History as a giant data set: how analysing the past could help save the future | Hacker News Reader

History as a giant data set: how analysing the past could help save the future

theguardian.com

24 pointsoska13 comments

Surprised the article doesn't mention Isaac Asimov's Foundation in which psychohistorians try to predict the future with mathematics, similar to Turchin's idea.

Hm, trying to predict the weather is probably much easier.

This is such a complex issue... On the most abstract level thermodynamic laws are defined in a closed space and those don't exist in nature. We would not have much history at all unless we existed in a space without divergence, tending to extremes instead of equilibrium.

History isn't an accurate account of what has happened. There is a lot of ego involved and later an ego-nation complex. History is basically an account of biases and extremes instead of a model of human behavior.

Perhapse we can make global laws to tend to equilibrium, designed with the fanciful use of AI and psychology/biology. Currently laws tend to extremes, motivated by profit.

There is history and there are patterns derived from. Biased and incorrect history will likely be an outlier or it will form a weaker pattern.

E.g.: communists win over the hearts of university students and rural folk. Then instability, coup/revolution and a controllable communist satellite is created. Is that true? Is any of it propaganda/fabrication? Which pattern is most common and can it be used to analyze Russia and China geopolitics against Ukraine ,US and UK?

Communists most definetely did not win the hearts of the rural folk in my country before coming to power with the Soviets’ help after WW2 (I live in Eastern Europe). The communists had relative success among the urban working classes and among some of the students/young folk (that was correct).

But even among students, they mostly managed to attract what were perceived back then as “non-native enough” students, that is the Jewish students mostly (because the Jewish were let’s say more than ostracized back then and the communists were one of the few organizations who received them with no questions asked), but still, the great majority of the student class was still right-wing and nationalistic.

All this to say that it is impossible to draw up generalizations (which most of them are made up based on Hollywood movies anyway).

I know a few cases where the rural class analogy is correct,including China. Any guerilla force will use that approach if the military is siding with the current government [1] , so the patter is only partially accurate,we can analyze when it is accurate and apply that lesson in the present. So,it has nothing to do with communist ideology, does that mean we can apply that to neo-nazis and other terror groups (al-qaeda is in there). This sort of analysis allows us to ask why patterns exist,the patterns are a starting point,not the conclusion.

[1] https://en.m.wikipedia.org/wiki/List_of_guerrilla_movements

The Nazis were not a terror group, they were a fully functional European government whose leader had even been democratically elected, that’s why makes them (and the idea of a Nazi State) an order of magnitude more dangerous compared to a terror group like Isis or AQ.

Back to the subject, choosing a pattern for seemingly subjective reasons (why should we prioritize the Chinese example you gave compared to the Eastern European one?) and then saying that everything else “that doesn’t fit” should only make us “adjust” the pattern and not ditch it completely (or the idea of using a pattern in the first place) is pure historicism. More generally, there are no patterns in history, there’s no “logic”, there’s no “historical spirit” (or whatever fancy term Hegel used for this), there are just things that happen for some reasons of which we haven’t got the slightest idea generally speaking (200 years on since the French Revolution and there still isn’t a consensus about why it happened, and we’re talking about one of the best known historic “events”).

The Nazis are not a terror group so they did not use guerilla warfare. Alqaeda was a terror group which did use guerilla warfare. NeoNazis are terror groups unlike actual Nazis so the historical pattern tells us they are more likely to use Guerilla warfare and focus on uni students and rural population (initially) if they escalate to a full blown campaign to affect change via terror. So the pattern is very useful there.

You don't chose a pattern for subjective reasons, you choose a pattern in relation to an observation you are making about current events. You look for patterns between what you see now and similar contexts in the past so that you can form a hypothesis and then a theory after appling critical thinking to it.

There most certainly are many many patterns in history, that's the whole point of the academic study if history. You learn from historical patterns and see if you can learn from the past. For example you can look at the Gin Craze in 1700's england and the prohibition era US and see what laws and measures to control the drug of alcohol failed or succeeded and why. You can then draw parallels with modern day drug war and legalization efforts so that policy makers will make an informed decison instead of just winging it. Attempts to implement a heavy tax and implicitly ban gin made things worse with improper distilation poisoning people,as did US prohibition era controls and I can also theorize the fentanyl and heroin controls of today is a historical pattern repeating itself. The Gin craze died when laws were implemented to allow only higher end establishments to trade Gin, making it affordable and good quality but not as cheap as it was before. Some countries like portugal have legalized most drugs allowing addicts to pay for it and use it so long as a trained medical professional administers it -- they report success and drop overdoses and poisonings.

Your refusal to accept and learn from historical patterns is not new. People naturally think they are special and as you can see with my substance control example, the US refuses to accept history and as a result thousands continue to die of drug abuse overdose.

There’s a thin sliver of space outside of a black hole where light is bent approximately 180 degrees. I have this fantasy that in the next 5,000 years we will build a telescope in space of sufficient resolving capability that we can peer deeply into that sliver and see Earth as it was eons years ago.

The traditional historian arguments against this really show the lack of rigor inherent in the social sciences, or alternatively, a self-interest bias of some sort. Complaining that “humans can’t be reduced to numbers” completely misses how probabilistic analysis works. Anyone with basic statistical knowledge wouldn’t dismiss this out of hand unless someone’s ox is getting gored. I agree that it is a complex system with many variables involved, but humans aren’t magical. Their social systems likely follow patterns like any other complex system.

Niall Ferguson has written an entire book on how Network Science/Graph theory concepts can be applied to historical events - The Square and the Tower. So it's not all bad. More like people are just waking up to possibilities.

Not much difference between historians and fiction writers.

Both tell a story of the world that's interesting, often insightful and totally made up.

Post newspapers you might be able to find data, removing all the BS like what we know exists in today's newspapers.... but it'd be hard.

We have just one history. There's a real danger of over-training or over-fitting the data.

Reply on news.ycombinator.com