I will first state some general arguments, and then state what happens in my playthrough. It's my first full playthrough, but not my first time for humankind as I have played an incomplete playthrough before and also participated in lucy open. I did not have time when victor opendev is online. Plus, I'm a veteran player for civ5 and civ6, skilled in both city managements and combats.

For the general arguments, the first is always about AI. like many other similar games, (perhaps in high difficulties) AIs are always fucking annoying. If you are weaker to AI, then you are doomed; you can not win the game while being weak all the way. If you are more powerful then AI (in humankind, there really isn't anything that characterizes "powerful in only one aspect". If you are powerful then usually you are powerful all around), then it's impossible to maintain your relationship with AI unless you have already crashed his armies or made him vassal. There are 4 types of crisis that often occur in my playthrough: 1. cultural: a powerful empire has lots of emblematic districts, which yields lots influence and converts ai cities into your culture, which cause them grievance; 2. religious: a powerful empire has lots of territory & population, which inevitably generates lots of faith and starts converting ai cities, even you build zero holy site; 3. territory: AI will never hesitate to claim territory right next to your door step, while when ever you do so you incur grievance and they demands your territory; 4. Military: all AI leaders are likely to attack you on neutral grounds and there are some AI leaders (Beowolf I believe) always do so. Since the current game the attacker has the first move, as long as either side has heavy cavalry or ranged unit, the attacker always has the tactical advantage, which means it's not possible for you to successfully defend every time without attacking AI.

All the above crisis will lead to AI making demands on you. The demands will suspend trade (which is perhaps the only value of AI to human player), which you should either reject or accept. Some demands you can never accept, like giving territory; and for me I never accept any demands. This leads to unavoidable deterioration of relationship with AI. As long as your relationship starts getting bad, it will just become even worse unless you decide you never move any of your units into neutral territory. An AI with bad relation will attack human player on sight, and as I said before, sometimes you have to attack to survive. This circle continues and at some point, the only difference between war and peace is the place of battle field: in peace you fight at neutral territory while at war you fight at the whole map.

The second general argument is about the penalty for war. Surely Humankind has put lots of restrictions on the spoils of war you can get, but again compared to a fucking annoying AI who will attack any of your units exploring the map, that penalty is nothing and can be easily handled. There are two major restricts, the first is the force surrender system, which forces you(winner) to end the war when AI (loser) lots all his war support (I have discussed in this thread: 

https://www.games2gether.com/amplitude-studios/humankind/forums/169-game-design/threads/44511-thoughts-on-force-surrender-i-keep-ransacking-conquered-cities-to-eliminate-ai?page=1#post-335371 )

The other is the city cap, which limits the spoils you can get from war. Both can be well handled by simply razing all the non-essential cities and replace them with an outpost right at the ransacked city center. This completely solve the problem of war support, as you do not need to ask AI for anything except the last conquered city and you should have at least 80 war score to take that city, thus eliminate the AI. This also directly solves the problem of city cap, because you just do not have many cities.

One important note is you do not really need that many cities. You only need several core cities to generate production, population (these two for units), money, science and influence. Due to the following two facts: 1.districts get powerful only when there are lots of them providing good adjacency; 2.infrastructures provide city-wide bonus, which becomes more "cheaper" if you have large cites, a city is simply a burden if it has yet develop into considerable size. That means at some early eras, it's not economic to develop too many cities. 

Many territories have attractive luxury and strategy resources, but you only need a cheap outpost to access them. The outpost serves in nearly everyway as a city: you can heal and upgrade your units at outpost; you can build harbor and easily embark your units at outpost. It's just it can not produce anything and can not defend against attackers. But usually an outpost takes 3-5 turns to ransack, at which time AI's armies can not move and will easily become your experiences of new units.

The third argument is about the great reward for expansion. The game is in crazy lack of strategic resources and nearly all units need them. Therefore in order to make a army matched by your technology, you have to either trade a lot or expand a lot. As I said before trade is very unstable; moreover, it's a serious national security problem if your war supplies depend on AI. Thus expanding seems to be the only reasonable solution.

The forth argument is about independent people. They can be of great value in the early game as mercenaries and cheap source of resources. But after that they soon lost their values. You can not upgrade mercenaries or merge them with your own troops; their units will soon be obsolete. The AI absolutely loves to assimilate your best independent neighbor even the ai is located on the far end of the map, and your trade routes are soon in danger. Therefore why not just conquer them? Independent people are very easy to conquer. Just park an army at their doorstep, wait for some units to move away then attack the city. Or simply use a unit to block the reinforcement and siege the city. Once city is taken that people will be "in decline", will not attack you (unlike ai player) and will soon disappear. You can either keep the city or raze to the ground and replace with an outpost. Conquering independent people cost you nearly nothing, while assimilate has great opportunity cost since AI can well finish the assimilation process before you.

The fifth argument is about territory control. My suggestion is, whenever you have culture and idle units, claim all territory around you or between your cities (I tend to found city on the coast and there may exist "gaps" between my cites. Not only a simple, cheap outposts provide access to resources and friendly territory for you to fight AI (healing and units upgrade), but also this prevent both AI and independent people from claiming such territory. Even you are in good relationship with both AI and independent people, keep in mind that it's well possible a violent independent people may spawn and killing your wandering units.

Now comes to my playthrough. I play on a normal map with 6 player (including me). The map has 3 continents with the existence of new continent. That being said 2 old continents each has 3 players and the new continent is empty. The land form is chaotic for better position of certain water-related districts. The difficulty is civilization. 

My pick of culture is: Harappans-Carthage-Teutons-Dutch-Siamese-Soviet. 

I finished the game (eliminate all AI player) at turn 130, which is not terribly fast since someone finishes it at turn 55. But I actually has a normal game play in the sense that I do develop my cities and explore the maps. I never have more than 30 units and I only lost one unit in battle (explained later). I do load a few autosave in the early war against the huns as I do not expect their hunnic horde to have combact strength of 31. But ever since I have Teutonic knights the game is just a flat push with a few cases of emergency reinforcements. I did not build any military units ever since I entering early modern era. The following image generally explains the result of the game:

Now I state by era.

1. Neolithic era. The goal is to get as many population while still be able to choose my desired culture. I decided to try Harappans before I start the game, and after a few reloads (as it was taken by AI) I entered early classical era at turn 9 with roughly 10 scouts (this requires a little bit luck; or precisely a little bit more sanctuary and more mammoth; deers, bears or berries are just not too rewarding; Keep in mind ever since you have 2 population, the growing is exponential as you will ransack and hunting at multiple directions, it's not really hard to achieve that). I also settled my first city on a rich river region, with city center adjacent to 4 river tiles. I also already planned my first administrative center to be the same river but on a territory with fast coast. Generally speaking, my goal is to settle on river  (as many river tiles adjacent to city center as possible) while if possible on coast territory with some land tile that has 5 adjacent coast, to build cothon in classical era.

2. Ancient era. I can tell I'm the first to choose culture at the culture screen. Then I encounter a back nomadic tribe. I immediately realize he is still stuck in the past. Having runner and numerical advantages, I belive I eliminated this black nomadic tribes many times. However I think the game does not allow an AI player to be eliminated in nomadic era, as every time I believe I eliminated him, he just respawn some nomads with full health out of nowhere. At some point I believe I should just let him develop, so I parked a total of 4-5 runners to restrict his movement until his outpost provides him with enough population. The turn he gets into ancient era, I immediately declare surprise war on him and he is eliminated in several turns, with his capital ransacked (I already has 2 cities; it exceeds city cap).

Having an AI on my continent eliminated that early is a great advantage, but the cost is I spend a lot of time waiting for him to enter ancient era and the Hittites is kind of out of control. Despite his units has +2 combat bonus against mine (1 for difficulty and 1 for Hittites bonus), I realize he settled his capital at the northern tundra area. Despite rich of production, luxury and science, it is relatively in lack of food. Being Harappans, I have reached 10 population at turn 28 while AI only has 5 population at his capital. See the screenshot of my capital below (another 2 population at my second city)

I made a simply decision: getting more and more armies. I keep producing units until I have 3 warriors and 3 archers, a force quite formidable at around turn 31-32. The AI does not go with a full stack of four units. It's usually 2-3 and hence easily killed, especially you have ranged units and the don't. The large population also helps me get my tier 1 tenet at turn 35, where I choose the tenet that grants you +10 war support whenever enemy gets defeated or retreats. Thus I nearly always have full war support and can easily go to war after peace. I researched organized warfare before turn 40 and then more units simply means decisive victories, as warriors do the kill while runners provide the rear attack and friendly units bonus.

My city development at ancient era is very simply. Build the emblematic district at every place I would like to. Some early workers plus building a lumber mill, coupled with good administrative center locations can handle the lack of production. Production 88 at turn 28 is not bad (though not comparable with Egypt or nubia), it can produce one unit each turn.

3. Classical era. I entered classical era with Carthage at turn 46. I immediately purchased some cothons in the best locations,  Since I did not build more units, the war against Hittites is longer than I expect. But I did not loose any units anyway. I finished Hittites at turn 50, when I just met Greek from another country with mercenaries of an independent people running on my continent. At first I was friendly to him but soon he assimilates my best independent people and attacks my units with no reasons. Then I load a save, hiring some mercenary of another independent people, together with my own runners, taking the independent people where Greek is friendly with. There are three independent people on my continent. Though I'm happy trading with them, learning the previous lesson, knowing Greek is friendly with them and it takes too much influence to assimilate them plus my armies has nothing to do before I'm technologically prepared for overseas warfare, I will later eliminate all these independent people, ransacking their cities and replacing them with my outposts to get their resources. I always kept a few units on my continent to handle the random spawning of independent people until I occupy all my continent (with either city or outpost) very late in the game.

The classical era is mostly peaceful except those independent people who unluckily choose to be friends with the Greeks. I focus on building cothon on every territory, even those that is still an outpost (you can purchase harbor, and thus cothon, with faith; however it won't produce yield until it is attached by a city or evolved into a city). I entered an early medieval era at turn 63 with cothon in almost every costal territory on my continent. At that time my first two cities are already urban centers. See the screenshot below (the selected city is my second city; capital Harappa is even strong). The cathage's cothon (which provides 12-15 production per territory) and 25% gold discounts, coupled with Harappans food bonus really helps to make your city strong all around. It's worthwhile to mention that Harrapans +1 food also applies to coast so your cothon will be powerful both in food and in production.

I also learnt what happens across the sea. The huns(blue) are consistently winning against Mycenaeans(brown) and is definitely the largest threat (despite they have less than half my fame). Their hunnic hordes have a crazy 31 combat bonus. All their civs are geared towards military. Every of their units have +7 combat strength: 1 for professional army civ, 2 for homeland ideology, 1 for difficulty, 2 for narrative events (possibly bug), 1 for great lighthouse (yes, great lighthouse provide combat bonus for all units, which is definitely a bug). Their hunnic horde has +9 combat strength due to culture traits. Even medieval units will have difficulties against them. But with Tuetons it's not a problem.

4. Medieval Era and later

I choose Tuetons because I have lots of population, all following my state religion. The juicy faith will become gold and science. They emblemetic district is just so so since I have too much faith. However the unique unit Tuetonic knight is crazy. It has +6 combat strength when fighting units from other religion. Plus +3 from charge and the fact it is 2 more powerful then normal knight, it can one shot ko nearly everything standing outside the city walls (except early modern era units and pikemen). I went to war with Hunnic when I just have crossbow and great swords men. The footsoldiers are strong but easily tricked by those noisy hordes who shoot and then retreat. I got Tuetonic knights at turn 89, at which time I was already slowly conquering with my foot soldiers. The arrival of knights greatly speed up the conquest. The fast movement of knights also help reinforce any armies that are locally outnumbered by the AI and clearly AI forces around the city before my footsoldiers arrive and lay siege. Most importantly, my knights are essentially upgraded from runners and already have some levels. 

In the middle of the conquest Aztec (Huns choose aztec for medieval; why not choose mongolia? I have no idea) lost all his war support and force surrender was trigger. However, since I claim the territory on AI's continents with outposts, his units can still wandering on my outposts, where I have vision. I quickly catch him up with my knights, surround him and win him multiple time in the same turn, directly increasing my war support to 100 and declared war again.

I'd say the time all my army landed on AI's continent with full medieval warfare, the game is already win, since I have both numerical and technological advantage. I later upgraded crossbowmen into firearm units but that's not essential. AI is all stuck in the past with medieval units, and my great swordsmen/knights plus my religion and civ geared towards war is enough to handle them. (+1 from professional soldier, +2 from homeland ideology, +2 from tier 4 religious tenet, plus numerical advantage nearly always grant +3 or more from friendly units). I only loose one unit, an arquebusier at a siege. The reason is: every attack, no matter how weak it is, will deal 5-25 damage with an expectation of 15. AI has a particular hate towards gunpower units and all his peasants just suicide on my arquebusier. He has 6 peasants and several hunnic hordes, as long as one peasant suicide, the attack position is empty and another peasant can come and suicide again, resulting multiple melee attack in the same turn. This death seems just a bit lucky an unavoidable, but except that the war against all other three AIs are complete victory.

I choose Dutch for early modern because their great ability of generating money. Their emblematic district gets +20 gold for each adjacent harbor. Noting cothon replaces harbor only when you are Carthage. So after classical era I can again build harbor, resulting one harbor and one cothon per territory and +40 gold for that emblematic district. Dutch grants me tons of gold, but I never got a chance to spend them on warfare as I never produced any military units ever since I sail to fight the hunnic. The Siamese has a unique district that can be built on coast, which synthesize extremely well with cothon and Dutch's emblematic districts and yields me even more gold. The final pick of soviet is very powerful at war, but I'm simply waiting for English's cities to get sacked and do not need any more combat strength. The last screenshot shows the yield of my second city at the end of the game (the crazy amount of food is from Angor wat; everything else is normal):

In the end, I would say the game is absolutely fun to play. The force surrender is not that annoying if you know how to pick fights with AI and ransacking all his cities. I believe in all historical strategy games, conquest is always at least on of the most profitable winning strategy, as that's just what happens in human history. An sustainable empire should never get bankrupt from war; in fact, it should get richer and richer through the spoils of war and control of more territory and resources. In Humankind there is very limited spoils of war, but the command the vast territory will greatly increase your economy (luxuries are really powerfully; AI is not always willing to trade especially in the presence of demands and grievance) and your military (strategic resources are exceptionally rare in Humankind). All these things adding up will still result in some sort of snowballing, though not as great as that in civ 6.