Why the infosec community is ahead of the curve and rationalists and nihilists

>Throwing hands up in the air and saying "it always depends" isn't a solution.

Admittedly, postmodernism isn't really inclined towards engineering problems. It does provide frameworks and lines of criticism, but I also didn't see many new ideas being deployed in that sense. This was also a point that I struggled w/ in the OP.

>I know those examples, precisely because of the probability-wielding rationalists, who point at such cases as curiosities, and show how to put numbers on them correctly.

I'd argue there's no useful way to put numbers on, say, borders. Even in cases where the geographic location of a border was definite, if I were to stand across, say, the California-Oregon border, it's not as if 50% of my body would be subject to California's laws and 50% Oregon's. Thinking of everything numerically and engineering features to support that framework seems like a Macnamara fallacy. I would also agree that some models are useful, but the emphasis on utility would in turn be problematic in a vacuum. You could reasonably predict, say, housing prices by racial makeup of a neighborhood (recalling the Boston housing data set controversy), but to do so would invite structurally racist applications. There are probably rationalist ways of de-emphasizing utility and accuracy, but again that's less my point and more to just illustrate that the biases of a model do not inherently remove from its usefulness to some end, and that critical frameworks can allow us to articulate problems. A map could just as easily center the world on Great Britain, Jerusalem, Rome, or China - to a certain extent they're equally useful and justifiable - but even with this arbitrary decision you affect how the viewer will see what the map represents.

>I still can't imagine to whom this distinction seems "obvious in aggregate". I never found it obvious.

Probably the word obvious wasn't the right term, but "definable at all" would be closer to what I intended. There _is_ an idea of what is left and what is right, but the words themselves are deployed arbitrarily. Perhaps a better example of where postmodern thought would be deployed would be something like exploring/critiquing the American right's notion of "postmodern neo-Marxism". This would be where concepts of floating signifiers and performance might be useful to explore how speech informs identity (not just vice versa).

It's true that in any attempt to orient oneself to the world, probabilities will be involved. Some of these are modelled implicitly and autonomously by your perceptual system, some are experienced as intuitions, and some are modelled explicitly as part of a problem-solving task, and some are considered reflectively.

In the problem-solving scenario, we are taking a certain model of the situation as a given, and assigning probabilities based on that. Sometimes that's good: our model is accurate and gives us things that are relatively easy to assign probabilities to. In some situations, like casino gambling, the map/territory distinction is almost nil. We still have some problems, though: not all maps are accurate, and some accurate maps don't make it easy to assign probabilities. We can still do it - we can always assign some probability - but we might not be getting very good results any more.

Changing our map requires, in Chapman's terms, changing our "stance": our idea about what the hell is going on in the situation, including our ideas about who we are in relation to it. Roughly speaking, stances are cached sets of techniques, assumptions and perspectives which make certain actions and perceptions available to us. They are associated with schools of practice, and have evolved for suitability for certain situations. We may have reason to believe that a given stance is correct for a situation, but our reason for thinking so is not that the stance gives us access to an explicit model of probabilities.

A lot of this is actually pretty obvious, and we do it all the time. The feeling of "oh, this is one of those situations, I know what to do now" is familiar to most of us. To the extent that a stance gives us a probabilistic model to work with, it is by updating our perceptual and intuitive senses of what to do and what kind of result to expect. "Meta-rationality", in Chapman's terms, is the skill of maintaining a constant awareness of one's stance, the possibility of changing it, and some well-trained intuition about when to do so.

Introspection on these stance-associated intuitions can be very valuable, but this is a reflective practice, requiring a much greater intensity of attention than merely acting on them (and this is generally not compatible with acting on them in real time, c.f "choking" in sports). It is more "computationally" (and hence metabolically) expensive, and is generally something one would do only after gaining familiarity with the practice. One cannot use a priori probabilistic reasoning to imagine what one would do in a situation if one were, say, skilled in the art of kung fu - one can only understand the implications of being good at kung fu by actually becoming good at it (see Agnes Callard's book Aspiration[1] on this point). After which point, it may or may not be good for your kung fu practice to examine it with the tools of probabilistic rationality. I imagine Chapman would argue that, for many stances and their associated practices, it isn't really worth it, but you'd have to ask him.

Of course, we can certainly describe the entire process using probabilities: my decision to treat this as a kung fu problem rather than a diplomatic problem ultimately rests on some model I have of the likely outcomes. But this is merely descriptive: I am not "shutting up and calculating" in order to make the decision, merely rationalising it after the fact.

Perhaps this does not seem particularly insightful to you, and I would agree that perhaps it should not - you may well have figured this out for yourself, and experienced rationalists may be familiar with all of it. Most of these ideas are within the grasp of thoughtful teenagers. And yet, at the same time, there really are people who act as if they believe that knowledge of probability theory will enable them to make, in near-real-time, optimal decisions even in unfamiliar domains. And most of us, if we're honest, forget just how much of a difference a change in stance can make to our perception of a situation. Many of us are naive realists who have trouble accepting that our very ideas of what is "real" in a situation is mentally constructed and dependent on our sense of what is relevant. It's not coincidental that Chapman's writings are often a re-working of certain Buddhist treatments of how our view constructs our reality.

[1]: https://academic.oup.com/book/4936

It's not even selinux specific - "everyone's a domain admin" is also a meme in security in the enterprise windows world where the permissions are actually a bit easier to manage. One of the main complaints about AWS complexity are IAM policies. There's just no way around it - this is inherently complex and a generic user does not care to learn how to deal with this. When you simplify it (like mobile phones do) on the other hand, there will be lots of people (rightfully) complaining about not being able to use the files between applications - a common complaint about iOS environment.

The point is that these generic-offtopic complaints are extremely repetitive. Maybe the first 1 or 2 (or 17) times one sees them they aren't a big deal, but after that it's bad to be overrun by them like weeds. It's not that they're wrong—rather, precisely because they have a point, they tend to get upvoted, crowding out more interesting/specific conversation. That then leads to complaints about the complaints and god help us. That's why we added that rule.

It's a tradeoff between the particular (yes the annoyance exists, yes the comment is valid—I personally agree with it) vs. the general (it's even worse to see these things over and over again, especially at the top of threads). Both are valuable, but only one can prevail and the global optimum is more important. If you really feel strongly about such formats, you should take it up with the people who are choosing to publish this way. Complaining to a link aggregator is like yelling at a photograph.

Thanks for the kind words, I think! - but... three links?

Also, I protest. The number of HN posts I "personally don't appreciate" is massively greater than the number I reply to. This is an internet forum, after all; I spend my days failing to appreciate things.