Only mentioning this because I personally didn't understand the value of Twitter for a long time and didn't understand why people used it, but now I think I do.
I've had very little difficulty contacting, i.e. academics whose papers I've read, or programmers who have written blog posts.
You do miss some randos interjecting into the conversation, but these interjections detract from the conversation more often than they contribute. When I'm conversing with Joe Armstrong I'm not interested in the interjections of some rando who did an Erlang tutorial once and feels that qualifies him to have opinions.
You also can learn unknown unknowns from things that are shared/discussed.
Direct email contact is great, but has a higher threshold to engage in. I think it’s different than twitter, it compliments it - but I think it’s also a lot less frequent because of that higher threshold.
Each with their own search, usually!
I can see the downsides to this, but in trade you're getting a bubble that protects you from a lot of nonsense.
> For FB if there was something that allowed event planning, but didn't require people to create accounts for it (basically event planning with RSVPs, but handled through email or text).
That sounds like a fairly small project, and I can see the value. Maybe I'll code it up in the near future.
> For Twitter - there's a lot of value if you curate people you're interested in and engage with their tweets. People you would otherwise have no access to and interesting discussions/links/discovery happen there.
> Twitter can be great, but requires some effort. I think this would be helped by 'index lists' of users that are interesting to follow for new people (though this would cause a similar bubble issue that financial indexes have) but at the moment looking at someone you find interesting for who they follow is a decent way to find people.
I'll concede this, but in my brief Twitter experience it was very easy for the noise to drown out the signal, and I didn't feel it was worth it to work around that problem. It doesn't solve a data problem I have: my data problem isn't that I don't have enough data sources, it's that I have too many, and too low quality.