A late start, thankfully, as there was horrible traffic this morning.

The Facebook Application Market
Tim O’Reilly (wearing a Zimbra t-shirt)

What they really do at O’Reilly: change the world by spreading information to innovators through books and conferences

Watch the alpha geeks to see where the technology is headed (e.g. pringles can wifi antennas implies ubiquitous network, screen-scraping implies web services, the pedal-powered internet implies alternative energy focus)

Social nets are all hacks working around not *really* knowing who our social network is (my phone knows, but I have to tell the app)


(Tim presents a lot of slide andvery fast … and mostly image based. Good choice)

87% of FB usage is 2% of the apps, which could be very bad news for all the excited app developers (take it with a grain of salt; it’s early in its life … could just be that you’ve got to be wicked viral and really quickly)

FB app uptake is not even a power law distribution (?) nor log-log.

Many apps cluster around the same users, competing for htem.

Top 5 categories of apps ranked by active users: sports, gaming, chat, fashion, fun (~10%)

Top 5 by number of apps in category: fun, messaging, gaming, video, chat

Top 5 with > 100K users: messaging, dating, gaming, video, fun

Top ten are predominantly is pwned by slide, RockYou.

Top apps adoption rates have slowed

A Web 2.0 Refresher

It’s all about openness, sharing, blah, blah, … and peace and love. ;]

Systems that harness network effect to get better as people use them.

Building a collective db: pay folks to do it (early Yahoo), get volunteers (Wikipedia now), architect it into the heart of the  (P2P file sharing, Google … and FB)

Harnessing collective intelligence: web2.0 companies are building collective databases whose value follows the number of participants.

Remember: the web has always been about users creating content.

Google’s breakthru was finding meaning in all that original, chaotic UGC and turn it into valuable real-time, user-facing services.

Wasabe (and Mint) can look at how you spend your money as an implicit action aka vote. [me: Interesting thing to aggregate]

What’s the equivalent of intelligence when filtering my feed? example: geni.com knows about family, company directory knows who’s working at O’Reilly, amazon knows who’s written books for O’Reilly.

The phone company knows who your friends are: who’d you call and who called you and how often and such. They won’t do anything with it (c’mon, they’re the phone company). It’d be interesting to get into that data. They’re doing it with email a little; Xobni is doing it a lot.

Web 2.0 is about the future of the Internet Operating System (data subsystems)
Location, id, time, products, media types, relationships, price, tags … what else?

Platforms beat apps every time: Lotus 1-2-3, WordPerfect, Netscape all got beaten by Excel, Word, and IE

Two types of platform: One Ring to Rule Them All (CW: Two keynotes in a row!) or Small Pieces Loosely Joined. The latter seems the robust model; no one will own it all, but we can all join in.  6A is opening the social graph (David Recordon), XFN microformat, etc.

Learn everything you can from your users. If google was a restaurant they’d be weighing every unfinished plate, adjusting portion sizes, etc.

Make your app get better in proportion with the size of your userbase, not just more busy and more crowded. FB is struggling with this a bit.

Make sure you’re getting a steady diet of data. If data is the “intel inside” of web 2.0, what data do you own? Make sure your apps are gathering interesting data.

Offer compelling user-facing services.

Be sure you offer more value than you worry about capturing. Does platform give me and my users control? Users are smart, developers are smart.

What I want from the social graph

Reflect my real social network. 5K address book entries, hundreds of FB friends, mostly folks reaching out to Tim. [me: I bet he says his phone knows who’s really his friend]

Help me manage those contacts (how to reach them, how they like to be reached, how they’re doing) including my friends, acquaintances, and even people I don’t know at all but have heard about and want to follow. The amount of info I need is inversely relational to how close we are.

Help me group people meaningfully.

Relationships are asymmetrical

Fine grained control over what to see vs. not (e.g. show me their photos and blogs, not their tweets, etc)

Discover interesting people

Look at geni.com: a special purpose social network that provides excellent reference material for both people and apps.

On FB’s relationship offerings when acknowledging a friend request with Steve Case, to whom he’d sold a company: “We hooked up? *laughs* Actually, that probably is accurate …”

Freebase/MetaWeb is doing interesting things with a structured taxonomy of info.

What does the person want to display about themselves? What of that do I wish to consume? I’d almost want to rearrange their profile to suit my purpose. [me: this is becoming a theme – profiles are for the consumer/viewer, not the person]

Jaiku was acquired by Google.

Your phone knows where you are. How often have you called someone and found out they’re in a different timezone … or in your timezone [me: gathering intelligence about who’s doing what according to where they are/were, viz. Jaiku founder/developer spending a lot of time in Silicon Valley lately]

“I’m an inventor. I became interested in long term trends because an invention has to make sesne in the world in which it is finished, not he world in which it is started.” — Ray Kurzweil

Think out to the possible future and design for it, not for now.


Q: You mentioned you study history, and Microsoft. They dominate a developer community. Are machines the new developer, or is it still people?
A: It’s still the people, though machines are doing a lot more dynamically in the mix. If you look back at the history of MS, … the real test is whether a large number of developers are able to make significant money on the platform and survive. At the end of the day MS absorbed its dev community, and that’s when it went stagnant. … whoever’s the winner of the web 2.0 wars will end up in that same spot.

Q: In regards to the criteria you’d advise for startups, on your own domain vs. on FB, etc.
A: I’m not sure it matters. One of my fundemental beliefs about web 2.0 is that no matter where it is it’s about distribution. Depending on your biz model more distribution may be better. The companies on the top are using all of these well; their fundemental biz is developing relationships with their users thru as many contact points as possible. (Incidentally, people seem to be choosing either web or FB; only about 5% overlap according to someone from iLike). App should be focused on the user and interaction with them. [Be mindful of what data you get vs. waht FB gets]

Q: You seem to be an advocate for the open social graph . […] Isn’t FB going to fight tooth and nail to resist their users moving out into the social graph [difusing their lead]?
A: There’s a change in the dynamics. Google’s a good example: they own some private data, but have the same public website info as any other search engine. You can share an awful lot and still differentiate. If Geni and FB were interoperating, FB could know a lot more useful things. There’s a lot of value in this openness; you’re actually getting more than you give up in most cases. FB has more to gain right now by the openness; it’ll enable them to mine more and more datas. All the trends say that openness is good for you.

Either I’m running slower this morning or Tim talks really fast. Maybe both. Really good keynote, and worth seeing Tim talk. Meanwhile, I find that I miss things in the presentation when I live blog so I’m going to ratchet down the level of detail. Apologies in advance.

blog comments powered by Disqus


09 October 2007


facebook graphing.social social networks