Graphing Social: Tim O’Reilly Keynote

watch the alpha geeks
– new tech moves through hackers, then entrepreneurs then platform players
examples include screen scraping and the peddle powered internet presaging data platforms and interest in alternative fuels

On Facebook (they have a new report coming out)
facebook is growing 1.14% a day
aps are growing 2% a day
87% of usage goes to 2% of aps
top 50 developers by usage looks like a more traditional long tail, but all 5K and the tail is way long
compares it to chris anderson’s research, including book sales.but facebooks long tail is essentially useless right now.
the power law is skewed, that may change, but thats the bad news.
many applications competing for the same users. dating aps have the best uptake, then messaging and chat, just for fun as a category isn’t strong.
the most successful category with active users is sports then gaming, chat, fashion, just for fun)
most active categories (what are people building) just for fun, then messaging, then gaming, then video (multiple categories, so may not be fully accurate)
aps with over 100,00 users messaging, dating, gaming, video, just for fun, (sports weaker here)
top 40- top friends, funwall, superwall, superpoke, video, x me, ilike, movies, graffiti — top aps seem to be topping out, growth slowing.

a web 2.0 refresher
the more users, the more value
building a collective database
* building on top of open source, yahoo pays people to extend
* learning from open source, wikipedia uses volunteers
* p2p sharing users build song swapping tools as a byproduct of their own self interest
* google works this way, and to some extent facebook too

key concept: harnessing collective intelligence. ajax doesn’t matter, what matters is value grows wiht userbase.
a network-effect-driven data lock-in, with accelerating returns. red-shift companies

Yahoo started with user generated content, and picked and chose best. google figured out how to automatically extract meaning from activity. They coudl automate what yahoo was doing.
page rank as true start of web 2.0
wesabe uses it too, with fan scores, recommendations, and data information being gathered and used for advice.
facebook is picking up data but you don’t have much control over it, there is not much intelligence in the data.
for example, a list of facebook invites
* knows sean is my brother
* my company directory knows I work at oreilly
* google knows I worked with Danese
* amazon knows who’s written books for me
– why should I confirm? can’t facebook learn to use databases?

How ridiculous is this? my phone company knows everyone I ever called, but my phone only knows the last ten. Phone companies suffer from churn– data could create lock in.

“are you my friend” anyone with email, phone, IM already knows who my friends are (Yahoo, are you listening???)
xobni is extracting data such as phone numbers and email, click to call, statistics on how often you communicate, let you know when you haven’t talked to someone in a while.

The Internet Operating system

the subsystems will not be devices, they will be data subsystems. facebook describes itself as a platform, it’s really a subsystem platform, not a platform yet. if you study history, a platform beats an application every time. lotus 123 to excel… wordperfect gets beat by MS word.

two types of platform
* one ring to rule them all
* small pieces loosely joined
facebook can’t do it all. hopes they will help open it up to a small pieces model
=> thoughts on the social graph read it!

questions you should be asking
* am I doing everything i can to build applications that learn form my users?
* Does my applications get better with more users, or just more busy and crowded
** consider filtering, smart filtering
* if “”data is the intel inside”:” of web 2.0, what adata do I own?
* what user facing services can I build against it?
* does my platform give me and my users control, or take it away form us?
** you have to create more value than you capture

Random thoughts about what I want form the social grpah
* I want social networks to reflect my real social network
* I want it to help me manage those contacts (how to reach them, updated status)
* I want it to manage my groups of people
** I need to put java people together, or facebook people, if I know them or not.
** people I know, people I don’t know, people I regret knowing
* I want it to recognize asymmetry in relationships
** how can I reach out to superstars in a field I don’t yet know
** I don’t want to just manage my friends. In fact, the closer they are, the less I need to manage.
* I want fine grained control over what I see and what I ignore
** some people I just want flickr feeds, other ones I want everything. I want to see this persons blogs, but not their tweets.
* I want to discover interesting people
_is Tim normal? Probably not, but good ideas here._ .. mothers maiden name no longer a good security question 😉
_I can’t recall if he had a point, except smart understanding of relationships_

facebook doesn’t fit my relationships — steve case: i sold him a company, what am I going to say, we hooked up? might be accurate.. _yes, that was a quote._
FOWA, should look at different tie describers
what do people want to say about themselves? What do I want to say about them? What if I could adjust my view of the people. How do I want to see them? could I rearrange modules to shape how I want to be updated?

jaiku has done great things, and just got acquired by google. takes idea of smart presence to mobile. your phone knows where you are. your phone should tell you if a friend is in berlin and you are going to wake them up. Or if a friend ins town, you cna ping them. _I do this with twitter, but obviously not as effective. But do I want my movements tracked?_
I’m and inventor. I because interested in long term trends because an invention has to make sense in the world in which is finished, not the world in which ist is started.” ray kurzwell
think far along the curve, think about new platforms, think about future of applications, think about taking the platform forward so we can say, wasn’t that platform quaint?


Q: criteria in companies distribution channel?
A: one of my fundamental beliefs about web 2.0 – it’s distribution, creating interfaces with your customers. The best use all channels, web facebook, etc. They want as much contact as possible. The need to understand each of those channels, and there may not be much overlap– ilike says only 4% overlap between web and facebook uses, they tend to choose. _thinking of twitter, everyone has a favored interface_ the uses is the asset, and the services you can offer to them, and you can figure out how to offer that.

Q: If Facebook will dominate, won’t they fight to keep their uses to themselves? Even if everyone wants it?
A: I’m ont sure, there are a couple answers. If you become truly domainate, no need ot share- facebook isnt there. Google is a good example. they own a lot of data they don’t share BUT they also share a lot as well. They spider the same sites as yahoo and ms. you can share and still dominate. if large graphs cooperate, say geni and facebook cooperate both sites become more valuable. There is value in openness, if you focus on building services for users, then you choose … it ultimately depends on the services and applications you build. Right now there is way more for facebook to gain by being open, as they try to crack open these deep mines of data. For now and for many years to come, all the trends say openness is good for you.

Dave McClure is useing fun movies ot intro folks. this was at the end of Tim’s talks