The biggest shortcoming in consumer applications right now: meeting people I care about

"Is there anyone I should meet near me?"

This is a squishy statement - imprecise ('near'? 'should'?) and not the type of thing we nerds excel in. But I believe this is an answerable question with technology. Some combination of rarity, closeness, extroversion, and other knowable factors help answer this question. Let me give you some examples:

• Do I care that my work acquaintance is at the bar a half mile away at 1 AM, when I've been asleep since 10pm? Probably not. (Especially if we've never talked outside of work)
• Do I care that my college roommate, who normally lives on Long Island, is about to fly out to San Francisco, 30 miles away from Palo Alto? Yes. (Especially if we've emailed or talked on Facebook or Google Chat since)
• Do I care that the woman who just walked into the bar also likes road biking, home brewing, skiing, and the Red Sox? (If I were unmarried) Unequivocally yes.

More than a few services have tried to solve items around this problem.

• On Facebook, you can search by "current city." Many users, however, don't fill this in and Facebook doesn't do radius search. It also doesn't cover weekend trips or other chance encounters.
• Foursquare has no concept of interestingness for a check-in. Users can select a radius for alerts, but there are so many check ins that it becomes too distracting - so users disable the alerts.
  
• Google Latitude got closest. For about a year, they had Latitude Alerts. Latitude kept your cell phone's location history and formed an approximate model of where you were at which time which day of the week. If you went somewhere strange for that day and time, it would alert people within an appropriate radius. For example, if I went home early one day, it would alert me of a friend in my neighborhood. If my friend flew from New York to San Francisco for the weekend, it would alert me. If they'd left it on, and if a critical mass of my friends used it, it would have been great.

None have handled friend-of-a-friend meetups, or potential compatible friends or partners. And none have handled a fine grain notification system that tailors itself to introversion level, work night partying tolerance, and ranking of contacts.

I used to check in on Foursquare religiously. Its game mechanics stimulated me, and I hoped to have chance encounters with friends. It started taking over 60 seconds to check in - geolocation takes too long, it showed irrelevant venues, and alerts were too noisy to leave on. I interviewed with them back in February for their new San Francisco office and pitched my idea - the reactions of the interviewers made it seem that they aren't interested in friends-meeting-friends.

It drives me crazy is that this information is there. As a technologist, I figure the solution is to make a new service that combines Latitude and TripIt to know where you have been, are, and where you will be, combs your email, Foursquare, Twitter, and phone logs to know who you actually care about, and make a service that works primarily with other people who use the service. But the set-up cost is high, the privacy implications are high, and I can't see enough people coming on at once make it useful - I imagine burning out early adopters as it seems that Foursquare has done.

This all came to me because I'm in New York City this week and I'm struggling to figure out who is in town, who might be in town, and where in town they are. I ended up using "current city" on Facebook serially for the five boroughs, remembering who lives locally, checking Latitude, and spamming Twitter, Facebook, and Plus to try to find those who might happen to be here for the weekend. (If you want to meet up with me before I leave, I'm hanging out at Ginger Man in the Flatiron Saturday from 12:30pm-2:30pm!)

I don't have the answers. What I do know is meeting other people, seeing the people we know and share history with, is one of the greatest things in life. Meeting more friends is also emotionally valuable. Technology can help, but it's falling flat right now.