If social skills are best learned through observation, then why are they so unevenly distributed amongst people who aren't hermits?
Or rather, I think they are learned as a child the way language is learned. It's a subconscious process that occurs naturally when those skills are observed by a young child. However, with social skills at least, different people seem to plateau in their development at different places, and further observation doesn't always lead to further improvement.
Most people know someone with excellent social skills, yet that person continues to outperform them. There are several reasons this could be the case. The skilled person (S) might be using skills that affect observers who are not consciously aware of the skill and its effect. This could involve subtle actions, such as skillful use of body language, more involved psychological manipulation, etc.
In some cases, becoming aware of those strategies may help observers employ them as well. This seems to be the premise of a lot of PUA blogs, as well as social skill self-help books. However, one notable pattern I've observed in these social skill advice dispensing media is that they all instruct their audience to actually practice the actions they describe, whether it's complimenting five people a day, purposefully taking a bold stance in a meeting, or negging a hot girl. Whatever the quality of the specific advice, actual physical practice is usually emphasized. Social skills workshops and therapy seem to often be taught in with a repetitive "observe and repeat" pattern.
This makes sense, I think. Some socially awkward people are clueless about their faux pas, but many struggle with excessive self-consciousness, often "overthinking" the situation until anything they do comes out strained. A solution for this is to perform a skill so many times it moves from challenging to subconscious, or from System II to System I. Models do this, practicing their best smile until it's the one they make when the camera is on them.
Overall, I feel like social skills are far more procedural than observational for adults. One interesting subset of strategies involve observation and practice at the same time. Mimicking body language makes people like you more. Sort of similarly, people often become more skilled in many dimensions when they think of themselves as actors acting out the part of someone who has that skill. For example, the advice to "think like a trader" reduces loss aversion. I know people who find it easier to go into social situations as if they are a scene from a play they are acting out. I'm not sure what the long-term psychological effects are, but it does seem to help a lot in the short-term.
So, I'm skeptical that a database of exceptional examples and explanations (although fun to collect) would be useful to people who could not be more easily helped by general advice they can already find on the internet. A database of simple strategies that have been rigorously demonstrated to improve social skills might be worthwhile.
I'm guessing that the best way to learn would be through some sort of software that could converse with you and provide immediate, specific feedback about one's deviations from socially normal or optimal behavior, although such software might well lead to excessive stylistic conformity. Another option might be one that led an individual to local maxima give zir current style. I know this is sort of on the way.
Overall, I feel like social skills are far more procedural than observational for adults. One interesting subset of strategies involve observation and practice at the same time. Mimicking body language makes people like you more.
Often that's true. Mimicking body language is about imposing intimicy with the other person. It's possible that the other person doesn't want that intimicy and then you can trigger resistance.
It's like physical intimicy. There are studies that suggest a woman is more likely to say yes when asked to dance when the person doing th...
There was a post (which I unfortunately couldn't locate) that argued that rationalists should aspire to more - that a successful rationalist should be able to master other skills too that make it obvious that being a "rationality master" is something to aspire to.
Several skills such as writing, speech craft and social skills are different from more procedural skills in that they are best learned by observational learning. Strong logic skills don't provide the same advantages here that they do in maths or science - in fact an intermediate level of logic is often a disadvantage as it leads to expecting other people to behave logically.
So if we wish to develop these skills as a community, we would need to develop a repository of examples with notes to explain what principles each example is intended to show. There might be some disagreement as to how applicable each principle is and whether a particular decision is correct - but it would still be far, far better than what else is out there. Unfortunately most communities on the Internet (ie. reddit.com/r/socialskills) end up with huge amounts of rather general advice. This advice is helpful in the beginning, but you quickly get to a stage where you have heard all of it before many, many times. But with examples, even people who are extremely talented may get something out of it.