There’s existence proofs of people clearly understanding geopolitics and using their predictive ability (and accumulated power) to direct the course: Bismarck, Hoover, etc.
From that, I think it is clear that useful understanding of geopolitics is not a superhuman ability. Maybe it’s a top 0.1% thing, but even so it is plausible that a given person writing on them does actually understand the system enough to have useful opinions.
If it's the top 0.1%, how do we distinguish them from the similar fraction of people that have been right on accident so far? I would expect that the number of lucky people is comparable to the number of skilled people if the skill is really really hard. Incidentally, this is how I think about investors. Some beat the market, but there are such a huge number of people it is nearly impossible to tell if they won by skill or by luck.
I would say that the possibility of analysis is not about the complexity of the total system, but about local wells of stability. So i would say that the fact that Country's tend to have coherent policies means that their actions can be analyzed without know everything that happens in every country
In geopolitics just as well as in all types of world modelling: A ton of things are trivial to reliably 'model' or infer (even for an uneducated dummy), and a ton are hopelessly too complex and erratic to plausibly say anything intelligible about even with the best 'modelling' or knowledge/intuition. Then there's a large mass of questions in between these two extremes where better knowledge or epistemic approaches/modelling may help you to moderately or maybe sometimes even importantly improve your predictions.
One can debate how large that subpart of the in between section is where significant extra sophistication is currently really helpful. But that geopolitical predictions per se would be impossible, feels almost trivially wrong.
In very complex systems there are usually emergent trends that make a simplified model somewhat workable, though not with a fine degree of accuracy. It's a very "spherical cows" scenario, though.
One of the things that’s always bugged me is the unstated assumption that its even possible for someone online to produce credible analysis of extremely complex systems, like geopolitics, if given enough time.
But upon closer examination it doesn’t really seem to be an assumption backed by anything, there’s no reason or force that guarantees at least 1 human being writing texts online… is in fact capable of doing that.
For example:
If we assume geopolitics is somwhere between 100x and 10,000x more complex than the typical dynamics in a typical large city, order of magnitude ballpark range. And that the effective real world gap between a very smart middle aged writer and a very smart 5 year old is also around 100x to 10,000x.
Then if we dont see any five year old’s analysis of city politics taken seriously… Not even for outlier 99.9th percentile five year olds focused on analyzing it with a large time expenditure… this would suggest we should give likewise treatment to the scaled up scenario.
Has anyone else thought about this before?