Skip to main content

How Many Friends Do We Really Need?! - ETech on Social Software

Posted by sspielman on February 11, 2004 at 9:56 PM PST

This place is a riot. I met three people today that said, “Well I was only coming for Monday, but this is so cool I decided to stay for Tuesday. Now it’s Wednesday, and I think I’ll stay till tomorrow.” That’s the kind of energy going on around here at ETech.

While it’s hard to admit, the Redmond machine has some pretty interesting stuff going on in the labs. As Marc Smith, from Microsoft research labs, noted in his keynote he was out on work release to talk about catalyzing collective action on the net. Actually he’s a research sociologist looking into how social cyberspace is an emerging environment that resembles a biological system. This of course is not referring to all (or at least most) of our desks that have any number of coffee cups growing something in it. We’ve seen communities, group(ware), networks, and worlds evolve, but what he’s talking about is a whole other class of systems.

Networks are coming together in ad-hoc way to get things done, but they aren’t really groups because they are too big. More like voluntary associations. In most collective efforts, there are a minimal set of people that actually get anything done, and others are ‘minimum’ members. This all add ups to Yphrum’s law – Murphy spelled backwards - which describes systems that shouldn’t work, but sometimes do, or at least work fairly well. Marc’s group is working on figuring out the type of plumbing that’s needed in an online environment for conversation communities. So things like discovery, selection, evaluation, motivation all come into play in the social networking. Really what we want to do is pick an online group like we pick a restaurant. Walk past it, read the menu, take a look at what people are eating, evaluate if it looks good, and ask ourselves do we feel like eating that tonight? There are a whole host of factors that go into our decision without thinking about it. For example when looking at usenet, (or used to use net), some wonder if it’s still around. Take a look at part of the netscan visualization project to see that activity by adding metadata to the information. It’s definitely still around.

However, what I found extremely compelling is the Aura project. Ok, I do realize that I’m linking to an aspx page here on a Java site, but it is worth the mention and should make us think and take a look at these types of applications in the Java space. The Aura project is dealing with linking online information to physical objects using barcode tags and other sensor data. You can create an application platform that is driven by the tag data. So if you have a barcode attachment on your cell phone, you can walk past an art exhibit in a gallery and get all the information about the artist, the type of media, maybe do a search on a WAP browser or click on links that might provide further information from someone’s blog who had already been through the exhibit. The question is how do you link people to their shared objects and places? Make it simple to build collective knowledge about things. This is a very compelling area for some very interesting applications.

In other points of interest, Danny O’Brien gave a most amusing session on Life Hacks: Tech Secrets of Overprolific Alpha Geeks. He did a survey of a 100 questions to various geeks to see how they organized themselves. He noticed that we are a bare minimalist group. I guess everyone was laughing at some of his slides (or more the images of his notecards) because we all saw a piece of ourselves up on the screen. While one might think that there is some secret killer productivity tool out there, it all comes down to what we all know and love. Todo.txt. That and the fact that the alpha geeks write everything in shell scripts that replace departments at companies.

In the social mobility space, take a look at which is a opt-in location tracking. It focuses on landmarks and takes advantage of text messaging to only push content to people that can act on it. They have some interesting ideas even though it seems to be focused on only the NYC area.

Nelson Minar gave some interesting insights into what really goes on under the covers at Google. Although most of the audience questions that really dug under the covers didn’t get a straight answer (to no ones surprise), but still it was interesting to hear about a couple of the 100+ factors that go into a page ranking. We got some insight into the indexing and doc server shards and replications as well as some interesting things that can be done through one box results in Google. Stuff like doing flight and package tracking, phone numbers and street maps, VIN #’s, UPC codes, FCC, and patents. And then there’s the google calculator that can do unit conversions and the all important hex to roman numeral conversion. Try it I’m not kidding. At least the engineers still have fun there.

Other things of fun included sitting in on the FOAF (friend of a friend) project. This is a project were people related to each other by relationships are described in a machine readable format. It’s basically a RDF markup for a persons inbox, homepage, workplace, etc. You can then link to other FOAF files. It’s interesting because FOAF is using RDF/XML documents, and has a vocabulary of over 50 terms. The FOAF files can then be used in other RDF files.

Tomorrow ETech wraps up, but there are some really interesting sessions all the way to the end of the show. I’ll get to as many as I can before I have to catch my plane, but check back for tomorrow’s wrap-up.

Related Topics >>