Jeremiah, like a lot of people, misses the point of Friendship. People aren't relegated to mere collections of topics to be parsed into hashtags or filters: they're above and beyond that. News and content discovery needs to be better removed from the whole social connection process, but it's not going to make "Friending" obsolete. Being able to follow people, not topics, is the appeal to the rest of the world outside of social media and web strategy experts. The intelligent web ought to include how to follow people we care about better, not removing them entirely from the equation. - Mark Trapp

Friending to me is saying I enjoy that person's perspective. If everyone tags using "#relatedmusic" then there is no perspective anymore and you are just getting a big dump of everything. The curator feature of relationships is difficult to do away with. - Todd Hoff

+1 @todd...i love that word 'curator.' people--as filters--will always matter in a so-called 'social' context. - .LAG liked that

mark: i agree w/ your perspective but i saw the post as stating that a better way to discover and auto-follow "friends" would be possible via this systemic approach to topics/themes/memes/interests when some of these capabilities mature - one of the potentials of semantic web right? - mike "glemak" dunn

Mike: but friends aren't merely collections of topics. Better filtering and content discovery is one part of the solution, but easier ways to figure out what your existing connections are doing are just as important. I go back to an example I used a while ago: I may have no interest in car talk, but I want to know when my friend gets a new car, and I want to hear what he has to say about it. Merely reducing it to topical analysis means I miss a big event in my friend's life. - Mark Trapp

love the idea of curator. Works especially well with faving photos that show up in FF. - Thomas Hawk

mark: totally agree, not ruling out presence aggregation (i think ff already works really well for this & the primary reason i like it so much) - saw this post as augmenting an already existing social net - btw, i don't think the current system of manually finding/following is broken - but that's just me ;) - mike "glemak" dunn

A lot of folks missed the point of the whole post. This isn't about 'friending' that's just but once example. The bigger point is = the web is going to be a sentient being. - Jeremiah Owyang

Jeremiah: you wrote about how the act of friending would be obsolete because we're teaching some nebulous system a set of rules to figure out how we determine friends. I'm saying Friendship isn't programmatic: it's not just a different form of interest profiling. Rulesets can apply to a lot of things in social media, but Friendship transcends a list of rules to be parsed. I've written about this subject here: http://marktrapp.com/tags... - Mark Trapp

Mark... spot on. I may not even want to friend the same people on different services, may use different filters on different services for the same friends, etc. I want this kind of thing to be manual. I want to retain control of how my time is used and what information I see. - Tinfoil 2.0

Mark. Have you seen Xobni? It's already tracking who my contacts are by email usage without my explicitly saying that someone is a friend. This is already happening. - Jeremiah Owyang

agreed jeremiah - web as sentient being is a stretch (which i'm sure is the way you meant it) but not the progress that will help to produce the semantic web which should be very revolutionary... - mike "glemak" dunn

I've not said the "S" word Mike. I'm trying to approach this 'next next' without using any buzz. Taking a pragmatic approach. But yeah, we agree. - Jeremiah Owyang

Jeremiah: your contacts are not necessarily your friends. I have a list of people, like vendors, who I have to email regularly, nevertheless, I don't care about their daily lives and I don't have a sense of real attachment to them. You're conflating two different terms here: Friendship isn't a keyword search, it's not your address book; it's more than that. - Mark Trapp

Jeremiah: I agree that there's a lot of stuff, programmatically, that could be done to figure out very useful relationship graphs: most frequently contacted, business associates, etc. But you're always going to need to explicitly state "This person is a friend." It's not reducible to a parse, and for that group of people, your actual friends and the people you care about the most, "friending" isn't going to be obsolete. - Mark Trapp

Well put, Mark. - Anthony Citrano

hashtag is a metadata. With enough metadata, you can find out relationships between people or information. Though I think the next step is finding relationship or connection without having explicit metadata. - Leon Ho

Might be instructive to take a page out of (formal) social network analysis...look at users' behavior to infer/interpolate relationship information. This has been done (for example) analyzing massive amounts of data from cell phone users' anonymized call logs. Not too hard to figure out which phone numbers belong to friends, which to family, which to Domino's, etc. Watching users' interactions can educate a "smart enough" system about the relationships (friendships and other kinds) among the users. - Andy Shaindlin

I have been (and still are) interested of how data mining allows to create maps of social connections and information they share between each other. There are huge privacy risks in that kind of things but it looks like that world is getting more and more full of all kinds of sources of data. Sooner or later people start to notice that there are tons of information about themselves even if they haven't shared anything publicly (or that is the way many people think). - Daniel Schildt

Jeremiah wrote about "Teaching the System" and while there are issues related to connecting different systems together, there will be more and more conversation about how to connect huge databases to create massive pools of data. Even if that data is located in separate locations, application interfaces allow systems to communicate between them and make distributed data mining by just transferring results to another service. - Daniel Schildt

I just find it distracting that in longer run at least some of the systems will become tools to track and control people. It's not the functionality, it's how features are used. There will be more talk about ethics of data mining but does that really change much of how things are getting to on later stages? - Daniel Schildt

People's identity in on the way to "higher level" as it's being digitized in many ways from credit card data to click and location tracking online and offline. Some people say that they don't have anything to hide. OK, it may feel like that but do they really think that in future? I don't want to be paranoid, not even close to that kind of feeling but I'm just kind of pessimistic of ways many things look they would be going to. - Daniel Schildt

In my opinion there should be more conversation of what are privacy aspects of society where information is openly traded between different systems maintained by individuals, companies and governments. Who controls the data or are things getting little bit out of hands? Or is the free flow of information best way to do things always when we are talking about amount of private and public information there are already in databases around the world? Who says the last word on how that info is used? - Daniel Schildt

So knowing that, isn't it our choice to decide how much to put out there? Also, there are sites / services that aren't indexed. Why not go that route? - just asking. - Mona Nomura

I'm not saying that it would be bad that public information gets indexed and combined. I'm just saying that in long run the difference between private and public is going to get mixed. It's our choice to decide what we publish or not but there is much information about ourselves that most people don't even know to exist. That is the main thing that makes me to be somewhat pessimistic of future. - Daniel Schildt