That's what this article highlighted the most to me. It's hard to make something that feels small and personal and quirky yet appeals to everyone. If you make something beautiful and targeted inevitably someone says "Hey! You should expand!"
I had this conversation with the owner of a tiny Japanese steakhouse in Tokyo where I had the best steak of my life. There were only like 8 tables and I was there alone so he sat down with me and we talked for a half hour. He told me how many times he had been approached to expand to other cities or the US over the 30 years they'd been open. But he didn't want the stress and he didn't think he could deliver the quality of meal if he did. Sushi and ramen places are like this in Japan too. They may be the best but they often stay small.
I think we've lost track of that in the USA and particularly Silicon Valley (or maybe we never had it.) Would we be happier if Facebook had stayed elite university only and there were 100 social sharing websites for different niches? I dunno. Probably not. But I think we're also learning that one size fits all social is not as interesting on a massive scale as we thought it would be. Path isn't doing that well either though. Maybe the best we can do is subreddits. God save us.
I think you're right about SV being focused on scaling ideas at the cost of losing the quirkiness of something small. But, rather ironically, I find that most of the smaller restaurants in SF seem to lack that special quality that can make mom-n-pop shops so great. Unlike the little udon shops you mentioned from Japan, these small businesses in SF seem to be focused purely on scraping by; I rarely find a place that seems like a hidden gem. (Maybe I'm spoiled; used to live in NY and the food options there are drastically better. Often I would stumble upon tiny restaurants that would be fantastic.)
I don't even know if I have a point to all of this, I guess I'm just saying that restaurants in San Francisco do charge high-end prices, but often offer middle-of-the-road quality.
Sub reddits are just Usenet on the web but worse, rather than anything social.
Maybe if Reddit added a profile page with some social stuff (for paying users?) they could destroy Facebook.
There are a few sub reddits where people are sharing real life information, so maybe confirmed profiles would be beneficial for them. I don't know if that just adds a bunch of cost and not much revenue.
People go to reddit to get away from reality.
The idea of subreddits applies more generally. One piece of software, one account, many communities with different foci and norms. A more social forum could do the same thing.
This may be why you are being downvoted, but:
Is of course very interesting, and probably much discussed at Reddit. Anonymity issues/pseudonyms is probably a determinant factor.
Melbourne does alright but could be better. The way liquor licenses are designed skews the market towards fewer, much larger bars.
EDIT: I wasn't really thinking about your larger point when I wrote this - but I agree with my sibling - the tradeoff is a feature, not a bug. I think Tokyo and Kyoto are remarkable cities because they are not winner-take-all markets: the amount of niches the cities support is incredible.
In contrast to SF, Melbourne is way more interesting in terms of great small bars and restaurants to find. I tell people that Melbourne is one of my favourite cities in the world to eat for the variety and authenticity.
Here's the tripadvisor for Gyu An: http://www.tripadvisor.com/ShowUserReviews-g1066444-d1677235...
And I remember the pleasure of sitting down at a cafeteria table with whatever collection of good friends had agglomerated there that evening. Maybe with one of their friends whom I did not yet know well.
Different axes. Different contexts. I enjoyed both, but generally didn't try to glue them together. (Though our centralized computing did lead to early, local social use of email, TALK, and whatnot -- amongst friends.)
Complex societies -- like complex programs, as it happens -- rely on multiple contexts, scopes, sets, etc.
This universal, "social" web can run contrary to the fundamental building blocks of our society(/ies).
After the initial rush of a dramatic expansion of some of these axes, I think people start to figure this out for themselves, even when absent from a broader discussion on the topic.
Google's search results are now almost-but-not-100% personalized, and that evokes a sense of annoyance in people.
[Disclaimer: Not the OP]
I'm not entirely sure what you meant to say in that last sentence, but I know you didn't mean to say "uncanny valley". Stuck in limbo, perhaps? (whatever that would mean)
Some people find this useful. Other people find it creepy.
Google is really good at universal systems that produce quantitatively better results when the user input is minimal and the output is easily optimized.
You could argue my restaurant example is an input-output problem. Good ingredients in. Solid, repeatable kitchen process. Good food out. Unfortunately, food preparation is a manual process involving high levels of individual skill and has an inherently large variance. That's why running a restaurant is a people problem. The tricky output is the ambiance. Food is important but restaurants are also social entertainment. Combine those two and that's why niche restaurants are almost always better than chains (imo of course.)