Moved construction scaling with gold reserve to max out at 50% gold reserve, down from 100%.

Changed bureaucracy cost estimation to be based on bureaucracy usage as opposed to production.

Decreased the military spending of subjects to account for their income transfer to the overlord.

Added a cap for AI military spending estimate at 30% of their total income.

Military spending will no longer lower army size below 10 units.

Lowered minimum navy sizes so countries struggling financially will not try to expand their navy.

Changed AI weight for capital ships to no longer depend on their country rank.

Added back AI weight for using transfer subject wargoal on Schleswig-Holstein since Prussia will eventually need them for unification. Liberate subject will be used if a transfer is invalid i.e. Holstein is a personal union.

Undid the AI strategy weight for manifest destiny since the more aggressive strategy caused the US to take more states than intended. The issue with army build-up is addressed elsewhere.

Slightly adjusted the target of Unify Germany's aggression towards countries irrelevant to the unification.

All political strategies now have a base weight of 10.

When choosing political strategy, the ruler IG is now factored out into its own clause and given a +10 weight across the board. It used to piggyback on the ideology clause which made strategies with a higher ideological weight more likely than others.

Weighted AI towards the conservative political strategy when relevant interest groups are in power. This is to maintain parity with other strategies that have the same behavior.

Lowered the law enactment success threshold to 10% for the reactionary strategy, similar to other strategies.

Qing will no longer choose the Preserve Mandate of Heaven strategy once nationalism is researched. Qing is still weighted towards a reactionary political strategy due to having a Traditionalist emperor, but is no longer barred from changing laws.

Lowered AI's base institution spending but increased its scaling with the number of institutions to compensate. The goal of lowering base weight is to encourage AI to invest in new institutions as they are unlocked, as opposed to spending all the budget on the first institution leaving nothing for the latecomers.

Undid AI's preference leveling up education and health systems institutions since the investment scaling addressed the problem more organically.

Lowered AI's threshold for downgrading institutions.

Lowered AI's absolute surplus threshold for increasing government spending.

Lowered AI's gold reserve threshold for constructing government buildings to 10%. Vanilla 1.8 lowered the threshold for production buildings from 20% to 10%, but didn't change the thresholds for government buildings.

Raised critical threshold for construction from 75% to 90%.

Raised critical threshold for innovation from 75% to 100% the innovation cap.

Lowered AI weight for construction sectors. This is to address the issue of AI building too many useless construction sectors post-scarcity.

Changed AI weight for constructing military buildings in states with non-accepted pops from -250 to +250. Since 1.8 all pops are considered non-accepted which lowered the AI weight for barracks across the board. A complete fix is impossible but this at least balances the scores for military buildings relative to other government buildings.

Lowered AI weight for constructing buildings that produce favored goods. This parameter caused certain high throughput industries (steel, automobile) to receive too much weight even when the price is already depressed. In testing, this change noticeably improves GDP, especially towards endgame and full employment, when the construction loop is less important.

Added negative AI weight for passing laws that would add institutions if when bureaucracy cost is higher than 35% of their normalized income. This is to prevent countries with insufficient taxation from bankrupting themselves from bureaucracy.

Removed AI weight for passing schools when Japan has the Terakoya journal entry.

Added negative AI weight for going back to Per-Capita Taxation once Proportional Taxation is enacted.

Relaxed the AI triggers for banning slavery to no longer require a movement. The abolition movement does NOT fulfill the trigger since it is only "neutral" towards banning slavery.

Raised AI value for kill rate and recovery rate so that the medical mobilization options would be activated.

Increased weight for Steam Turbine. The Coal-Fire Plant PM almost doubles AI score for building power plants, which is important for era 4.

Increased AI weight for researching Telephone and Radio. Radio is required for multiple military units so it should be researched first.

Changed how some building files are modified to hopefully be more compatible with other mods.

Slightly lowered the minimum wanted construction to match the should-have surplus threshold.

Military sizes scaling with income now has a floor at 50% the population factor, instead of at 15% income.

Undid the change to tolerance before downsizing military. It's no longer necessary with the military size floor. AI seems to ignore it in 1.8 anyways.

Changed navy size's income scaling to be less dependent on the proportion of population that is coastal. This means Prussia and Austria will maintain a bigger navy.

Made AI unable to select Placate the Masses strategy unless they start with it.

Removed Placate the Masses' strategy's altered tax level and building weights.

Changed the starting wage level to neutral for all countries. AI is unable to change wage level. As such the starting wage puts some countries in a handicap for the entire game. Notably Ottoman Empire starts in low military wage in vanilla. EIC is exempted to help them stave off Sepoy Mutiny.

Made Unify Germany strategy's elevated aggression start at an earlier point.

Made Unify Germany strategy no longer care about Baltic Governate due to the country's North German primary culture.

Added weight for naval bases for Unify Germany strategy in case it's needed to reach Schleswig-Holstein.

Weighted the US towards Expand Core Territory strategy while manifest destiny journal entry is active.

Weighted Afghan nations strongly towards Expand Core Territory strategy while the unification journal entry is active. Unrecognized countries tend to adopt the armed isolationism diplomatic strategy, which prevents them from choosing conquer state as a war goal despite being strongly weighted towards aggression.

Made AI only willing to start knowledge sharing if the subject is a direct subject with more than 40% of the overlord's population.

Increased max progressiveness of conservative and nationalist political strategies from 50 to 100. This allows changing directly from Traditionalism (-50) to Interventionism (+50).

Lowered the minimum base success chance before AI would consider changing a law across the board. It is now 20% for reactionary strategy and 10% for the rest. 10% movement support is used in various AI triggers for laws, so it seems fitting for that to be the threshold.

Incraesed change law chance for non-reactionary AI strategies. In vanilla, AI would have a higher chance for law changes when threatened by an insurrection. That additional chance is now added to the base chance.

Undid additional pro-IGs for various strategies. It was necessary prior to 1.7.4 to give AI an incentive to pass certain laws. After 1.7.4 AI is now weighted to pass various laws irrespective of IGs.

Relaxed the AI triggers for changing from Serfdom to Tenant Farmer to no longer require a political movement. There is no movement that satisfies the condition.

Relaxed the AI triggers for changing from Traditionalism to Agrarianism or Interventionism to no longer require a political movement. There is no movement that satisfies the condition for Agrarianism. Only modernizer movement wants Interventionism and it is created by events primarily.

Increased AI weight for tax reforms from 400 to 750.

Changed the trigger for transitioning to Proportional Taxation to be 1.5 pounds of GDP per capita, instead of peasant percentage.

Made AI want a buffer of 10 tax capacity when constructing government administration.

Lowered AI weight for placing government administration overall. This means institution expansion will have lower or equal priority compared to universities and ports.

Changed AI evaluation for lacking qualificaiton for government buildings from -90% to -25% to stop Japan from building government admins in Hokkaido.

Raised the threshold before AI ignores budget surplus percentage from 30k/50k to 300k. This means larger countries such as Qing will want more surplus before increasing spending.

Lowered base score for production methods from 1000 to 300. This makes multiplicative factors such as labor saving valued less.

Raised privatization threshold to 0.25 when a country has have over 300k fixed income surplus. This means buildings that produce government goods, notably tooling workshops, will be privatized towards the late game.

Increased the evaluation score for wanted goods with a high price (+25%) from +2.0 to +8.0. The weight is excessive so that power plants can be built even when there is no local workforce.

Removed automobiles from wanted goods in AI strategy. This helps balance AI's PM choice between automobile and engines.

Fixed AI weight for resolving the Schleswig Holstein question targetting incorrect country and wargoal.

Removed building weights in default strategy.

Made AI more willing to lower subject payments. In my opinion the current subject payment of 30% is unreasonably high. A lower payment is mutually beneficial in saving the overlord the trouble of putting down rebellions all over the world.

Fixed incorrect triggers for AI weights for Agrarianism and Laissez Faire.

Incaresed AI weight for government admin to prefer incorporated states with insufficient tax capacity to overcome insufficient qualification multiplier.

Undid a number of money spending and taxation changes. Those changes were made at a time when AI were worse at maintaining their finances, and when radicals were more problematic. I don't think they are necessary anymore.

Raised AI's wanted threshold for increasing convoys from 7.5% to 15%.

Increased AI's critical threshold for increasing convoys from -25% to 0.

Increased AI's bias for investing domestically from 2 to 5.

Removed stickiness when upgrading production methods. This is specifically to fix barracks PMs not getting upgraded, since the relative benefit is too small.

Lowered the weight for employment changes. This is specifically to fix urban center PMs not getting selected due to the increased employment.

Sync'd with 1.8.6 base game.

Removed rural folks from pro-IGs of liberal political strategy. 1.8.0 added them to anti-IG.

Reverted most of the tax level changes. 1.8.4 made AI much less likely to stay at high taxes.

Undid the change to number of IG that AI considers when reforming government. It is raised from 16 to 64 in 1.8.6 vanilla.

Undid the production method weight for labor saving as 1.8.6 changed how it is calculated.

Land reform weight fix is part of vanilla as of 1.8.6.

Amur fix is part of vanilla as of 1.8.6.

Sync'd with 1.8.4 base game.

Increased the conscript ratio of national militia from 3 to 5. This reverts a vanilla change in 1.7.4.

Undid weight changes to economic system and land reform as the new movement system makes the conditions in vanilla sufficiently easy to achieve. It now only fixes a bug in AI weight for changing from Serfdom to Commercialized Agriculture.

Undid AI's critical innovation threshold change. It was raised in vanilla 1.8.4.

Sync'd with 1.8.3 base game.

Raised AI's critical innovation generation to 0.625.

Slightly reduced ratio of military spending to total income minus bureaucracy cost. The bureaucracy cost was originally estimated to be higher.

AI strategy now only affects the population factor of army / navy sizes calculation and not the income factor. This softcaps how much Prussia would expand their army in pursuit of unification.

Being on the British Isles no longer shifts the balance from army to navy.

Lowered the cash reserve before queuing more construction from 10% to 1%.

Lowered the threshold for deleting army from 4 times to 2.5 times. It was originally raised to prevent massive army deletions at game start by Qing in 1.7. Now that Qing no longer starts with Poor Laws, I think it is okay to lower this value. It also helps EIC to stay afloat by downsizing its massive starting army.

Lowered Unify Germany strategy's recklessness from 75% to 25%. This stops Prussia from launching unwinnable unification plays.

Undid the increased chance to start passing a law when an autocracy pursues a liberal strategy. This was meant to help Alexander II of Russia. In 1.8 Russia got its own AI strategy to do exactly this.

Sync'd with 1.8.1 base game.

Added a multiplier for wanted construction that scales with gold reserve. It starts when gold reserve exceeds 50%, and maxes out at 2 times when gold reserve is full. This is to mitigate a bug where the AI refuses to build more construction sectors.

Undid AI weight for establishing schools, since vanilla 1.8.0 added one.

Renamed file prefixes from "kai_" to "ai_" to reduce mod incompatibility.

Sync'd with 1.7.6 base game.

Implemented AI weight for construction sectors to prefer states with iron potential.

Removed AI weight for construction sectors in populous states.

Implemented AI weight for government admin to prefer incorporated states with insufficient tax capacity.

Lowered success chance requirement for passing laws across the board to 15% for conservative / reactionary strategies and 10% for the rest.

Increased AI weight for establishing schools significantly and relaxed its conditions.

Increased AI weight for abolishing traditionalism and serfdom back to vanilla level.

Increased AI weight for investing in education and health systems institutions.

Increased AI weight for transitioning to propotional taxation once less than 40% of the population are peasants.

Decreased navy size scaling with income by 25%.

Removed AI's infamy tolerance scaling with Nationalism / Pan-Nationalism that was introduced in vanilla 1.7.0. Now AI will think twice before going over 25/50 infamy unless they have one of the strategies explicitly encouraging conquest.

Increased threshold for raising taxes from 120% income-to-expense ratio to 124%.

Increased desired innovation generation to 50% over innovation cap.

Lowered AI's desired taxation level by one level. This means AI will not raise taxes to high unless they need to deal with debt.

Reduced AI's navy size scaling with income by 25%.

Massively rebalanced evaluation of mobilization options. Compared to vanilla, the AI should be more willing to activate support equipments, and less willing to activate supplements and transportation.

Increased building weight for whaling and logging for AI with Agricultural Investments.

Made AI with Plantation Economy strategy wants to produce and export rubber.

Merged goods stances from industrialization strategy into the default strategy.

Limited all "wants_high_supply" goods stances to countries with over 1 million population.

Increased AI weight for abolishing Land Based Taxation from 100 to 400. The trigger is tweaked a little bit and graduated taxation is excluded for this purpose.

Changed AI weight for abolishing Serfdom and Traditionalism to be a constant of 400 rather than 1000 when a movement supports it.

Added AI weight for establishing schools.

Reverted the valuation of construction. This change was made redundant by other changes to AI's PM valuation.

Changed critical threshold for innovation to 0.