm sure the latest upgrades were held tightly by the elites, I also think that gives them a reason for support -- that they act as a testbed for upgrades that will eventually be spread to the offspring of the rest of the Mirai society. Also, the high automation level implies a low need for low-level people, so I doubt they had a lot of internal strife from disaffected masses. Everyone's upgraded, everyone's part of the system. The ones who can't or won't fit in probably get dropped into a volcano or something.

    Doubtful. Demonizing the opposition is standard practice, especially if the winning side thought they managed to kill them all already.

    I suspect she's really dead.

    I have to wonder if it's possible. It wasn't clear from the book that these could be added later, rather than having to be designed in from birth. I also wonder if it's even possible for her to reproduce with a non-Mirai. But she could certainly build all the really neat war machines and use the mercenaries and yakuza to cut a swath through the universe.

    Also, I'm guessing she'll start stashing "eggs" of herself around just in case someone does manage to kill her.





    Lastly, I found SV's worldbuilding discussion at the end of PW really enlightening. I wasn't sure I agreed with him about AI, and having thought about it for a few months now, I know I disagree.

    Caveat: the one giant gaping assumption I'm making in all of this is that it's possible to create an accurate model of a neuron, in analog form, in around 10^3 transistors.

    Simulation of organisms' neural systems today is slow and expensive because of limitations in our understanding of neurobiology. Even modeling the behavior of a single neuron is difficult and inaccurate, much less clusters of them. I've read that we now have a good simulation of a flatworm brain with on the order of 10^2 neurons.

    However, once we have an accurate, valid model of a neuron, and can form that model on silicon, scaling it up to lots of neurons becomes a copy-and-paste issue. CPUs currently are estimated to have around 7x10^9 transistors (Intel apparently refuses to release hard counts because trade secrets). Let's say a neuron can eventually be modeled with 10^3. That's around 10^6 neuristors (a nod to Heinlein and TMIAHM) per chip.

    Then the limitation becomes our knowledge of the connectivity map of a human brain and the initial state programming of it.

    To get that information would likely require actions that Western society would consider not merely unethical, but reprehensible and criminal. Specifically, dissection and eventually vivisection of living infant brains.

    There are two routes that go around this: (1) start with a simpler, smaller animal. I propose the African Gray Parrot. Dr. Irene Pepperberg's research has shown that they are capable of having far more complex thought processes than most people believe. (I strongly recommend her book, "Alex and Me". She discusses some of the criticism of her work in it, and my impression, perhaps biased by her view of it, is that most of the criticism boils down to "she can't be trusted because she was too personally involved.") Using these or another bird capable of speech and understanding, we can learn enough to bootstrap ourselves to the next stage.

    And the other route is: (2) China. Our AI future will probably be created on the graves of millions of dissidents, or more specifically, on the graves of millions of babies harvested from dissidents just as the Chinese government currently harvests transplant organs from political prisoners to sell to people wanting a healthy young matching kidney/liver/heart. Like it or not, the Chinese government won't care.

    A parrot has around 1.5x10^9 neurons. At 10^6 neuristors per chip, that's 1500 chips. Anyone who's been in computers for a few decades remembers the old memory boards for Apple ][ computers with 64 chips on a single board; it just isn't that much space, or heat. Current supercomputer clusters surpass that many total boards easily -- Wikipedia says SUMMIT has over 9000 CPUs and 27000 GPUs, so around 10^4.5 processors, all of which are arguably more complicated than an analog neuron model needs to be. But by then, we may be able to pack more transistors into smaller spaces, or go 3D with multiple layers.

    Once accurate maps are created, connected in silicon, and programmed, I suspect that even a parrot brain will be far more intelligent than we'll need to work with for the foreseeable future of non-upgraded humanity. Much of intelligence is nothing more than speed and memory. We can add more memory to the AI quite trivially, and speed of silicon is thousands of times the speed of neuron electrochemical propagation. Parrots can engage in general planning already, and even if more limited than humans, the added memory and speed should make even such a limited AI superior to most or all humans.
     
    Roark likes this.
  3. Roark

    Roark Usually Lurking

    Joined:
    Jul 6, 2017
    Messages:
    462
    Likes Received:
    1,165
    Ha. I totally forgot about this thread since no one ever responded. I found ShaperV through his Naruto fanfiction actually. "Time Braid" is complete and excellent, but the other one ("Indomitable", I think?) is not finished. Still a really good read, it just stops in a rather unsatisfying place.

    I felt like an idiot the first time I made the connection, because "a brighter future" just makes Alice's origin blatantly obvious, but it's good to see that this isn't as widely known as I'd thought. Not that the success of the book hinges on it or anything.

    The sad thing about this is that I intended all of the questions to serve as conversation starters, but never got any response. I have some pretty definite thoughts on what the answers are, but no one to share them with! So thanks!

    The Mirai definitely struck me as a people with a strong philosophy that promoted taking care of everyone. I imagine they were a lot like Alice writ large: Hyperprotective once you're family (read: nation), and utterly ruthless with anyone/thing that threatens family. Although you raise a good point about the internal dissidents. I wonder if they actually executed people for refusing to get with the program, or if they resorted to exile/lesser measures.

    ShaperV does seem to indicate this in PW, but I was thinking we should be careful not to swing the pendulum too far in the other direction. From what little we've seen of the Mirai they sound like a nation of philosopher-kings, but we don't actually have that much data beyond Alice's programming and the AI Hope.

    Really? I can see it going either way, but she would be a great tool to fit in a ton of different narratives, and it doesn't stretch my disbelief at all that she's survived and looking for Alice.


    Well the book states that they were still working on the process to transition an adult to the same level of upgrades as Alice when the fleet left Mirai, so it seems likely it would have eventually been possible, and more importantly, I think several of the "supersoldier" packages the Mirai had before Alice would still count as cutting edge stuff even in the Inner Sphere. So even if she can't share out her package, she's still got tons of goodies to hand out if she ever feels so inclined.

    Because the upgrades apparently got rolled out to the general population, so it seems unlikely that they didn't get the process down to "This machine here can handle it".

    As for the reproduction, given that the book indicates "wanting to have children like you requires a fabricator there" I think she can have kids just like her. The process is probably entirely taken care of on her side, with just a little DNA donation required. (Come to think of it, the eggs are basically the same thing with a downloaded brain copy and no mixture of DNA.)

    I really enjoyed those sections, but what stood out to me was how ShaperV went about the process. He made some assumptions, and then thought out the implications and extrapolated. So you're probably correct, and a workable AI won't be as impossible to achieve as PW presupposes. Hell, it might not even be as difficult as you explained just here! But for the case of the narrative of the Alice Long series, it makes sense.

    Especially because I think ShaperV wanted to write a story that still allowed humans (however changed by cybernetics/transhumanism) to have an impact on the world around them. A future shaped by your AI theory would rapidly outstrip human agency, and then you have to ask what kind of story can you tell that would even be recognizable to us mere mortals?

    I would be more worried about the future of our AI overlords, but I think counting us humans out is a fool's bet. Whether you believe in a god or evolution, humans are the only thing we know of that consistently come to dominate their surroundings, through the sheer attrition of time and effort. Maybe some AI project will get out of hand and go all Skynet on us, but I doubt it would be so easy as some people think.
     
    WyldCard4 likes this.
  4. NervousEnergy

    NervousEnergy Your first time is always over so quickly, isn't it?

    Joined:
    Sep 8, 2019
    Messages:
    2
    Likes Received:
    4
    I'm surprised there's not more discussion / speculation about Alice Long 2, especially given how popular the first book was in terms of sales. It was one of the best classic space opera stories I've read in a long time, and I'm very much looking forward to the sequel.

    From an AI standpoint, we will obviously learn a lot about terrestrial human/animal awareness and sentience from studying neurons, but the neuron is biological adaptation that doesn't translate well or efficiently into silicon (or future quantum entanglement computers.) I don't think we're going to get true AI by modeling neurons. Self-awareness / sentience is one of those concepts that currently lives at the intersection of science, philosophy, and mysticism. I personally suspect (as do many other Humanists) that sentience is a natural by-product of a certain degree of sophistication in an information storage and processing system. If that's the case, then true self-aware AI will arise as a consequence of expert systems getting as sophisticated or more than the example of the relatively slow but *massively* parallel processing human brain. They'll also be very, very alien to us. I really like how the AIs in the Alice Long universe are pretty much incomprehensible to the humans who built them.

    Can't wait for the sequel. The world is incredibly well-built, with a lot of mystery and variety, but the soul of the book is the character Alice herself, and she's a masterpiece.
     
    Roark and wasprider like this.
  5. Roark

    Roark Usually Lurking

    Joined:
    Jul 6, 2017
    Messages:
    462
    Likes Received:
    1,165
    I suspect we may be in the wrong place for this. I'll bet there is some discussion ongoing somewhere (like a reddit), but QQ isn't exactly where people first go to discuss space operas, especially if they're part of the wider audience who are unfamiliar with ShaperV's "A World of Your Own" quest hosted here.
     
  6. Xagroth

    Xagroth Your first time is always over so quickly, isn't it?

    Joined:
    Oct 3, 2019
    Messages:
    2
    Likes Received:
    1
    Hi! I'm in the same boat here, but with a little different point of view (since other works like the Takeshi Kovacs trilogy is something I read years ago, and I've been an avid fan of the Eclipse Phase RPG for years -not a player since the game seems to be unable to gain traction where I live, language being one of the reasons-, plus some other works).

    And yeah, I landed here too while looking for news on book #2 (last ones being from february)

    First, my impression of the book: Alice Long is a Mary Sue, but of the "good type", at least for now: while she is more than capable to get out from almost any situation, and certainly suffers little to no negative consequences from her lack of discretion after leaving the "opressing home", most of the time the book is less about her and more about worldbuilding, and she gets badly battered at more than one point, in a way that reminds me a lot of Honor Harrington (from David Weber's, and a recognized Mary Sue that works, again because she is less "the master of the world" and more "a plot engine". Certainly, Alice will start something big.

    Second, my reservations. While the book is certainly enjoyable, I fear that following books might suffer from an overpowered protagonist (as seen onboard Hope), or a Dragon Ball -esque escalation of enemies. Another option that worries me even more is a genre shift, from space opera to military sci-fi or empire building, something that requires a good writer and a certain frame of mind from the reader that I, personally, never had to shift from one genre to that one (having read almost all of the books that are about mil!SciFi as such from the very beginning).

    Now, onto what you asked:

    Cerrtainly the Lurkers bring a great deal of definition to the setting, because they can be assumed to be self-improving artificial intelligences. For any of the functioning ones to be unaware that those who do nothing are at the mercy of those who do anything. This means they have been shaping the galaxy for a future in which a meeting of fates between ASIs and "humans" is inevitable.

    The Mirai Kingdom... most likely WAS one of a number of measures taken for such a time. Upgrading humanity, or at least hoping for an "ambassador" state, would be a logical first step in the road that does not start with total war. And total war at this point in the setting would mean a ruined galaxy.

    The source/cause of this instability is, without much doubt, humanity with or without the Lurker's intervention. Remember: even knowing what happens when creating a superintelligence, groups still attempt it... How log until someone does something that makes him/them a threat to all sentient life? Call it a starkiller, or Soren in Star Trek Generations...

    The brigther future would involve, looking at SLS and Alice, full Posthumanity. Alice, when interacting with SLS, gets defined as Posthuman instead of merely Transhuman (at which level most of the spacers would be), because it is not possible to recognize her as human; not to mention, she herself thinks her bioparts are just a disguise.

    I find hard to believe only Hope escaped. Space is really big, even with the Swarmlords cranking their Von Neumann drones up to eleven, more so when you consider several layers of reality, stealth tech, and simply sending nanotech seeds: Alice can regrow from nearly total destruction, by Hope's words. Mere stadistic means there was more survivors, even if they had only a body and basic gear or even a basic ship at the most.

    Government: frankly, I think it has not been defined because it's more of a "4x simulation game's government" than a real one, before options like democracy or tiranny were coded. The Mirai obey their "player", simply that... and, as I said, a Lurker or a group of them could be "the player". Development's speed is quite hard to believe otherwise, and it certainly explains the viciousness showed by the Swarmlords.
    Of course, the SWLs could be the weapon of choice of another Lurker waiting for the best moment to erradicate humanity, so we get the standard hopeless battle with great sacrifice from the part of Lurkers believing in convivence with humanity...

    As for Alice and the Mirai Tech... she does NOT want to flaunt it. Few know about it... but those who do, she doesn't know, and can see her and give warning to the SWLs. So "Depowering" Alice in the 2nd book is logical, at least in the least covert gear.

    The rest of the questions here... I think I answered already ;)

    As I mentioned, at this point Alice is really overpowered, and an overpowered Mary Sue is boring. As the old Role-Playing adage says, we want our heroes to be like John McLane in Die Hard: battered, bloodied, exhausted, but ultimately victorious. So I would think that he managed to revise the book in a way that it's deserving of more than one read, like Perilous Waif, instead of some sort of interlude you suffer before the grand finale in a final, third book (because of trilogies); commercially, he might have managed to expand the collection from three to more books, which is always good in that sector.

    Now, if we go with the usual foreshadowing, we have the kitsune character Alice met while surrounded by Inugami and three tibdits of information quite interesting...
    1. Her missing mother
    2. SLS's mentioned "untrustworthy other" hat would speak to Alice
    3. Alice's backup system (self-cloning)
    So, is the misterious character behind the holoprojected kitsune her mother (and amused by her, but not wanting for her be tied to the Yakuza) AND/or the "untrustworthy other"? Or, more twisted, is said character... Alice? Being the one we followed in Perilous Waif a memory-less (or memory-locked) clone/backup? After all, the physical Backup system she has is quite useless in a way, yet she can split copies of herself as infomorphs to control a bunch of bots during the final fight... So it would make sense to deploy several bodies waiting to receive one of said infomorphs to update... or replace... the memories of said backup.

    It would be darker than what be have seen until now... but that's a part that I miss in Perilous Waif: real darkness, not simply... buffoon villains, besides a Jaffar (Akio's uncle).
     
    wasprider likes this.
  7. NervousEnergy

    NervousEnergy Your first time is always over so quickly, isn't it?

    Joined:
    Sep 8, 2019
    Messages:
    2
    Likes Received:
    4
    Well, if the book(s) are being discussed elsewhere, it's not being picked up on Google. This thread is still one of the highest results for 'Merciful Troubleshooter'.

    I hear this a lot, and I'm not sure I can agree with it. Power fantasies need a powerful protagonist, and this book is clearly in that genre. Alice may be overpowered, but not compared to what's going to be aligned against her. When it's one person against the majority of the rest of the universe, that one person better be a powerhouse, or the story threatens to go off a much worse (IMHO) cliff: the underpowered protagonist who wins through deus ex machina.
     
    alethiophile and wasprider like this.
  8. Elbrasch

    Elbrasch Getting sticky.

    Joined:
    Apr 21, 2014
    Messages:
    71
    Likes Received:
    71
    Agreed, a shining curbstomping supersoldier does not bad scifi make, see Halo 1 and the accompanying book.
    I don't expect us to get a "younger self/older self" conflict in the next book, that would be to much of a carbon copy of Woken Furies (Heh, even if the twist that she is the backup would be nice).
    Going from memory here, through I think there was never a strong motivation for her other than survival/curiosity/finding a better life. That one might be continued by sending spec ops teams from the inner sphere after her, attracted by rumors of the battleship being found.
     
  9. ShaperV

    ShaperV Experienced.

    Joined:
    May 17, 2013
    Messages:
    2,510
    Likes Received:
    38,765
    I think people who complain about this tend to overlook the fact that what matters is a character's power level relative to their enemies, not some absolute power ranking. In RL an organized crime boss can field a few dozen thugs armed with handguns, so a crimefighter's abilities can be scaled around the needs of a street fight. Alice lives in a setting where that crime boss can field a frigate with enough firepower to sterilize a planet, along with a couple of companies of warbots and marines in powered armor. So she needs to be a heck of a lot tougher than Batman, or else her victories won't seem believable.

    But part of the problem is probably that the first book is an introduction to the setting, and we don't see much of the groups that could squash Alice like a bug. The second book shows a lot more about the setting, and also has some scenes where we see other characters beating Alice in areas where they're more talented than she is. Hopefully that will address the complaints of my more thoughtful readers, although I'm sure there will still be some who complain about it.
     
  10. alethiophile

    alethiophile Shadowed Philosopher Administrator

    Joined:
    Apr 26, 2013
    Messages:
    4,075
    Likes Received:
    18,010
    In an RL context, I think this assumption is pretty unlikely to hold. The idea that a neuron is in fact well-represented by one of the elements in what we call a "neural network" is much more assumption than proven fact; given how incredibly complicated cells are, and how evolved biology doesn't feel any need to enforce hygiene across abstraction layers, it seems likely that each neuron is itself an enormously complicated molecular computation engine, not amenable to simulation as a simple element. They may even leverage quantum phenomena, as we already know some of the senses do; if that's the case, they may not be classically simulable at all.
    I think a lot of difficulty is being glossed over here. "Self-improvement" is not some magical fairy dust that automatically yields exponentially improving power; it's a difficult engineering problem, and in this universe we know that trying to create a mind that's all that much smarter than you -- even if it's meant to be a close copy of you -- doesn't end well.

    In fact, we already know that the Lurkers, or at least SLS, believe that humans will beat them in terms of AI tech over time; this is why they see humans as an existential threat.
     
  11. Ina_Meishou

    Ina_Meishou Making the rounds.

    Joined:
    Jun 5, 2013
    Messages:
    49
    Likes Received:
    1,735
    As I recall it's less that the Lurkers believe humans will eventually outclass them in AI tech as a whole, and more that eventually humans will be able to reliably recreate intelligent beings on the Lurker's level. Which given the scale of human production capacities would lead the the Lurkers losing due to pure numeric concerns.
     
  12. Xagroth

    Xagroth Your first time is always over so quickly, isn't it?

    Joined:
    Oct 3, 2019
    Messages:
    2
    Likes Received:
    1
    There doesn't need to be a conflict, and her "older self" might just be playing the "patron goddess" role we see on the Daniel Black books (I confess I read all that collection between my last post and this one. Fortunately, regardless of anything else, William's books are easy and fast to read).

    My problem with going all out in the first book is simply that things need to grow, or else there is no conflict nor grow, just a stream of enemies that can't truly press the protagonist, thus becoming bad Mary Sue territory. And starting high in the scale of growing conflicts means escalate up to 11 regularly, which for me ends being boring... The most common example is Dragon Ball, if it were to start with Goku being able to reach Saiyan from 3... Or in D&D/Pathfinder, getting to lv20 in the first adventure.

    Mind you, seeing how the 4th book in the Daniel Black series (Thrall) goes on, I can guess that Alice will enter into an escalation of gear for herself and her retainers (currently just Emla... Quite funny how there is an Embla in books 2 or 3 of Daniel Black xD).

    As I said before, the problem is when instead of starting with a few low level mooks and see the main character growing in the first book to take on a weak supervillain, we see the climatic Batman Vs Joker in the first half of the book, and then Bane gets beaten in the final half, what's left for book 2? Superman? Darkseid? What worries me is that in the Daniel Black collection has more in common with a classical D&D 3.5/Pathfinder character with the Leadership feat (and a pool of possible main reatiners to take into adventures, using first Cerise then Alanna) with industrial scale magitech, and that is not an advantage in a SciFi setting even considering she has better tech and Hope's libraries (and a good excuse to make very cautious use of them). Thing is, all the book 1 points to a military conflict, and that is not something the author has much experience with (dipping his toes with the Daniel Black collection, but with a protagonist that prefers to work as a hero and avoid being involved into mass combat unless it's from afar as magical artillery).

    There are, essentially, two improvement fields for an AI. First is hardware, and here we can agree improvements are very straightforward (energy generation, heat sink, more efficient circuits, etc...). The second is the software, and we see in the first book how long it took to SLS to reverse engineer and recompile Emla, so very small changes seem possible without entering into some sort of doubling capacity every month or so. I would bet that SLS can improve faster than even Alice, and has been doing so for a long time.



    Another set of worries for the Lurkers is, of course, that humanity creates some super-super AI that starts wiping humanity out, makes humanity hunt for it, discover lurkers, asume the lurkers are allies, backups or duplicates from the original genocidal AI, and hunt the lurkers completely.
     
  13. alethiophile

    alethiophile Shadowed Philosopher Administrator

    Joined:
    Apr 26, 2013
    Messages:
    4,075
    Likes Received:
    18,010
    Sure, but this is basically the same thing. Your progress in AI (or any other) tech will be determined by how many people you can put on it times how smart those people are, modified for resources available maybe. If humans can eventually mass-produce loyal and well-functioning AIs as smart as the Lurkers, then from that point on they'll always be able to outclass the Lurkers in speed of development.

    Emla is 1. much less sophisticated a mind than SLS is, and 2. based on a well-known mind design that was only superficially modified by the imposition of her clumsy control method. The fact that SLS can quickly copy and recompile Emla doesn't at all mean that he can make any notable improvements to himself easily or reliably.
     
  14. Elbrasch

    Elbrasch Getting sticky.

    Joined:
    Apr 21, 2014
    Messages:
    71
    Likes Received:
    71
    Also, there is always the problem with motivation. Sure, I could get a second Master and a PHD, not even talking about hypothetical brain surgery, through I would need more motivation for it. And as a lot of Humanity show (myself included), a somewhat decreased mortality rate in X years is not enough.
    We might have more biases than an AI, through the cost/benefit analysis of self improvement doesn't fundamentally change.