This infographic from the New York Times is a beautifully rich visualization of something that often gets reduced to a single statistic: the number of soldiers and security officers who died last year in Iraq violence. The stark small multiple of little people--one for every casualty--calls out the thousands of individual tragedies and shows how each died: some in firefights, some in bombings, still others by torture. The other thing that grabs you is how many of these deaths aren't even Americans--the American media tends to focus only on the 901 U.S. troops who died last year. Iraqi soldiers and police officers seem to be bearing quite a burden themselves.
Whether or not you believe the Iraqi war serves some greater good, it's hard to look at this and not be in awe at the sheer scale.
Monday, January 7
Wednesday, January 2
Google Reader's gaffe: you can't fake a social graph
Over the holiday break, Google snuck in a little "tweak" to ther RSS reader, Google Reader. The idea was simple: why not make it easy to share your favorite posts with a single click? Before, there was a little Share button that posted items to a public feed with a hidden URL. Only people you gave the URL to could see these items. But what if you could share things with your friends automatically?
Problem: who exactly are your friends? Google let Facebook eat its lunch in the social networking space, but it does have zillions of Gmail and Google Talk users. Any Gmail user you send email to gets added to your Google Talk contact list. Not all of them are actually friends. So a number of users were a bit miffed to discover their shares suddenly broadcast to all these people. People wear lots of different hats, and just like teenagers don't necessarily share their LiveJournals with their parents, you might not want to share every amusing Digg post with your grandma.
To be clear, Google doesn't appear to share Reader items with people who don't have Gmail or GTalk accounts. (It still does share things with people I know I've never chatted with.) And Google has now offered a workaround to restrict sharing, and there's a setting to remove people from this service. And to be fair, I've discovered a few interesting tidbits from my friends. But the interesting lesson here is how communication doesn't necessarily imply friendship.
Since their social network Orkut doesn't get much use (outside of Brazil at least), Google's trying to infer a social graph with their communications services. They're certainly not the first--people have tried using email to deduce social networks for ages. But email is a poor proxy for friendship.
To see why, if you've been using Facebook and Gmail for a while, try this simple experiment: Go to Facebook's Friend Finder and give it your Gmail account and password. (You trust them, don't you?) Be sure to uncheck the box "Select All Friends", then scroll through the list. Everyone you've ever emailed with Gmail who has a Facebook account is in this list. How many of them are really your friends? For me, only about 40% of the Facebook users I've emailed are actually friends--the rest are random cc's caught up in a "reply all", university staff, recruiters for jobs I no longer want, etc.
The answer, of course, just like with Facebook Beacon, is opt-in: out of all the people you've ever emailed, you might want to share things with some of them. Or perhaps framed another way: out of all the people you've ever emailed, only some of them are friends.
How many of your Gmail "friends" are Facebook friends?
Problem: who exactly are your friends? Google let Facebook eat its lunch in the social networking space, but it does have zillions of Gmail and Google Talk users. Any Gmail user you send email to gets added to your Google Talk contact list. Not all of them are actually friends. So a number of users were a bit miffed to discover their shares suddenly broadcast to all these people. People wear lots of different hats, and just like teenagers don't necessarily share their LiveJournals with their parents, you might not want to share every amusing Digg post with your grandma.
To be clear, Google doesn't appear to share Reader items with people who don't have Gmail or GTalk accounts. (It still does share things with people I know I've never chatted with.) And Google has now offered a workaround to restrict sharing, and there's a setting to remove people from this service. And to be fair, I've discovered a few interesting tidbits from my friends. But the interesting lesson here is how communication doesn't necessarily imply friendship.
Since their social network Orkut doesn't get much use (outside of Brazil at least), Google's trying to infer a social graph with their communications services. They're certainly not the first--people have tried using email to deduce social networks for ages. But email is a poor proxy for friendship.
To see why, if you've been using Facebook and Gmail for a while, try this simple experiment: Go to Facebook's Friend Finder and give it your Gmail account and password. (You trust them, don't you?) Be sure to uncheck the box "Select All Friends", then scroll through the list. Everyone you've ever emailed with Gmail who has a Facebook account is in this list. How many of them are really your friends? For me, only about 40% of the Facebook users I've emailed are actually friends--the rest are random cc's caught up in a "reply all", university staff, recruiters for jobs I no longer want, etc.
The answer, of course, just like with Facebook Beacon, is opt-in: out of all the people you've ever emailed, you might want to share things with some of them. Or perhaps framed another way: out of all the people you've ever emailed, only some of them are friends.
How many of your Gmail "friends" are Facebook friends?
Subscribe to:
Posts (Atom)