Hello, these thoughts are after completing numerous science speed runs in the game culminating in a 58 turn science win on normal speed.  This post is not meant to be about how I did that, go to reddit if you are curious.  This is meant to be about how to balance the game to help prevent this as I feel pretty strongly that the strategies abused to accomplish this are completely unbalanced and make large portions of the game completely irrelevant.

I describe my strategy in this thread on reddit fir anyone that is curious: https://www.reddit.com/r/HumankindTheGame/comments/pr0m73/humankind_science_speed_run_turn_58_win_normal

Neolithic Era

I would make a small tweak here, mainly to help balance multi-player.  I would add a soft cap to scouts and influence by having an increasing amount of food required for every additional scout and a decreasing amount of influence earned after certain thresholds.  I think a good place to target is having 10 scouts and 100 influence by around 10 turns.

Collective Minds and Land Raiser

Both of these abilities are completely broken and unbalanced.  Converting one resource directly into another leads to strategies where you can completely ignore science and build only industry at an extremely fast rate using land raiser and then convert it all to science later in the game to science and win in just a few turns.  Both abilities need to be changed.  I have seen a number of good suggestions for collective minds.  My idea is to completely disconnect the industry from the science and make it more of a catch up mechanic:

Ancestral Minds - This city receives +50% science when researching technologies from prior eras or the ancient era and receives -25% industry.  Minimum activation time = 5 turns.

Fixed Land Raiser - Science and Money worker slots in this city function as Industry worker slots while active.  Minimum activation = 5 turns.

Tech Tree

The tech tree is currently a bit poorly designed in my opinion.  I was able to win a science game while only researching at most maybe 40% of the tech tree or less.  All of the techs to win the game are only a single prerequisite line that runs along the top of the tree.  The line leading to fusion reactor has zero prerequisites and is an auto win once you can choose to be France.  My suggestion here would be to do a complete study of the tech tree and adjust pre-requites by quite a bit.

The second thing I would do is to change the mechanics of the tech tree.  I feel like after everything else is fixed that science cultures will still be too strong because they can unlock units, nukes, everything an era ahead of everyone else.  Late game it almost only makes sense to choose science cultures for this reason or you can get locked out of patronage or fame from techs.  I would suggest removing the era cap on techs, maybe then add a penalty where researching techs 1 era ahead results in -50% science.  I would require all techs to have all prerequisites instead of only one.  I would then change the science ability to just be that science cultures do not receive a penalty for research in future eras.

Negative Influence and Money Balance

Influence

Currently if you do not care about wonders, civics, or attaching territories, you can freely go into massive negative stability by city spamming.  I had 39/7 cities at the end with -353k influence per turn.  The stability penalty is laughable.  Even if the penalty were -infinite stability, you'd have 20 turns before receiving the first signs of trouble and you can currently win far before that happens or take advantage of it for 20 turns and then just liberate to stop any negative effects.  I was able to win in ~16 turns after going negative influence and my stability never even started to drop.  The problem is that stability as a mechanic has a 20 turn delay.  It is good for controlling district spamming, but not for penalizing abusive tactics.

Proposed penalty fix

Influence balance of -1 to -1,000 = significant increase to chances of negative events

Influence balance of -1,001 to -5,000 = -25% to industry, money, science

Influence balance of -5,001 to -10,000 = -50% to industry, money, science

Influence balance of -10,000 or less = -75% to industry, money, science

Money

Money is similar but used in a different way.  Instead of spamming cities to get crazy heigh FIMS yeilds, you can instead just spam an insanely large army and have a huge money deficit.  The stability penalty would take around 40-50 turns to kick in, in which time you can conquer most of the map or at least your neighbor and then just disband part of the army.  Your army is the only thing that uses money upkeep so anyone abusing this is doing so via a large army.  The penalty should directly start affecting this immeditately.

Proposed penalty fix

Your army receives -1 combat strength per turn that your money balance is below 0.

This allows you a brief window where maybe you build a reserve of cash, so that you have a few turns before running out of money similar to how empires might have hired mercenaries.  After you stop being able to pay them though, they start losing morale in the form of lower combat strength.

Wondrous Luxuries

Currently beelining patronage is the strongest thing that you can do in the game.  The rewards are just too good for the cost and also the fact that other players are completely locked out is just completely unbalanced.  Wonderous luxuries are currently better than every world wonder except Hanging Gardens, which is a wondrous luxury.  This makes it a game of whoever gets to Patronage first wins and everyone else does not know they lost yet.

Fix 1 - Leave them powerful, but significantly increase the cost.  They should be on par or more than similar wonders.  I would also institute a system whereby they are claimed like wonders, but maybe with gold signifying the investment made to create these powerful luxury resource trade networks.

Fix 2 - Significantly weaken them or have their benefit only apply to the city they are in.  Allow each player to build 1 copy of a wondrous luxury.

Settlers

Settlers are way too cheap compared to how strong they are.  They should probably cost at least 3k to 5k production and then scale up for every settler produced.  At each new colony technology, there should be an additional cost increase.  I would also recommend limiting them to only giving infrastructure up to 2 eras behind instead of 1 era behind.

District Cost Scaling

Currently building anything except makers quarters is like shooting yourself in the foot as it will increase the cost of the next district, but not help you build the next district like a makers quarter does.  I would suggest having the cost scale off 2 separate variables.  The first could be total districts in the city as now, but have that be an extremely small cost increase.  The second variable should be number of copies of that district type.  This way if you start off building some farmers and science quarters, you don't hurt your production later when you want to add some other district types.