Tuesday, February 23, 2010

Augmented Reality Face Recognition

Update : Answering Ben and John's comments :

Ben: I guess the question is, now the tech. is out there, how long does it remain "opt-in". Let's suppose it was, say, Facebook that bought this technology and incorporated it into their iPhone / Android client ... ie. see which people around you have FB accounts.

Then, one day, they find that not enough people opted in, so decide to make the system opt-out instead.

John : Yes, I love Github; and the social aspect of Github is very cool. ( http://github.com/interstar )

I use it for work, and will start putting social projects up there. But you have to be careful with making too many inferences (eg. about code quality) : sometimes I check-in code written by colleagues, and they check-in my code ...

On the broader point, I'm more convinced than ever we're heading into netocracy. And one of the symptoms is when people start making inferences from the social graph itself as opposed to using just navigating the social graph and then analyzing. Ie. when people start saying "X is probably a good potential hire *because* he's connected to these people and worked on those projects" rather than "I found X via these people, now let's see if he's a good hire".

What I really think is that there's little chance for society as a whole to retreat from netocracy.

And when we arrive, there'll be no way for individuals to opt-out. We could try to avoid making our social network explicit and publishing it, but it will become increasingly difficult to function in society without doing so.

In a real netocracy, not being in a social network will be as debilitating as not having any money under capitalism.

What is much more likely though, is that smart operators will work through secret or more private social networks (much as smart operators under capitalism have Swiss bank accounts or use otherwise dark payment schemes)

However, such secretive individuals and organizations will need to break cover and become partly visible, in order to operate at all in netocracy (witness this story about a criminal gang, who court certain kinds of visibility and connections)


Ben said...

Hard to see why you would opt in to the service though. It also seems a bit antisocial. "I'll point my camera phone to find your twitter feed because I can't be bothered to ask you."

John Powers said...

I know you're working too much Phil. Links are always optional, but I'm quite curious first if you've seen this --an article in the Guardian Tech Blog about using the GitHub API to do network analysis on coders there (of course with a link to the source code).

What made me so curious to hear your take on it. What has me interested is first that you'd be in in some of the iterations of that algorithm, so it's personal for you. And second it seems to me evidence of your netocratic ideas--which is the part I'm really interested in.

Kind of OT, but we both have quite common names so a simple search of our name isn't that exciting for either of us--though you do come up high in the results. But for friends with more unusual names the name search is sometimes sort of horrifying because they had no idea how much is public about them.

This GitHub social network analysis is portends the future for the rest of us. I curious where this is leading.

John Powers said...

Geez, that story on the Koobface gang! But your point about being "partly visible" has a lot of ramifications. Just on a surface level everyone is concerned about managing different degrees of visability.

Ethan Zuckerman champions xenophilia proclaiming that homophily makes us stupid. I know that many of the online friends I treasure most I met seemingly randomly.

Pornography on the Web is a complicated subject. On a certain base level the flow of pornography within groups can provide enough traffic which enables other interactions to happen. There are probably other examples of networks people are in where the benefit or reason for being in that network isn't so apparent as it might seem, but I think the presence of porno in online networks is an example of things being not quite what they seem.

It is true that for the most part birds of a feather flock together, but many people,if not most are open to difference to some extent. Indeed most people seek out chance encounters online.

The idea of being cautious about making too many inferences extends in several directions. The trouble is that group identity is seen as: A not B. I'm not sure we immediately grok: sometimes A and sometimes B. But as our social graphs are exposed, rigid identities--or the appearance of them--are harder to maintain.

John Powers said...
This comment has been removed by the author.