Conversation

A widely accepted model of collapse (Tainter) is that societies collapse when they get “too complex” in some sense. But this is an almost uselessly solipsistic tautology. How complex is too complex? Logs are interesting is that they are an external complexity signature.
7
11
Imagine you’re a poor immigrant in a too-complex society, who is barely literate in the local language. In a way you’re constructing your mental model of the society from an underbelly “log-level” view of it. And then your chain-migrating peers imitate and build on it.
1
4
For a sense of this process, if you’ve traveled abroad in a country with unfamiliar culture/language, think about how you pieced together a working understanding. That’s log-based reconstruction. If that process is too hard, the society/system is “too complex” and will fail.
1
3
This is something like a system being too complex to penetrate for a typical outsider *without* need for a policed boundary. Ironically, my writing is almost an example of this. Hard to get into unless you were with me for the growth phase before it vanished up its own... navel.
1
4
A logging process is a designed version of the natural process of complex systems producing a heat signature with an extractable information component. By information theory, all systems do this. Perfect secrets-keeping is impossible.
2
3
Even black holes produce a “log” process (google black hole information paradox/Hawking radiation). So complexity-collapse is like a black hole formation. A singularity creating event that should produce a “death shriek”
You’re unable to view this Tweet because this account owner limits who can view their Tweets. Learn more
Replying to
It’s usually an insurance policy afterthought I suppose. The one time I designed a guilty complex system, the thing got much easier when I designed my desired logging format first. But that’s because it was research software and log files were what I was using for the research.
1
1
Show replies