Growing up, I played a lot of strategy games. My favorite series include Paradox games, Total War games, and others like Sid Meier’s Civilization. The settings range from ancient times to the far future, and the game mechanics are quite different from one another. Despite this, the general strategy remains similar, involving a blend of economic growth powering military growth and vice versa.
There are two essential cycles. The first involves raising an army to explore/conquer territory, harvesting resources from that new territory, then raising more armies to explore/conquer more territory. This is basically what happened with states like the Ottomans, Spanish, British, and U.S.A. The second cycle involves using money to improve technology and capital1 (technocapital for short), then using that improved technocapital to get more money than you started with. This is M → C → M’ in the classic Marxist formulation, where M is money, C is what that money buys, and M’ > M.2 Again, we see something similar play out in real life, as venture capitalists fund technology companies in return for even more money. Although both cycles are complicated historically by uneven progress across time and space, games are much more direct: rinse and repeat these faster than the rest and you should eventually win.
These cycles are complicated by the addition of one or more additional opponents, resulting in diplomatic relations which can balance out any runaway powers. But even where politics replaces outright war, there is still a race where the powers striving for victory have to constantly increase their economic and military power in order to remain competitive. The two are always linked, as the former powers the latter while the latter secures the former. Improvements in technology enhance both. Anything else like “culture points” or whatever are only pursued if useful for victory in their own right or for the player’s own amusement.3
No matter what your goals are, in order to achieve them, you need some combination of resources and control over those resources. “Control” can be interpreted more broadly than just direct force: given a balance of power between several entities with distinct interests, politics emerges as a means to influence the use of resources that are not directly controlled by you. Hence Clausewitz’s assertion that war is the continuation of politics by other means.
Strategy games illustrate how means and ends simultaneously converge and remain distinct. In theory, if you are able to achieve the means to permit them, you can achieve whichever ends you desire. But in practice, to accelerate the development of your means as much as possible, you wind up having to prioritize further development of means over the original ends you started with. That’s why no matter what your goals are in Civilization, Stellaris, or Factorio, the most competitive strategies mostly boil down to some combination of tech rush and mil rush.
This point applies far more widely than just games. AI companies are locked in a similar race, where Anthropic, OpenAI, Google, xAI, and others may have different goals for the future yet they all converge on accumulating as much means (compute, money, revenue, etc.) as possible as fast as possible over any other course of action. It may be the case that the future looks very different depending on who “wins,” but if everyone needs to prioritize means above all else to do so, the end result could be very similar. What that might be is beyond the scope of this post. Suffice to say that any strongly-held goals lend themselves to this Terminator-esque focus on seizing the means (as it were) above all else, whether those goals are in-game or irl. At least the AI companies aren’t able to mil rush yet (at least directly).
Footnotes
-
capital can be improved machinery, improved skills in humans, or some other means of improving the productive process. ↩
-
strictly speaking Marx’s C is commodities, including the means of production and labor power that get consumed in production, rather than capital in the everyday sense. ↩
-
amusement is often the goal in sandbox games where no formal “victory” conditions are present. ↩