Log in

No account? Create an account
What is google up to this time? - Arvind Narayanan's journal — LiveJournal [entries|archive|friends|userinfo]

What is google up to this time? [Oct. 9th, 2007|06:37 pm]
Arvind Narayanan
[Tags|, , , ]
[Current Mood |curiouscurious]

When brad put up his plea for opening up social network data so that everyone could live happily ever after, I assumed it was a last-gasp effort to keep livejournal from being consumed by the myspace/facebook duopoly. But apparently, Brad was hired by google shortly before that document was published. Very quietly. I had no idea. That, of course, puts a whole different spin on things.

Techcrunch randomly speculates that google is planning something really major for Nov 5, when it will announce a open social network platform that third party apps can utilize. This will be a better deal than writing a facebook app because you have total control of your app instead of making it a little widget that lives inside facebook.

That sounds great. There's only one small hitch: where are the users?

Apparently google's going to leverage orkut. Wait, what? Orkut? Seriously? Orkut is apparently competitive with facebook in terms of worldwide page views, but that means nothing. It has a 100% penetration in exactly two markets: Brazil and India (and Brazilians and Indians abroad.) Everywhere else, it's a nonstarter. In particular, orkut gets clobbered by facebook in the lucrative North American and European markets, and I don't see this changing any time soon. I mean, users couldn't give a rat's ass if a social network is open or not, they go where their friends go.

Even allowing for the fact that google is probably going to open up the API to the smaller networks, I don't see it making a difference. Who's left? Friendster? I thought they've been dying for years. Bebo? I hear they're real popular in Ireland. That's just one step away from world domination. Livejournal? Way too small. That leaves twitter, I guess. Should be fun writing web applications in 140 characters or less.

Nevertheless, it's early days yet in the world of social networks. I wouldn't at all put it past google to pull this off. In any case, we're no longer looking at the doomsday scenario of myspace killing off everybody else, and the world is a better place for that.

Meanwhile, Microsoft is feeling left out in the cold, and is trying to buy 5% of facebook for $500 mil, valuing facebook itself at a whopping $10 bil. Wow. Their IPO is going to be very interesting to watch!

[User Picture]From: mcfnord
2007-10-09 11:58 pm (UTC)
i'm pretty sure orkut has been associated with openid in recent months. all google needs is an authentication platform that they can combine with social graph features. facebook has tools for building a social graph, but openid potentially presents the capacity to ingest whole existing networks including facebook into one graph. i think that's the grand scheme. facebook's adoption and habituation are still valuable, but they might soon face real competition for control of the social graph.
(Reply) (Thread)
[User Picture]From: arvindn
2007-10-11 05:42 pm (UTC)
I agree that if google can get the facebook graph then they're totally in the game.

Dunno if openid is the way to do that though. It's too complicated to use for the average user.
(Reply) (Parent) (Thread)
[User Picture]From: mcfnord
2007-10-11 10:05 pm (UTC)
compared to what?
(Reply) (Parent) (Thread)
[User Picture]From: rfc9000
2007-10-11 04:38 pm (UTC)
Interesting article by Brad. Thanks for linking.
While it sounds a great idea from a social network startup's POV, from user's POV it doesn't seem a great idea. Reason - privacy.
I use orkut only for finding old school-days/college-days friends, facebook for my current friends, and IM for my close friends. I sure wouldn't want to sync these up, let alone let a 3rd social nw use this data!

Of course, if you can use IMBD ratings to de-anonymize netflix records, syncing different social nws without the user's content is a far more trivial task I guess :)
(Reply) (Thread)
[User Picture]From: arvindn
2007-10-11 05:37 pm (UTC)
Oh, I'm sure nobody's gonna share your data with others without your consent. Nevertheless, it's a lost battle if you want privacy because it only takes one of your friends to share their data.

You have to realize that you and I are already old farts. Current college kids live a twitter-using exhibitionist lifestyle. Privacy is a foreign concept to them. They comprise the single largest demographic of online social networks.

That's neither a good thing nor a bad thing. It's just a different choice.

I actually have a paper draft on de-anonymizing social networks. It was rejected from KDD with fairly good reviews (3 weak accept, 1 reject.) I think this would probably be a good time to dust it off and send it somewhere.
(Reply) (Parent) (Thread)
[User Picture]From: rfc9000
2007-10-11 06:37 pm (UTC)
I agree. I should stop using myself as an example if I want any kind of objective analysis :)
After all, there are so many people who will readily give their emailids and passwords to random sites (facebook friend finder, meebo, et.al.), just to increase their online-sociability quotient!

Yeah its sure a good time to dust it off.. social networks seem to be in vogue, Networking conferences nowadays are seeing so many papers which use social networks to solve routing, sybil attacks, email reliability, search, u name it!
(Reply) (Parent) (Thread)