Automation buildings are weird and should be changed.

[Sweeze]

i was about to make a thread on this topic, when i saw it already exists. here's things i find weird about automation:

• automation seems to try to do the same thing as robot construction, but it's somehow instant? what's the point of robots then?
• you can automate specialist jobs before you can automate worker jobs, because you have to research district specialization first
• pricing is weird, it scales with amount of districts, not amount of jobs. so automation is most price effective when you have a single city district with all the building slots filled.
• you're at some point incentivized to REDUCE automation? that seems incredibly backwards.

Yeah, didn't mention a lot of this because its kind of just a vibes and feel thing, but its really weird that automation exists next to robots.

If automation merely *enhanced* native population's working ability, so that 1 pop could count as 2 pops of workforce, it could arguably make sense. It's like investing in better drills for your miners. But instead it just outright filled 25%/50% of the job even if the whole planet is empty. If the building can do 100% of the work as a pop then why is it not a pop? And if it can do 100% of the job then why can it only fill 25%/50% of the workforce?

I just realized two important things:
+75 from basic resource districts is always more efficient than +50 from city districts,(and same efficiency)
and unless energy efficiency, automation almost always results in a net loss — except in very specific situations.

Yep, the thing is that you're basically trading energy for pops so you want to load automation at whatever point in your economy has the most pops per district.

In theory differences in job productivity could alter this (if your pops had +200% to energy production then you'd want them working in energy and your automation in alloys), but bonuses like that aren't practically large enough to affect the overall calculation.

 

[Abdulijubjub]

In theory differences in job productivity could alter this (if your pops had +200% to energy production then you'd want them working in energy and your automation in alloys), but bonuses like that aren't practically large enough to affect the overall calculation.

Bonuses like that don't exist anymore. 100 workforce is now 100 workforce, no matter where it's from.

If your pops have a +300% efficiency bonus (4x) it just means that 100 automated workforce is only replacing 25 pops, but it's doing the same thing those 25 pops would have. The tradeoff between the two only matters once your planets fill up, and you're choosing whether to have automated jobs and unemployed pops or just have pops work the job.

 

[Critical Ethics]

The automation energy cost really should be per (100) pop replaced, although the coding might be unreasonably difficult.

My ideal solution would be for the automation cost to key off the pops replaced and also not just be an energy cost. Currently automation pushes pops downwards, pulling them out of high tiers while increasing your need for technicians. This seems backwards to me. The obvious solution would be an alloys cost, but that further buffs alloy worlds since they're the only ones that could avoid the logistics costs just by doing their day to day. Same issue with Strategic Resources.

Having them cost engineering research could be interesting, or perhaps engineering + physics or society or more engineering to represent it being more difficult to maintain and improve upon institutional knowledge when people are no longer interacting with the day to day. More automation requires more scientists. Or heck - every automation building consumes energy, alloys, and at least one form of science, so outside of extremely, I'd even say overly, generalized builds you're always going to be importing something, and the general result of a highly automated empire is pushing people from menial labour into specialised machining and tech roles.

 

[Critical Ethics]

Bonuses like that don't exist anymore. 100 workforce is now 100 workforce, no matter where it's from.

If your pops have a +300% efficiency bonus (4x) it just means that 100 automated workforce is only replacing 25 pops, but it's doing the same thing those 25 pops would have. The tradeoff between the two only matters once your planets fill up, and you're choosing whether to have automated jobs and unemployed pops or just have pops work the job.

I may be misunderstanding you, but while pop_workforce_mult is almost gone from the game *_bonus_workforce_mult is in heavy use. This doesn't fill the job up quicker, but it overflows, costing you more job inputs for a matching increase in job outputs but no increase in pop or district requirements. So a district with 300 alloy jobs and a +20% pop_bonus_workforce_mult is still consuming 300 pops and 3 district slots, but also consuming 360 workforce of minerals and outputting 360 workforce of alloys. A +20% *_produces_mult (like the mining techs) only applies to the output side so it's strictly better in isolation, but since they're more common than bonus_workforce and the two multiple from each other it's a bit more complex. Research bonus multipliers are entirely *_bonus_workforce_mult (that is actually relevant to my last post - we can't actually have a +20% *_produces_mult for research because it's game breaking)

*_automated_workforce of 0.25 fully replaces 25% of your workforce, but that 25% does not get pop_workforce_mult bonuses. So in our 300 example above the total effective workforce drops from 360 to 75 + (225*1.2) = 345, both for mineral input and alloy output. Automation means you need less pops but more land because your maximum effective workforce, and therefore maximum effective output per district, goes down. e: But yes you then still get to multiply the full 345 by your output_bonus

This also answers this question:

• automation seems to try to do the same thing as robot construction, but it's somehow instant? what's the point of robots then?

The game mechanics difference is that a robot with harvester hands and various other pop efficiency bonuses does a lot more farming per square meter than an automated system and can be job swapped quickly if needed while the only way to boost automated production is building an entire new district. Flavourwise a robot is a semi-autonomous agent that can adapt and optimise to complex situations while an automated system is integrated into the district's facilities and can only perform rote tasks within strictly defined parameters.

 

[WJJellO]

Building upkeep reductions apply to automation buildings. If you stack enough of those (read: roll the Oracle governor), it drastically changes the cost/benefit math in favor of using them.

 

[Abdulijubjub]

I may be misunderstanding you, but while pop_workforce_mult is almost gone from the game *_bonus_workforce_mult is in heavy use. This doesn't fill the job up quicker, but it overflows, costing you more job inputs for a matching increase in job outputs but no increase in pop or district requirements. So a district with 300 alloy jobs and a +20% pop_bonus_workforce_mult is still consuming 360 pops and 3 district slots, but also consuming 360 workforce of minerals and outputting 330 workforce of alloys. A +20% *_produces_mult (like the mining techs) only applies to the output side so it's strictly better in isolation, but since they're more common than bonus_workforce and the two multiple from each other it's a bit more complex. Research bonus multipliers are entirely *_bonus_workforce_mult (that is actually relevant to my last post - we can't actually have a +20% *_produces_mult for research because it's game breaking)

*_automated_workforce of 0.25 fully replaces 25% of your workforce, but that 25% does not get pop_workforce_mult bonuses. So in our 300 example above the total effective workforce drops from 360 to 75 + (225*1.2) = 345, both for mineral input and alloy output. Automation means you need less pops but more land because your maximum effective workforce, and therefore maximum effective output per district, goes down. e: But yes you then still get to multiply the full 345 by your output_bonus

I think we have the same understanding of how it works, but are just framing it differently. i.e. no actual disagreement

I was thinking.. so long as you have the space, turning X energy into Y jobs worked is either a good trade or not. The pops' bonus_workforce_mult bonuses don't enter into it, only non-pop-dependent output and upkeep modifiers. Thus, how efficient your pops are doesn't matter.

But if you don't have the space, then you have to start making tradeoffs for e.g. whether you'd rather turn off or downgrade automation in order to get another pop with +X% efficiency (more jobs worked in the same space). And the pops are such a clear winner there that there's not even a discussion: of course you just turn off the automation and save yourself energy in addition to getting more output.

By contrast, in 3.14, pop-specific output modifiers changed the e.g. input/output ratios, such that you may even prefer to have no job at all vs. automation (ex. better to hypothetically have 30 pops split 10/20 between +100% output metallurgists and miners than to automate 20 +0% output metallurgy jobs and need 40 miners in upkeep (more pops, same output, plus automation costs). And it was easier to run out of space, for research especially.

 

[Red Death]

I was thinking.. so long as you have the space, turning X energy into Y jobs worked is either a good trade or not. The pops' bonus_workforce_mult bonuses don't enter into it, only non-pop-dependent output and upkeep modifiers. Thus, how efficient your pops are doesn't matter.

But if you don't have the space, then you have to start making tradeoffs for e.g. whether you'd rather turn off or downgrade automation in order to get another pop with +X% efficiency (more jobs worked in the same space). And the pops are such a clear winner there that there's not even a discussion: of course you just turn off the automation and save yourself energy in addition to getting more output.

Or in other words, automation might be a good idea when you have more space for development than you have pops (Or you would rather have your pops work something else). But if you have excess pops, there's not much reason to be using automation. Unless you for example end up with the situation where you would be better off paying for the automation because you've modified Civilians to be more productive.

 

[Critical Ethics]

I think we have the same understanding of how it works, but are just framing it differently. i.e. no actual disagreement

I made a fairly major typo which is now remedied, typed 360 pops and 330 workforce instead of 300 pops and 360 workforce, but based on your response I think you got what I meant and yeah we're mostly on the same page.

I was thinking.. so long as you have the space, turning X energy into Y jobs worked is either a good trade or not. The pops' bonus_workforce_mult bonuses don't enter into it, only non-pop-dependent output and upkeep modifiers. Thus, how efficient your pops are doesn't matter.

But if you don't have the space, then you have to start making tradeoffs for e.g. whether you'd rather turn off or downgrade automation in order to get another pop with +X% efficiency (more jobs worked in the same space). And the pops are such a clear winner there that there's not even a discussion: of course you just turn off the automation and save yourself energy in addition to getting more output.

By contrast, in 3.14, pop-specific output modifiers changed the e.g. input/output ratios, such that you may even prefer to have no job at all vs. automation (ex. better to hypothetically have 30 pops split 10/20 between +100% output metallurgists and miners than to automate 20 +0% output metallurgy jobs and need 40 miners in upkeep (more pops, same output, plus automation costs). And it was easier to run out of space, for research especially.

Quick caveat to the first paragraph I want to make is some emphasis on the difference between land and developed land. But basically yeah, automation should work as basically trading in lower land efficiency for higher pop efficiency.

Everything else you said also holds true, but only because there's no meaningful early game opportunity cost to the building slot itself. If your opportunity cost is not building a 200 pop job building that you don't even have pops for then yeah building the automation is a no brainer. If building the automation building requires you to give up building an upkeep reducer or an output increaser then they're not just costing energy, they're worsening the input : output ratio of both the automated jobs and your remaining, non-automated pops. You get the 3.14 setup but even more so, with an obvious and easy to understand tradeoff... except you don't because, well, see previous.

See what I meant by death by a thousand cuts?

 

[Abdulijubjub]

The second is that your first paragraph only holds true if there's no meaningful opportunity cost to the building slot itself. If your opportunity cost is not building a 200 pop job building that you don't even have pops for then yeah building the automation is a no brainer. If building the automation building requires you to give up building an upkeep reducer or an output increaser then they're not just costing energy, they're worsening the input : output ratio of both the automated jobs and your remaining, non-automated pops. You get the 3.14 setup but even more so, with an obvious and easy to understand tradeoff... except you don't because, well, see previous.

In the current build, I don't think that ever really comes into play except on mixed output worlds (though I do wish there were enough choices and ways to customize it that it would happen more often).

If you're automating a rural district, there are only 1-2 buildings that actually boost things, so there's no conflict. And in an urban district, you've got 11 buildings slots that can take supporting buildings and obligatory planet uniques like Genomic Research Center. That's a lot of building slots. Only if you have 2-3 major job types that warranted multiple set of supporting buildings can you actually run into building slot pressure (ex. both researchers needing all 6 support buildings, plus CG needing its 3).

(agree on everything else)

 

[Sweeze]

Incidentally, with how automation works now once you have sufficient tech and megastructures, assuming both energy and room for jobs is functionally unlimited, this would mean that any empire which can't stack a really high -empire size from pops should genocide their own people and rely only on automation. Sure, only half of your jobs are filled, but at zero empire size.

This would make for a funny FE type.

 

[Red Death]

I guess Empire Size might also be a factor: I think, unmodified, a district adds less to empire size than the pops required to work all the jobs produced by it.

 

[Abdulijubjub]

Incidentally, with how automation works now once you have sufficient tech and megastructures, assuming both energy and room for jobs is functionally unlimited, this would mean that any empire which can't stack a really high -empire size from pops should genocide their own people and rely only on automation. Sure, only half of your jobs are filled, but at zero empire size.

This would make for a funny FE type.

Honestly, they should just do this for the FE buildings.

Make them start half automated instead of producing base resources, such that having a regular FE building and an automation center makes them fully automated.

Then the Xenophile FEs will just be sitting on their planets full of Hedonists, with automated buildings producing everything for them. The Xenophobe/Materialist FEs will have their Nerve Stapled or robot slaves work the non-automated rural districts, while only the city is fully automated, And the Spiritualists, of course, won't use any additional automation at all, preferring to leave the labor to Acolytes of the Plow, Acolytes of the Hammer, and Acolytes of the Class-3 Singularity Power Generation Center.

 

[Critical Ethics]

In the current build, I don't think that ever really comes into play except on mixed output worlds (though I do wish there were enough choices and ways to customize it that it would happen more often).

If you're automating a rural district, there are only 1-2 buildings that actually boost things, so there's no conflict. And in an urban district, you've got 11 buildings slots that can take supporting buildings and obligatory planet uniques like e.g. Genomic Research Center. That's a lot of building slots. Only if you have 2-3 major job types that warranted multiple set of supporting buildings can you actually run into building slot pressure (ex. both researchers needing all 6 support buildings, plus CG needing its 3).

(agree on everything else)

We also agree that that's how things are, because my complaint is that that's how things are!

 

[Abdulijubjub]

We also agree that that's how things are, because my complaint is that that's how things are!

Fair enough.

For what it's worth: I think adding more buildings that are on the same level of power as what we currently have (without pulling the power from somewhere else) would be a mistake, but for a game pacing rather than building design reason. Stellaris already has major issues with the economy rescaling to an absurd degree as you go through the tech tree, which leads to things like difficulty tech scaling being needed to make the end game techs not blaze by at 1 month each. Unlocking even more "everything is 20% more effective" buildings as you go would exacerbate that.

What I would love would be more mixed blessing buildings that were more about customization, rather than raw increases.

Ex. Instead of more "+15% efficiency", "-20% upkeep", or "+20% output"...

• "+1 alloy, +6 mineral upkeep per 100 metallurgists": not even nominally net positive without upkeep reductions/output modifiers, but generally useful with the right support (and a godsend if you can print minerals and just need somewhere to dump them). And make it mutually exclusive with the % upkeep reducers.
• "-0.5 alloys, -2 mineral upkeep per 100 metallurgists": good if you really want to drive upkeep down, but not even nominally profitable in the general case (and, ironically, even less profitable if you have upkeep reducers).
• Let different empires want to customize differently, instead of building all the same generic boosters that are always good.

 

[Critical Ethics]

Fair enough.

For what it's worth: I think adding more buildings that are on the same level of power as what we currently have (without pulling the power from somewhere else) would be a mistake, but for a game pacing rather than building design reason. Stellaris already has major issues with the economy rescaling to an absurd degree as you go through the tech tree, which leads to things like difficulty tech scaling being needed to make the end game techs not blaze by at 1 month each. Unlocking even more "everything is 20% more effective" buildings as you go would exacerbate that.

What I would love would be more mixed blessing buildings that were more about customization, rather than raw increases.

Ex. Instead of more "+15% efficiency", "-20% upkeep", or "+20% output"...

• "+1 alloy, +6 mineral upkeep per 100 metallurgists": not even nominally net positive without upkeep reductions/output modifiers, but generally useful with the right support (and a godsend if you can print minerals and just need somewhere to dump them). And make it mutually exclusive with the % upkeep reducers.
• "-0.5 alloys, -2 mineral upkeep per 100 metallurgists": good if you really want to drive upkeep down, but not even nominally profitable in the general case (and, ironically, even less profitable if you have upkeep reducers).
• Let different empires want to customize differently, instead of building all the same generic boosters that are always good.

OK so this is what I keep trying to explain. You're talking about zones as they were originally described, just with more steps.

Zones as originally described:
Doubling up identical zones did not double up building slots. Just the amount of jobs.

Pretty much everything you could build in a zone was locked to the zone. No double dipping into city slots.

This means you get three upgrades to each zone and that's. It. You can research 50 different buildings and all that does is give you options. Once you have three upgrades you've capped out. Not 11. 3.

This results in what you described but requiring, ironically, fewer buildings. Each zone is basically a build your own building. If I put a +20% booster and a +25% jobs and a 25% automation (with some energy cost on it), and you put a +25% pop efficiency and a -20% upkeep and a +2 output/+2 upkeep building, we've built two very, very different districts. There's 20 different possible combinations from just 6 buildings, never mind planet specials.

The kind of boost stacking you're worried about are one of the things that zones were supposed to solve, and the tradeoff buildings you're describing were one of the things they would innately facilitate.

But instead...

 

[turtleturtle111]

Also automation buildings have no empire size penalty so if you are min maxing they are actually the bees knees. No pops no size just energy costs. Necrphage to prevent pop growth and just go cybernetic to get the cost reduction and planet spam your way to victory.

 

[honigbrötchen]

Incidentally, with how automation works now once you have sufficient tech and megastructures, assuming both energy and room for jobs is functionally unlimited, this would mean that any empire which can't stack a really high -empire size from pops should genocide their own people and rely only on automation. Sure, only half of your jobs are filled, but at zero empire size.

This would make for a funny FE type.

I was trying to do that in my first game with 4.0 as a virtual empire, since pops just vanish once you forbid the non-automated jobs so there's no need for self-purge or "release-all-your-pops-as-a-vassal" shenanigans, but the +empire size for planets and -empire size for pops is kind of counter intuitive for that playstyle.

Does anyone have an idea if or how such a (meme-)build could work?

My idea was to focus on building upkeep reduction, so:
-10% from the prosperity tradition
-40% from three coucilors and one ruler with Architectural Sense III (so gestalts are a no-go, I guess)
-10% from urban designation / -25% from ecumenopolis designation

That would be -60% upkeep for normal planets and -75% for ecus

On a tall build with only few sectors, you could have another offical as sector governor, -10% and another -25% for the whole sector from Architectural Interest and the Urbanist destiny trait respectively, but this whole setup seems like it wants to be wide, so needing a high level official for every sector isn't sustainable.

Then there's either Synthetic Virtual Dictatorship for another -20%, which would fit well with all the officials on the council since virtual leaders are immortal,
or choose any authority you'd like with Cybernetic ascension for -25% upkeep reduction from a cyborg ruler

without any official governors, that would be:

-80% upkeep reduction for virtual dictatorship on all planets and -95% on ecumonopoli
-85% for cyborgs and capped reduction on ecus


That's only the upkeep side of things, though, and despite the extensive discussion in this very thread about efficiency and building choices and whatnot I have no idea what buildings or civics/traditons would be ideal for a nobody-is-working setup like that.

There's also the problem of what to do with your pops.
Probably a civilian build would be kind of mandatory if you insist on having nobody work? But I still kind of want to make this work with virtual somehow despite the empire size problem.

There's just something funny about a completely automated, fully digital empire that produces lots of stuff with more buildings than pops on a planet.

On the other hand, if you don't go virtual you could play with the mammalian sloth portrait for maximum meme capacity...

 

[Abdulijubjub]

Also automation buildings have no empire size penalty so if you are min maxing they are actually the bees knees. No pops no size just energy costs. Necrphage to prevent pop growth and just go cybernetic to get the cost reduction and planet spam your way to victory.

I was trying to do that in my first game with 4.0 as a virtual empire, since pops just vanish once you forbid the non-automated jobs so there's no need for self-purge or "release-all-your-pops-as-a-vassal" shenanigans, but the +empire size for planets and -empire size for pops is kind of counter intuitive for that playstyle.

Does anyone have an idea if or how such a (meme-)build could work?

My idea was to focus on building upkeep reduction, so:
-10% from the prosperity tradition
-40% from three coucilors and one ruler with Architectural Sense III (so gestalts are a no-go, I guess)
-10% from urban designation / -25% from ecumenopolis designation

That would be -60% upkeep for normal planets and -75% for ecus

On a tall build with only few sectors, you could have another offical as sector governor, -10% and another -25% for the whole sector from Architectural Interest and the Urbanist destiny trait respectively, but this whole setup seems like it wants to be wide, so needing a high level official for every sector isn't sustainable.

Then there's either Synthetic Virtual Dictatorship for another -20%, which would fit well with all the officials on the council since virtual leaders are immortal,
or choose any authority you'd like with Cybernetic ascension for -25% upkeep reduction from a cyborg ruler

without any official governors, that would be:

-80% upkeep reduction for virtual dictatorship on all planets and -95% on ecumonopoli
-85% for cyborgs and capped reduction on ecus


That's only the upkeep side of things, though, and despite the extensive discussion in this very thread about efficiency and building choices and whatnot I have no idea what buildings or civics/traditons would be ideal for a nobody-is-working setup like that.

There's also the problem of what to do with your pops.
Probably a civilian build would be kind of mandatory if you insist on having nobody work? But I still kind of want to make this work with virtual somehow despite the empire size problem.

There's just something funny about a completely automated, fully digital empire that produces lots of stuff with more buildings than pops on a planet.

On the other hand, if you don't go virtual you could play with the mammalian sloth portrait for maximum meme capacity...

These builds work even without leaning into them. Just colonize planets you wouldn't otherwise need, take Prosperity for -10% upkeep, and set every designation on these otherwise empty planets to "urban" for another -10%. You can also take Expansion to shave empire size from planets down from 5 to 2.5 (after Imperial Prerogative, which you should take in almost all builds anyway).

4.0 has replaced "colonize every planet to get more pop growth" with "colonize every planet to get more places to slap down automation buildings". They don't even get negatively affected by habitability.

It's why the base design of these buildings annoys me so. They ought to scale your pops, not replace them.

---

But for that build, specifically, I'd go for something that boosts output, rather than reducing upkeep. Shaving off another 0.9 or 1.0 energy per rural/urban district is less valuable than a 10% boost to output.

Shelled would get you the same housing reduction, and cybernetics have tons of output boosting governments/edicts. Oligarchic Overclocking would let you have 100 100% happiness rulers as the only pops on the planet (maxing out stability) while also cranking out as many resources as you can from the automated worker/specialist jobs.

Aka, Solaria.

 

[Sweeze]

I was trying to do that in my first game with 4.0 as a virtual empire, since pops just vanish once you forbid the non-automated jobs so there's no need for self-purge or "release-all-your-pops-as-a-vassal" shenanigans, but the +empire size for planets and -empire size for pops is kind of counter intuitive for that playstyle.

Does anyone have an idea if or how such a (meme-)build could work?

My idea was to focus on building upkeep reduction, so:
-10% from the prosperity tradition
-40% from three coucilors and one ruler with Architectural Sense III (so gestalts are a no-go, I guess)
-10% from urban designation / -25% from ecumenopolis designation

That would be -60% upkeep for normal planets and -75% for ecus

I don't think you need to focus on upkeep reduction that heavily. You need enough to make the automation decently profitable, but bonuses to output also do that.

What's more important is being able to just create a lot, lot of jobs ASAP. As soon as you unlock Ecus or Ringworlds and get them down your power scale goes off the charts. You're basically Virtual with no drawbacks. Before then I'm not sure what the best play is.

 

[Falrinn]

Overall it does feel like the current implementation of automation felt more like a placeholder to test the technical principles of the feature rather than a fully-fleshed out feature to be included in a final release. It's notable in that it's the only building that both interacts directly with districts but doesn't come in unique flavors for each district, and uses the same icon as robot assemblers.

Personally I would consider abstracting out automation. Instead of automating 25/50% of job slots, it just increases job efficiency by a significant amount while decreasing the number of job slots available. The ratio of job efficiency bonus to job slot reduction would be significant enough that only the most heavy efficiency-stacking builds reach the point where automation might be suboptimal.