Is there a way to easily change all second age techs quickly without having to change each tech individually? 

Nope.You'd have to change each individually, and for each tech tree. You'd also need to spend a lot of time figuring out the tech choice weights, and it would still gimp the hell out of the AI (because it would insist on trying to upgrade its first, solitary factory all the way to the highest level it can on every new colony, using the present global econ settings regardless of how inappropriate they are - both things we can't even control in the xml right now). 

I'm not opposed to the idea in principle, but it's not something that's practical to mod in right now.

If the update you want to apply is something that can be expressed with a math operation and its targets can be identified with an xpath-rule it's possible to use XML patch tools. Simple updates are easy, like "modify ImprovementDefs.xml by multiplying Value=Value*2 in all //DefaultImprovementList/Improvement/Stats XML-blocks that have EffectType=ManufacturingCost", which would double all production costs.

Of course, such a simple modification isn't a solution for this because you don't want to increase the production costs of all improvements but just the late game ones, and you may want to increase them by different factors depending on how late they appear in the game. The latter could be done with multiple passes but the real problem is coming up with a rule that tells you which improvements are the late game ones. I don't see any XML tags or values in ImprovementDefs.xml that could be used as keys... The existing manufacturing cost? Assuming they already increase towards the end you could perhaps update those ones that are above some value?

How about lowering the bonuses of the initial factory and increasing the bonuses of the later version? Instead of changing the AI to fit the game (which we can't do) how about changing the game to fit the AI? It's like what you did with terraforming.  

I already reduced production by half in IAB. Much more than that and the start of the game becomes extremely slow.

'Production' is basically 'man-hours'. You're thinking of manufacturing, which is 'man hours * factories'. Research is likewise 'man-hours * labs' and economy is 'man-hours * markets'. Manufacturing bonuses have nothing whatsoever to do with research or wealth bonuses, but production bonuses effect all three. So, for example, manufacturing bonuses would largely be a matter of better equipment, which enhances the productivity of the workers on just one specific task (industry). Extending the work day by an hour, on the other hand, would be an example of a productivity bonus. It impacts on all three other resources, because they're all basically products of it. 

What if we moved the later tier production buildings forward in the tech tree? If it only had tier-1 improvements available at the early development phase then maybe it would go "wide" instead of "tall" in its planetary planning.

Doesn't really solve the problem. As Karl says, it just moves it to later planets.You could also get the same effect by just lowering the weight on those techs so they AI doesn't pick them til much later... but again, it's just moving the problem. 

Ultimately, SD really, really need to fix the upgrade and terraforming to be lowest possible priority. But in the mean time, just about the only way to fix excessive output is supply-side - reduce the bonuses, make the buildings cheap upgrades that don't give a large increase in output but also don't take very long to put into place. There's no reason whatsoever that a single industrial sector should give a 75% bonus. These multipliers are simply crazy increases from spammable buildings; gaining 50% production from a wonder is barely relevant when you can get 30 times that much manufacturing bonus.

Really, I'd have no great objection to seeing factories nerfed right down to give 5-10% production increase + 5-10% per upgrade, with costs at a flat 30/tier. Then reduce adjacencies to a more realistic 1-2% per level, since it's also crazy to see bonuses being doubled or tripled through placement. This would mean there's no real functional difference between building all your factories early or building one and upgrading it to the highest level you can, and a really well-laid out world would only see some 10-20% increase in output over a random mess; a top-level wold would be getting only +25% per tile and so you'd almost never see a planet with +1000% of anything. 

A lot of problems with the economy basically boil down to colony capital + buildings being powerful enough to do most things effectively in it's own right, regardless of population. Karl's early-game example was flawed, in that he missed out the +5 production you get from having a capital building - a modifier applied before almost anything else (only food bonuses are applied earlier), so in fact his 66.4 manu planet would output 99 military manufacturing. This is why colony spam is so OP atm. The colony capital is basically 5 free production before bonuse, and given that it can be 150 turns before planets start to reach 20+ pop, that's an unfathomably immense boost - so immense, in fact, that population is largely optional until the point in the game where the AI starts it's end-game strategies.

1 planet with 0.1 population and 6 top-tier factories, with no adjacency bonuses or power plants will produce 6.4 base production (25% bonus from approval), and then apply a 450% bonus from the factories. That's 35 production before military or social manu bonuses. From 0.1 pop. With 10 times the population (still only 1), it will only produce  41 - you've increase population, the supposed driving force of production, by 900% and only achieved a 20% increase in output. With 5 population, 50 times the first example, you get 61.9 output - still not even double the 0.1.

Get rid of colony capital production bonuses - or, better yet, convert them to social manufacturing - and slash the impact of buildings to 5%/tier, and the problem completely changes. My 0.1 pop world with 6 top-level factories now suddenly has 0.1 production, which is multiplied by just 250% (150% bonus plus the base 100%) to... 0.25. The identical planet with 1 pop produces 10 times as much from it's 10-times larger population. The 5 pop planet produces 50 times as much. The early-day planets aren't completely gimped - they can still build buildings from the 5 social manufacturing on the colony capital, but cannot exert influence on an empire-wide level (by contributing military production, research or wealth) until they have large, significant, productive populations to have do that. As an added bonus, by fixing colony capitals as a flat social manu bonus, you don't need to micro-manage sliders to upgrade research or econ worlds. Just point the population 100% at whatever it is you want them to do on day one and let the colony capital do the building. As a final positive, the AI now also cannot cock up it's development, because that social development MUST be spent on buildings - it's can't divert it into research or wealth due to inopportune slider settings.

If SD want to have population=production as the basis of the economy, then it should mean exactly that - the sole source of non-multiplier production should be population. Not techs, not Hives, not colony capitals. And if they want to use cumulative %-based multipliers on that from buildings for final output, the individual multipliers should be small. An factory-covered world set to 100% manu should produce 4-5 times the manufacturing output of a lab-covered planet set to 100% manu, not 40-50 times as much; otherwise, in order to make research-based planets actually capable of upgrading themselves, costs have to remain too low to be relevant to manufacturing.

The bad wording is largely SD's fault for making manufacturing and production as ambiguous as possible. Most of us have now dropped 'Total Manufacturing' as a reference to production (the way the game refers to it) and refer to 'production' and 'manufacturing' as separate concepts - production being general output of the population prior to specialization, and manufacturing being the general output of factories.

I don't really buy the idea of spending research to build research buildings or wealth to build wealth buildings, tbh. Banks need money, sure, but you don't have a bank til someone shows up with a cement mixer. As for research labs... the really expensive bit of the Large Hadron Collider was not the scientists who run it. It was the equipment, which comes from factories rather than science labs. Besides, factories aren't specifically tooled for military manufacturing - they're general manufacturing units which are presumably equipped with at least 3d printing technologies and can pump out whatever the hell you choose to ask for. There are military manufacturing bonuses elsewhere; these really don't have anything to do with research or wealth even in the production of the buildings.

Truth be told, there really is nothing much wrong with the fundamentals of the production * manufacturing bonuses * mil/social bonuses model. What we have here is a balance problem - the system is fine, but the numbers being fed into it are out of whack. Compare and contrast with the issues around LEP, which is a systemic problem - in that case, the system itself is unworkable because it's riddled with ineffectual compromises and perverse incentives, so no amount of tweaking the values can fix it; even if we accept that you can keep approval as the limiting device (which is itself highly debated, and the consensus view is moving ever-increasingly against it), it still requires a complete change in how approval is calculated just to make a semi-workable system.

And in real life, even if the research lab which comes up with the better toilet doesn't mass-manufacture them, or distribute them to other research labs. They send the design to a toilet factory, which does that. And this is how the game is set up - your research lab generates research, which you use to acquire 'toilet specialization 3', and then your factories can mass-produce them and equip your labs with the best toilets that your civilization can devise. And that's largely how civilization has worked since Adam Smith cottoned on to the advantages of the division of labour.

The alternative you're proposing is a group of scientists gathering together, doing some maths, and a laboratory materializing out of their equations. They then, making use of the computers and blackboards and suchlike in their new lab, write some more complex math that causes an additional research lab to pop into existence next door. Converting factories to only produce military production is one thing, but you'd then need to give research labs and markets a social manufacturing value... which would mean the best way to build a wonder or upgrade a planet would be to cover it in research labs. Including upgrading factories - your massive industrial base would be useless for upgrading itself until a bunch of bankers showed up, presumably with a lot of construction equipment.

It turns out that producing toilets and aircraft is actually a pretty similar process tbh - ultimately, all forms of manufacturing boil down to taking some matter and then re-shaping it. Research, conversely, doesn't involve taking matter at all in a lot of cases, but rather giving people enoguh time and comfort to sit and think really, rally hard about something. Wealth largely consists of taking something and changing it's location.

first I'm against anything that nerfs Ai. If you didn't. Start out with giving the capital production. I would recommend excempting the yor from this, and this is why it is already hard to increase population with them. They need a fair amount of starting production. I don't. Have a problem giving different races different playstyles.

@Naselus

You say it's like that in the future like it's a fact. Why would we need a toilet factory to make a toilet with 3d printing+? We are in the future after all, and this design is better for the game. Most likely with robotics and 3d-printing it's a matter of allocating resources (the wheel in Galciv3), to get whatever you need. The infrastructure is most likely not needed except for the big stuff. (like spaceships etc.) See the point?

So you are saying that raw production is just a whip to get them working? Like some lab in a poor country could come up with some revolutionary particle physics discovery out of nothing, if only they had this fancy building? It's an ongoing investment in research, this game just treats it all wrong. You can have the most fancy equipment in the world, and research nothing if the scientists do not know what they are doing. 

If we are getting started on production and IRL, Malevolent etc. production bonuses are just plain ridiculous (whip to work harder? yeah that works), and should be an example of why any argument defending the current system with IRL examples is not viable at all.