Between Popular and Personal there is Social

Every time I see Gabe Rivera of TechMeme, I ask for the same thing -- MeMeme.  Give me TechMeme where the core index is based on who I read, about 150 people at any given time, to show me what my friends are interested in.  I used to ask this from people who make Newsreaders.  Because simply somedays you are too busy to read everything, but you want to make sure you haven't missed something big.  That's the real value I derive from TechMeme today.  But what I really want find something that is big with my friends, which in the larger blogosphere is actually something small.

Today we have two new and seriously great kinds of attention tools.  Newsreaders give us the ability to personally personalize.  Combined with persistent query feeds, you can follow the people and things you know you want to read.  Similarly, social networking services with a purpose let you aggregate the objects of your friends, be it pictures with Flickr, or posts with Vox.  Tagging then lets you pivot for social discovery, but that is digging deeper than you often have time or interest for.

TechMeme and others show us mass popularization.  Different communities help things bubble up.  In Social Software, you first saw this with Blogdex and DayPop.  What in the blogosophere has the most attention within a given time period.  Now we have Digg, del.icio.us/popular, Reedit, Netscape, Technorati, YouTube, Dabble, Last.fm, Flickr Interestingness and a gazillion other increasingly rich examples.  This is a Wisdom of Crowds we couldn't gain before for discoverable knowledge.

As an aside, I wonder how original Slashdotters feel about Diggers' favor for a popular  answer rather than a leading question.

So more concisely, what I hope develops:
* Tools that let me personally personalize should give me just one more degree of interestingness and popularization. 
* Tools of mass popularization should give me social popularization

Since Flickr has both kinds of attention tools, let me give specific suggestions for extension.  For within My Contacts's Photos, show me the most viewed, favorited and commented by my contacts.  Then show me the most viewed, favorited and commented pictures by my contacts in Everyone's Photos.

Now, this is just one user's greedy suggestion, and there serious usability and algorithmic challenges to overcome.  But what I'm getting at is part of the future of media.

The other night I watched the evening local news broadcast for the first time in a while.  Its funny how local news attempts to localize national news.  The idea is that if they show you a Mom in the Bay Area of a Soldier in Iraq, you can relate to that and it brings the story home.  But unless the story originates from that Mom or Soldier, it is just an overlay with too much of a contextual shift.  Similarly, when an item of local news is made national -- it is too shallow for our local tastes and we are attracted to it simply because or fair city is made popular. 

I empathize with the expert editors behind these mass media and their attempts to connect the interesting for me, when me is lost in a demographic.  But I've had a taste of going direct.  When I carry the burden of discovery, and float around YouTube's popular and related clips, I can compose a broadcast for myself.  The outstanding political commentary, funny stuff and best soccer highlights from around the world.

But after a long day of work, I'm tired, and want the network to work for me.  Cue up not what is popular, or what the people I subscribed to produced.  Cue up what my social network has found interesting.  At any given time it may be local, national, international, topical or mundane.  Of course, in the process of actively consuming it, I'll leave behind breadcrumbs of attention to make it better for my friends.

UPDATE: I've been ranting about this for years.  Sam Ruby hacked together a nifty MeMeme and the result shows a clear and simple foci of attention (a post by Spoksky at last glance), FeedDemon has something in the works and Tailrank has something close.

Comments

[Kevin Burton]

Hey Ross.

You've been able to do that with Tailrank for around six months now :)

You can even add the filter back into your aggregator...

I an set you up if you want.. you can also just upload your OPML file...

Let me know if it isn't exactly what you want and we'll just fix it.

[chartreuse]

Dude, you should not have posted this on a holiday weekend.

It is exactly what all the social media creators need to read.

nice.

[Ross Mayfield]

Kevin, uploaded my OPML. Of course, it seems to turn up interesting stuff. But guessing the algorythm, it is one of my subscriptions linking into a cluster of popularity determined by the mass.

What I want, being greedy as I am, is what my subscriptions (or friends in a YASNS) are buzzing around. Not the mass (someone once said, the masses are asses).

Chareuse, if I had the tools I want, perhaps I could find time to blog on tuesday for a hamburger today.

[Kevin Burton]

Ross.

Yes... actually we STILL do this but I think we had about 50% of the market want the filter approach (show me memes my community links to) vs the social clustering approach.

Intially we LITERALLY had local/social/global ranking tweaks but we found that people didn't use them...

Anyway... I think the biggest problem I have is UI... where do I add this into the site.. I'm willing to bring it back and start supporting it again if you can help me figure out this whole filter vs social ranking thing...

Kevin

[Nick Lothian]

Hi Ross,

I was working on something like this a while ago: See http://www.mackmo.com/nick/blog/tech/2006/2/15/More-personalized-meme-tracking-screenshots.html

I you've got an OPML file online somewhere I can show you what it would look like for you.

[Zoli Erdos]

How about Megite, also customizable based on your OPML?

[Paul Robinson]

Spooky. I wrote something very similar on this topic myself just a few hours ago when I started riffing about the Web 2.0 bubble:

http://vagueware.com/2006/9/4/web-20-gapingvoid-cartoon-inspired-rant-on-social-sites

Weird. We must be related or something. If so, can I spend Christmas at yours? :-)

[John Furrier]

Why I like this post is that it highlights something that I've been thinking about and working on for sometime now....that is algorithmic verses representational discovery and presentation systems/models. To me and others working in the Internet 2.0 world it's about new ways to understand user behavior that span the ranges that Ross talks about .... from mass communications to completely customized and from popular to personal.

http://podtech.wordpress.com/2006/09/04/new-social-model-between-popular-and-personal/

[Justin Mason]

Hi Ross --

This exists, at least for link-blogs and del.icio.us feeds, in the form of SpicyLinks: http://taint.org/wk/SpicyLinks . Here's the blurb:

'SpicyLinks is an automated link-blog summariser. It reads other people's link-blogs, so you don't have to, and reports the stuff that prove popular in your personal collection of sources.

It's similar to Populicious, but that really misses the point, in my opinion. I don't particularly want to know what _everyone_ is pointing at; I want to know what a selected set of trusted sources, with good taste, are pointing at.

It is partly a clone of HotLinks, but whereas that uses a static set of sources, SpicyLinks lets you specify your own.'

Give it a go -- it's open source, just a simple perl script.

[steve epstein]

do you have SMS or email notification as part of the idea ?
To be informed immediately..?

thanks..
Steve Epstein

[Eric Wahlforss]

Interesting post and thread! I've had similar ideas myself for some time now. Here are a few:
* Just like you point out in the article, there's almost always too little time for reading blogs. It'd be great to have a slider in the reader that says something like 'read for 5 min - 15 min – 30 min – 1 hour' with articles being filtered according to a social algorithm as one moves the slider.
* Interesting comments by Kevin on the algorithm+UI issues. Good question whether the app should feature advanced mail-style filters (but with options such as: 'has been bookmarked on del.icio.us (etc) more than X times', 'has been read/commented on by my friends', 'is by friend', etc, or whether the algorithm should be opaque, e.g. some smart, weighted combination of global popular+local+friends posts.
* One UI idea I like in some collaborative filters in music services is the slider 'niche - mainstream'.
* In general, it'd be nice to have a reader that is not 'live' in the sense that it does some processing/evalutation on posts before it shows them to the reader. in the UI, 'un-interesting' posts could simply be greyed out. interesting posts could be color-coded from hot to cold.

again, nice post. i'll comment here again if I get the time to do a mock or so.

[Rob]

not ME but POP ;)
http://popurls.com