If you were to design AI that can predict the rise and fall of countries, how would you go about it?
If you were to design AI that can predict the rise and fall of countries, how would you go about it?
You can’t. I see two immediate problems:
- In the world today, countries do not really fail. Instead, they become national zombies. When countries start to falter, other countries or organizations step in to support the country’s citizens—often effectively keeping the country on life support. Therefore, I think you mean civilizations/societies:
- For example, by most measures, Venezuela has failed. However, in name, it persists;
- Whereas, collapsed civilizations/societies are completely gone, sometimes leaving little trace (Societal Collapse from Wikipedia) –
- Western Roman Empire (surviving only in a minor part as the Roman Catholic Church);
- 1170 BC Eastern Mediterranean Region Late Bronze Age Collapse;
- Akkadian Empire;
- Hittites;
- Mycenaean Greece;
- Neo-Assyrian Empire;
- … some others we know about, and likely many others we don’t (collapsed civilizations do not leave good records);
- It is not possible to train AI (or more precisely a “machine learning” system) without data –
- Issue 1– When civilizations fail, they usually take the data with them (for an interesting reason);
- Issue 2 – Successful countries within modern civilization support failing countries, so we get no new data there.
SIDE NOTE 1: I believe, as an alternative, an AI programmer could substitute corporations for societies to train a possible system (see discussion below). Corporations don’t tend to save other corporations.
Without substantial data, human reason is required.
Here’s why:
- Reason is model-based prediction—only small amounts of data are required in order to DISPROVE, if possible, any proposed models or hypotheses. Bias is not possible in reason, but the reasoner cannot be sure that they have the absolutely correct answer. There is always the unknown theory.
- Machine learning is data-based prediction—That is, large amounts of data are required to CREATE accurate algorithms. Unfortunately, the programmer can introduce bias (i.e. error) by training with incomplete, wrong, or misleading data.
My hypothesis/model for civilizational collapse:
Human reason creates civilization (problem-solving). Irrationality (or the inability to problem solve) and authoritarianism puts it at risk. And, problems end it.
From Wikipedia:
Cognitive decline and loss of creativity
The anthropologist Joseph Tainter theorized that collapsed societies essentially exhausted their own designs and were unable to adapt to natural diminishing returns for what they knew as their method of survival.[82] The philosopher Oswald Spengler argued that a civilization in its “winter” would see a disinclination for abstract thinking.[2] The psychologists David Rand and Jonathan Cohen theorized that people switch between two broad modes of thinking. The first is fast and automatic but rigid, and the second is slow and analytical but more flexible.
Here’s what rigidity and authoritarianism look like in political leaders:
Leaders think that they are helping with the above behavior. Instead, leaders are producing dependent and thoughtless citizens. For example, the Eastern Mediterranean Region Late Bronze Age civilization was comprised of a large number of palace city-states. The elite leaders of these city-states did all the thinking for the people. This worked until a difficult problem emerged. In the 12th century BC, the Eastern Mediterranean region endured a serious and extended drought. Neither the leadership of the city-states nor the people could cope. Eventually, enraged and hungry, the people tore these city-states to the ground. It happened very fast, and the city-states were never rebuilt.
However, corporations are similar. Employees do not destroy their companies in failure since employees may move to another company.
Therefore, an AI team could test the above hypothesis with corporations.
I would assess leadership team problem-solving skills and authoritarianism. I expect that the more dogmatic and controlling the leadership team becomes, the more likely the company will fail. The same is true for civilizations.
For societies, I would also assess the level of citizen participation in the governing process—i.e., how much freedom they can exercise. Free people can problem-solve. This final consideration, however, must be civilization-wide. This is because creative and productive countries can sustain some failure among the rest. When all countries become authoritarian, and their thinking is undone, civilization is ripe for collapse.
