Vox Vitae: Warhammer AI quest

[Haganeko]

Seems like a good reminder that we need to be flexible about longer term planning.

Personally I see no need to throw bongo into a sun when destructive analysis is right there. We should at least get some benefits from ending it.

Side question- I've seen people say that binding it means we get the vault space back, or that we will be done with it after that? My impression was that we'd still have to carry it around, it would just be bound and caged instead of just caged. (And my bet is still that we will never run out of research options on Bongo, probably priced just cheap enough to be enticing, for as long as we keep it.)

Click to shrink...

Valid. And I think the idea is both that a lesser vault can hold a bound demon and that Demonology might lead to some tech allowing permanent disposal.

We have a lot of other factors to consider... But you make a good point. Lets see about researching the other stealth techs and then the druhkari craft before we go for sensors, though? They might discount sensors, they're cheap, and they'll probably end up as prereqs for the next level of sensors anyway.

Click to shrink...

Exactly. Stealth tech may also help avoid us being noticed by trouble we don't want.

Bongo physically should probably fit in a small portion of a single-space chamber.

...We might want a structure that's a set of individually shielded compartments?

Or, I mean, we could bind Bongo as an experiment and then maroon it someplace inaccessible, use it as a disposable Cia target dummy, whatever. IMO owning Bongo isn't really the point of researching daemonology, researching daemonology is the point of owning Bongo.

Click to shrink...

Very true. Also, part of the point of owning Bongo is it's potentially more more dangerous out there than in the Vault.

Actually, here, let me put together some lists of stuff I think we want:

[Snip]

Total cost here is 1450, for 18 techs. But I think we can cut that down by a *lot*, if we prioritize well. My preliminary selection would be:

-[] Psychic tripwires (25 RP)
-[] Faith is my shield? (75 RP)
-[] Machine Spirit Chaos resistance (100 RP)
-[] Machine Spirit-controlled Psychic Shields (50 RP)
-[] Psytech Machine spirits (100 RP)
-[] Drugs? Drugs. (75 RP)
-[] Improved Passive Stealth (50 RP)
-[] Basic Ground force stealth (50 RP)
-[] Basic Active Stealth (75 RP)

For a total of 600 RP for 9 techs, with 22 already applied, before any discounts applied this turn, which is really not bad at all. That should get us a good chunk of progress on boosting Cia, protecting against the warp, discounting sensors, progressing stealth, progressing Machine Spirits, understanding the warp and hopefully discounting the lead in techs for warp sensors and warp comms, and progressing biology.

Click to shrink...

I think auomated manufacturing should also go in the "want but wait for Denva" list. It plus stealth means we can drop some void installations somewhere with instructions to make us stuff and just come back for the product later.

Also, we have those Eldar biology samples, so that might go somewhere amusing.

And maybe it'd toss a discount to Juvenat tech.

Plus, being able to bombard humans with safe knockout gas would probably come in handy sooner or later.

Click to shrink...

Makes sense! We have a bunch of juicy, but expensive, medical tech available and they might help with quite a few.

Exactly so, yeah. My question there is in vitro - if we have enough understanding to grab some eldar cells and try cultivating them in petri dishes just to see what they do, does that mean we learn more from dissecting them? Or does it not really matter?

Click to shrink...

Makes sense. The question is: would this research exhaust the sample? If not, the narrative of the research might point us to when we might want to do it again.

 

[Angle]

I do not have a good handle on the 'you don't have research A to research B, but you should' dynamic. It worries me a bit.

Click to shrink...

My personal rule is mainly 'cheap stuff that looks like it might give a discount, or might provide tools that would be useful'. Or if B straight up says 'discounts all <Stuff>', then researching cheap <stuff> beforehand to unlock later, more expensive <Stuff>. Which is the case with the Druhkari craft, and the Immaterium Understanding techs.

I think auomated manufacturing should also go in the "want but wait for Denva" list. It plus stealth means we can drop some void installations somewhere with instructions to make us stuff and just come back for the product later.

Click to shrink...

Those are *expensive*. If we want those, then I'd look at getting the 100 point machine spirit manufacturing techs first, and hoping they give discounts. Though as you said, waiting to get back to Denva and see what they've worked out is sensible too.

 

[SightUnseen]

Actually, here, let me put together some lists of stuff I think we want:

Click to shrink...

Missing AI research techs and the cognition implants, I think. I know you're fishing cheap stuff for bonuses, but Neablis has said "do not game the system" and at some point we'll need to go for the pricey stuff, or be built broad but too shallow.

250 for the cognition filter, that's the fastest item and well worth IMO.

Edit: AI stuff refresher. Need both for checksum/rollback protection.

-[] Intelligence Coding (400 RP) You don't really understand how your own code and hardware works. Maybe putting some effort into understanding that would pay dividends. It's never been something too interesting, but hey, maybe you should give it a shot. (Helps you understand your own code. Prerequisite to building new AIs, starts command point tree and automation tree, also unlocks personality-checking research)

- [] Personality-Checking Routines (200 RP) All of this chaos-corruption is making you nervous. People changing according to the will of some warp-entity? Becoming different in ways that would disgust their past selves? That gives you the heebee-jeebies. Maybe - just maybe - it would be good to develop some ways to automatically check if aspects of your personality are diverging ways you don't want them to. Then you might be able to fix the problem before you become somebody who doesn't want (Gives you some algorithms to check for personality divergence in AIs. Doesn't do anything, just alerts you to changes and lets you decide what steps to take. Unlocks further research to prevent/rollback personality change, as well as potentially extend this research to humans with neural implants). Locked behind Intelligence Coding.

Click to shrink...

 

[Angle]

Missing AI research techs and the cognition implants, I think. I know you're fishing cheap stuff for bonuses, but Neablis has said "do not game the system" and at some point we'll need to go for the pricey stuff, or be built broad but too shallow.

250 for the cognition filter, that's the fastest item and well worth IMO.

Edit: AI stuff refresher. Need both for checksum/rollback protection.

Click to shrink...

That stuff is *waaaaay* too expensive right now, and it gives us less resistance to chaos than just getting the psy shield and psytech stuff I picked out. And there's a wide margin between gaming the system and simply researching things in an effective order.

 

[LightLan]

You know what's the worst the daemon could do this turn?

Simply keep charging up from the warp-exposure and not expend it until it has enough to break through at once.

 

[SightUnseen]

That stuff is *waaaaay* too expensive right now, and it gives us less resistance to chaos than just getting the psy shield and psytech stuff I picked out. And there's a wide margin between gaming the system and simply researching things in an effective order.

Click to shrink...

Absolutely there's a difference - I just mean that there is likely a limit to the discounts we'll see on some of the expensive items.

I'm not suggesting a triple-research for three techs - I'm suggesting that we do the psy-shield but *also* cognition filters so Anexa can do Immaterium Understanding safely, and/or maybe put 100-200xp down on Intel Coding.

 

[Aineko]

I think there's a case for Intelligence Coding. I don't think we can count on finding a lot of discounts, and refusing to ever buy expensive techs may be an error.

I'm not as interested in personality-checking specifically as other things behind the entry gate, though once we were over the Intelligence Coding hurdle it would probably become more attractive.


Intelligence Coding isn't only for the Chaos paranoia. It's the first step into a huge field.

 

[SightUnseen]

personality-checking

Click to shrink...

Can't do Personality-checking at all until we get Intelligence Coding anyway - but this is a way to get a Corruption Meter, which seems important (but not necessarily Urgent) to me since we biffed on scrapcode immunity (at least for now).

Intelligence Coding isn't only for the Chaos paranoia. It's the first step into a huge field.

Click to shrink...

Absolutely. This is how we set up a remote base of autofactories, for example.

 

[sciencegeek]

Considering how much seems to be bottlenecked by research. Do we have any way to increase how much we can do every turn beyond the Tech priest's? Getting some research from them was a real boon but hopefully either Human Design Interfaces (200) or Companion Cogitators (450 RP) might be able to further boost the amount of research per turn. Although for the latter a bit of waiting might be advisable.

 

[Angle]

Absolutely there's a difference - I just mean that there is likely a limit to the discounts we'll see on some of the expensive items.

I'm not suggesting a triple-research for three techs - I'm suggesting that we do the psy-shield but *also* cognition filters so Anexa can do Immaterium Understanding safely, and/or maybe put 100-200xp down on Intel Coding.

Click to shrink...

I don't think cognition filters makes that much difference, honestly, especially if we're just gonna keep her under psy shields all the time. I'd much rather focus on things that offer more direct resistance to chaos, we can go for cognition filters later, once they get cheaper.

I think there's a case for Intelligence Coding. I don't think we can count on finding a lot of discounts, and refusing to ever buy expensive techs may be an error.

I'm not as interested in personality-checking specifically as other things behind the entry gate, though once we were over the Intelligence Coding hurdle it would probably become more attractive.


Intelligence Coding isn't only for the Chaos paranoia. It's the first step into a huge field.

Click to shrink...

Neablis straight up said that Large Scale Machine Spirits and Advanced Materials could discount it on crits. I'd be willing to add LSMS to the research agenda, and we're already hoping for discounts to AM from the druhkari craft and from superconductors, or from other stuff we find while exploring. And remember, the main thing that IC unlocks for us is even more expensive techs to research, so I think we should focus our efforts elsewhere for now.

Considering how much seems to be bottlenecked by research. Do we have any way to increase how much we can do every turn beyond the Tech priest's? Getting some research from them was a real boon but hopefully either Human Design Interfaces (200) or Companion Cogitators (450 RP) might be able to further boost the amount of research per turn. Although for the latter a bit of waiting might be advisable.

Click to shrink...

We're hoping to unlock something behind MS Design. And going back to Denva for more tech priests is an option, we can get at least 8 RP that way, and I'm hoping another 50 Tech priests could get us 10-25 RP beyond that.

I really want to try and focus our efforts. I think our primary goals right now should be:

• Finish exploring Ascalons neighbors and then return to Denva.
• Develop Cia and support her properly.
• Further protect ourselves against warp threats.
• Wrap Bongo up for long term handling.
• Further our understanding of the warp, and improve our ability to make use of it.
• Develop Warp comms.
• Develop machine spirits as a cheap way to boost the effectiveness of, like, all our tech.
• Develop better sensors and get a handle on stealth tech, so the Druhkari are less of a problem in the future.

Luckily, the middle six all go together really well, and the first and last two aren't too expensive. Other stuff should mostly wait, with some minor investment here and there where appropriate, such as with Drugs.

 

[SightUnseen]

I don't think cognition filters makes that much difference, honestly,

Click to shrink...

Intense Corruption arc flashbacks

I understand, and disagree. This stuff gets researched all the time in-setting - and feel like that usually is seen resulting in an Icarus moment. AdMech are the literal poster children for poking things that ought not be poked and losing their shit over it, so I think the caution is warranted.

A psyshield may not protect from a cognitohazard - and it certainly won't protect her the next time she leaves the ship to visit a planetside lab or anomaly.

 

[Aineko]

Neablis straight up said that Large Scale Machine Spirits and Advanced Materials could discount it on crits. I'd be willing to add LSMS to the research agenda, and we're already hoping for discounts to AM from the druhkari craft and from superconductors, or from other stuff we find while exploring. And remember, the main thing that IC unlocks for us is even more expensive techs to research, so I think we should focus our efforts elsewhere for now.

Click to shrink...

...Could discount it on crits is not promising a lot. Not that I don't like LSMS, we'll want it by the time we're looking at building a fleet at least...

Yes, IC unlocks more techs. Not necessarily all expensive techs - personality checking is significantly cheaper than Intelligence Coding, so other techs behind it might be as well.

Absolutely. This is how we set up a remote base of autofactories, for example.

Click to shrink...

We should be able to bypass it for that purpose thanks to having gotten access to Basic Automated Manufactories through the Machine Spirit tree.

 

[Angle]

Intense Corruption arc flashbacks

I understand, and disagree. This stuff gets researched all the time in-setting - and feel like that usually is seen resulting in an Icarus moment. AdMech are the literal poster children for poking things that ought not be poked and losing their shit over it, so I think the caution is warranted.

A psyshield may not protect from a cognitohazard - and it certainly won't protect her the next time she leaves the ship to visit a planetside lab or anomaly.

Click to shrink...

I don't think it's useless. But worth 250 RP? I'm not seeing it, sorry. We could get Tripwires, Faith, MS CR, and MS CPS for that price, which seems *waaaaay* better.

Still, once we get back to Denva we'll probably get some tech unlocks and discounts from their cybernetics research - it's not what we asked for, but they almost certainly did a whole bunch of it anyway cause it'd be so useful for them. Then we can start clearing out low hanging cybernetics techs, and I think the price will come down substantially.

...Could discount it on crits is not promising a lot. Not that I don't like LSMS, we'll want it by the time we're looking at building a fleet at least...

Yes, IC unlocks more techs. Not necessarily all expensive techs - personality checking is significantly cheaper than Intelligence Coding, so other techs behind it might be as well.

We should be able to bypass it for that purpose thanks to having gotten access to Basic Automated Manufactories through the Machine Spirit tree.

Click to shrink...

PC is still 200 points, which is pretty expensive compared to a lot of the other options we have available. I'd definitely prefer to focus on MS for the foreseeable future, they seem way more promising compared to what they cost.

 

[SightUnseen]

I really want to try and focus our efforts. I think our primary goals right now should be:

Click to shrink...

I broadly agree with this, even if I differ in details.

We should be able to bypass it for that purpose thanks to having gotten access to Basic Automated Manufactories through the Machine Spirit tree.

Click to shrink...

Excellent point - and MS tech in general is strong and has overlap with IC.

Again, I'm not saying to ignore MS - just that investing some points in the longer term stuff is going to be important - and I think having a Corruption gauge is more important in the short term than growing the bean.

A Navigator is a poison pill if we can't tell how badly warp travel is affecting us.

 

[LightLan]

Considering how much seems to be bottlenecked by research. Do we have any way to increase how much we can do every turn beyond the Tech priest's? Getting some research from them was a real boon but hopefully either Human Design Interfaces (200) or Companion Cogitators (450 RP) might be able to further boost the amount of research per turn. Although for the latter a bit of waiting might be advisable.

Click to shrink...

RP is hard to acquire our options are:

1) increases:
a) Machine Spirit Design, and potential follow up researches. Begun this vote.
b) Human Design Interfaces, and potential follow up researches. To be done.
c) Level up Anexa.
2) discounts
a) Streamlined ship design, 50% RP discount for ship designs (we will be designing plenty of ships for our upcoming new fleet)
b) Machine Spirit Design, and potential follow up researches. Begun this vote.
c) Human Design Interfaces, and potential follow up researches. To be done.

As you see, that's very little.
Presumable once we get the AI-tech tree down to AIs those may also do research.
They might even do research we want them to do if they like us or we raise them right.

 

[Angle]

Excellent point - and MS tech in general is strong and has overlap with IC.

Again, I'm not saying to ignore MS - just that investing some points in the longer term stuff is going to be important - and I think having a Corruption gauge is more important in the short term than growing the bean.

A Navigator is a poison pill if we can't tell how badly warp travel is affecting us.

Click to shrink...

This I'll agree with! I'm willing to prioritize cybernetics research over the navigator, yeah - we need a bunch more warp understanding before we unpack them anyway, in addition to biology.

RP is hard to acquire our options are:

1) increases:
a) Machine Spirit Design, and potential follow up researches. Begun this vote.
b) Human Design Interfaces, and potential follow up researches. To be done.
c) Level up Anexa.
2) discounts
a) Streamlined ship design, 50% RP discount for ship designs (we will be designing plenty of ships for our upcoming new fleet)
b) Machine Spirit Design, and potential follow up researches. Begun this vote.
c) Human Design Interfaces, and potential follow up researches. To be done.

As you see, that's very little.
Presumable once we get the AI-tech tree down to AIs those may also do research.
They might even do research we want them to do if they like us or we raise them right.

Click to shrink...

Once we have Warp comms, we can also pursue a research agreement with Denva and get the benefits even while travelling, which would be pretty good.

 

[SightUnseen]

I don't think it's useless. But worth 250 RP? I'm not seeing it, sorry. We could get Tripwires, Faith, MS CR, and MS CPS for that price, which seems *waaaaay* better.

Click to shrink...

Yes. Like I said, I'm not saying not to research those things - in fact, I think any or all of Tripwires, MS CR and MS CPS might discount the cognition filter.

I think it'd be worth, if we were to do a 3x research action, to get your entire 450pt psy-shield list, Improved Gellar, overflow into Cog Filter, excess overflow into ... something else. LS MS, maybe, or one of the cramming techs?

 

[Haganeko]

Missing AI research techs and the cognition implants, I think. I know you're fishing cheap stuff for bonuses, but Neablis has said "do not game the system" and at some point we'll need to go for the pricey stuff, or be built broad but too shallow.

250 for the cognition filter, that's the fastest item and well worth IMO.

Edit: AI stuff refresher. Need both for checksum/rollback protection.

Click to shrink...

As has been pointed out, we can get some discounts from Advanced Materials, which itself we can probably discount by making contact again with the natives of Denva Primus and bargaining for some ore.

I think there's a case for Intelligence Coding. I don't think we can count on finding a lot of discounts, and refusing to ever buy expensive techs may be an error.

I'm not as interested in personality-checking specifically as other things behind the entry gate, though once we were over the Intelligence Coding hurdle it would probably become more attractive.

Intelligence Coding isn't only for the Chaos paranoia. It's the first step into a huge field.

Click to shrink...

That's true! But it's also a large commitment and there are lower hanging fruits.

Intense Corruption arc flashbacks

I understand, and disagree. This stuff gets researched all the time in-setting - and feel like that usually is seen resulting in an Icarus moment. AdMech are the literal poster children for poking things that ought not be poked and losing their shit over it, so I think the caution is warranted.

A psyshield may not protect from a cognitohazard - and it certainly won't protect her the next time she leaves the ship to visit a planetside lab or anomaly.

Click to shrink...

You are correct re: heading out planet side, unless we get personal shields. That said, however, we do know psyshields work somewhat against cognitohazards—they protected Vita until she put the knowledge away—the issue is not having the cognitive filter would mean dealing damage to the shields every time the issue is thought about; in fact, even with the filter thinking about the stuff is dangerous. Having the filter just means one doesn't have to be under constant threat all the time.

 

[Angle]

Yes. Like I said, I'm not saying not to research those things - in fact, I think any or all of Tripwires, MS CR and MS CPS might discount the cognition filter.

I think it'd be worth, if we were to do a 3x research action, to get your entire 450pt psy-shield list, Improved Gellar, overflow into Cog Filter, excess overflow into ... something else. LS MS, maybe, or one of the cramming techs?

Click to shrink...

I'd still hold off on cog filter... I'd be down for LS MS and I. Gellar, though. Maybe some other Psy Shield techs, if any come down in price? Or any other stuff that looks like it leads to discounts for the warp comms stuff. Or Psitech amplifiers, that ones pretty cheap right now and I bet it'll further a bunch of stuff.

As has been pointed out, we can get some discounts from Advanced Materials, which itself we can probably discount by making contact again with the natives of Denva Primus and bargaining for some ore.

Click to shrink...

We already got the stuff - and not just ore, refined metal. It discounted superconductors, not AM. Though superconductors itself might discount AM. Should wait till we get back to Denva though, and see if they figured out how to make them, though.

 

[LightLan]

Yes. Like I said, I'm not saying not to research those things - in fact, I think any or all of Tripwires, MS CR and MS CPS might discount the cognition filter.

I think it'd be worth, if we were to do a 3x research action, to get your entire 450pt psy-shield list, Improved Gellar, overflow into Cog Filter, excess overflow into ... something else. LS MS, maybe, or one of the cramming techs?

Click to shrink...

If anything tripwires are more likely to give discounts for Small-scale design-integrated psychic shielding and personal-sized psychic shielding.
We are going small and cheap with those after all.
I don't see any of the others discounting cognition filters either, they neither make cybernetics better nor give understanding of cognito-hazards.

I'd still hold off on cog filter... I'd be down for LS MS and I. Gellar, though. Maybe some other Psy Shield techs, if any come down in price? Or any other stuff that looks like it leads to discounts for the warp comms stuff. Or Psitech amplifiers, that ones pretty cheap right now and I bet it'll further a bunch of stuff.

Click to shrink...

We've just done improved gellar field and at least gotten a success.
Let's see how that turned out in the update before we decide if it needs further improvement.

 

[Haganeko]

We already got the stuff - and not just ore, refined metal. It discounted superconductors, not AM. Though superconductors itself might discount AM.

Click to shrink...

Oh. My bad. I guess we need to find actual alloys, maybe buildings, not just cool base ores.

If anything tripwires are more likely to give discounts for Small-scale design-integrated psychic shielding and personal-sized psychic shielding.
We are going small and cheap with those after all.
I don't see any of the others discounting cognition filters either, they neither make cybernetics better nor give understanding of cognito-hazards.

Click to shrink...

Could be! Thankfully, those are also useful. Maybe we can say we'll get Cognition Filters at the next opportunity starting on turn 26? Assuming nothing urgent comes up?

 

[Angle]

If anything tripwires are more likely to give discounts for Small-scale design-integrated psychic shielding and personal-sized psychic shielding.
We are going small and cheap with those after all.
I don't see any of the others discounting cognition filters either, they neither make cybernetics better nor give understanding of cognito-hazards.


We've just done improved gellar field and at least gotten a success.
Let's see how that turned out in the update before we decide if it needs further improvement.

Click to shrink...

I agree... I'll note, Neablis did say that getting Faith Studies and Improved sensors could offer a warp based form of Personality Checking, which sounds promising - especially as I want those two anyway.

 

[LightLan]

I agree... I'll note, Neablis did say that getting Faith Studies and Improved sensors could offer a warp based form of Personality Checking, which sounds promising - especially as I want those two anyway.

Click to shrink...

Empathy at Range (200 RP).

Improved Warp sensors are directly available.
We also have gotten discounts from warp travel in the past before (Abacus), now with literal warp sensors there we might get something discounted again.

 

[sciencegeek]

I agree... I'll note, Neablis did say that getting Faith Studies and Improved sensors could offer a warp based form of Personality Checking, which sounds promising - especially as I want those two anyway.

Click to shrink...

At a bare minimum I think we should do those two as they are very cheap for the potential reward. We especially should dive into the research on faith to see what else can be researched and if we can make maximal use of Denva. If we get a crit or high roll we might even be able to mix faith-tech with MS for maximal chaos blocking.

As an aside, does anyone know if getting a research discounted to 0 is possible? Because that sounds both plausible and possibly useful.

 

[Nightlord256]

And then there was my "next turn" pitch from voting to get triple nested shielding and then put a bound bongo in one slot inside of the existing obliette while using the other three slots for the warp lab.

Click to shrink...

Hmm, a thought. Do we want to attempt the Daemonology research before getting the warp lab now that we know the lab makes it safer? If there is one tech I'd want us to have containment measures for, Daemonology would be it.

Considering how much seems to be bottlenecked by research. Do we have any way to increase how much we can do every turn beyond the Tech priest's? Getting some research from them was a real boon but hopefully either Human Design Interfaces (200) or Companion Cogitators (450 RP) might be able to further boost the amount of research per turn. Although for the latter a bit of waiting might be advisable.

Click to shrink...

Yes, the machine spirit design we just completed had the potential to unlock a new research tech that would improve RP per action.

Human Design Interfaces also has a lot of hidden potential. Discounting our design options further would be very nice, but I think it would be a huge gift to Denva upon our return. Auto-CAD that responds to your thoughts would be a game changer for engineering.

-[] Human Design Interfaces (200 RP) You've adapted your simulation software to work with advanced neural implants, but what about your design software? If humans could play with your auto-cad as well as you can then that would help you design or even entirely farm out the design of things. (Discounts design options, may unlock technology to improve the RP generated from humans with simulation implants, either per-person or simply raising the cap.)

Click to shrink...