Back late last night from Defrag, which I hope to attend again next year. It's been noted elsewhere that most speakers at the conference repeated several main themes; this seems true to me. What follows is a long list of highlights, sorted in alphabetical order, which seems as good as any.
Art helps us share experiences we don't have in common.
As long as the gaming is transparent, the community will figure it out and deal with it.
Communication has gone from e-mail, open by default, filtered through a black list--to instant messaging, closed by default, filtered through a white list.
Computers are explicit; it's digital, so it either is or is not.
Dark Webs will cease to exist. If it can't be found via search, it's not out there.
Data becomes information when you need to know it.
Don't sugar-coat what it is we do for a living. Don't dumb it down. Managers are voracious consumers of theory; they want to be able to make decent predictions.
Every new parent company understands that it can't have a traditional command-and-control relationship with a distributed sales force.
Gaming comes from structural issues around a site. Transaction costs are minimal with Digg, so it's hard to analyze past behavior. There's no construction of a person doing a thing. Richer systems will emerge, where you can see what a person has done in the past.
Half of all information is disinformation. Is the preceding statement part of that 50%?
How can I experience insight and intepretation? How about surprise?
How can we do a better job of making people feel okay that they're wrong all the time?
How do I make sure I don't miss anything?
I grew up in Detroit and I live in Oakland now. Where I'm from, "user" is a bad word.
If everybody knows what I do, what is my value?
If you try to prevent collusion, you prevent collaboration.
In the 1950s. 60s, and 70s, computers were about reducing us to facts. "Informationalization" is about classifying everything. Nowadays we do it ourselves, every time we fill out a Preferences form.
In the future, "I'm not showing you my data" will be far less important than "I'm not showing you my connections."
Information is just stuff I need to know. Knowledge is information with semantic content.
Interestingness is data mining around people's activities, and the social relationships between them and the photos. It's not a voting scheme, and not susceptible to gaming; interestingness is an organic measure that can be applied retroactively, before the feature came online. The magic word is "implicit." Even thought the average Flickr user doesn't think of himself as an editor, his progress through the system makes it better for everyone.
It's amazing that Wikipedia still works. Under every page is a flame war. It doesn't feel like a scalable system.
Kids today are confident enough that they'll be able to create a new idea that they don't need to lock away the rights to old ones.
Language consists of words, paragraphs, punctuations, and now links. Being able to make content permanent--first words, then writing--was a founding event. Links make the relationships permanent, for the first time.
Last year we were talking about breaking identity out of the walled garden. We were terribly concerned with the explicit: this is who I am. This year we're talking about breaking networks out of the walled garden, and are just starting to think about the importance of the implicit: this is who I'm connected to.
Links reveal the world in a way that matters.
Links are the opposite of information. They are impossible to understand without society.
Networking sites take off for tiny reasons about the site, not the intended architecture of the network.
Often the most loyal employees know that their company is wrong. They will treat customers well even though they may lose their jobs for doing so. We need to legitimize the position of these people. They are putting themselves at risk for no personal gain; if caught, they will be fired, but if not caught, the company will profit.
Our children need to understand the difference between potential and possibility.
Overload is a choice: you choose to be overwhelmed, or you focus. You use tools.
Poets have to decide how much to make explicit. So do we, when we make a site, or write a program.
Social software accrues value to the users, the readers, and the owner of the system. In that order.
Some decisions should not be coded.
Sometimes it's a mistake to declare your preferences up front. Life is complex!
Team blogs seem to fit better at work than individual blogs or wikis. Blogging is writing in a way that contributing to a wiki isn't. You find an authorial voice.
The for: tag is a gift. What my friends and contacts are bookmarking is incredibly powerful.
The Semantic Web is self-describing, linked data.
The Web works outstandingly well ... it enables us to do the basic thing that peace wants us to do: live with other people in difference, yet share the fundamental things that make us human.
The social graph connects people. The semantic graph connects everything.
The take from electronic crime is now higher than the drug industry. Within a remarkably short period of time, they'll be messing in our space.
Think augmentation, not automation.
To be implicit is human.
Using social software will not homogenize the ties within a network. The impact will be differentiation: things will be less alike than they were before. The world is spiky, not flat.
We focus on the explicit. It misleads us.
We're at the end of 5000 years of yang ... convincing users to embrace the ying is going to be very, very hard.
What people don't want from search is a bunch of pages. They want a visualization based on their query.
What we care about turns possibility into potential, and makes information inadequate. What's between us is what we don't say.
Where's the history of the Web? When Google updates, it ceases to exist.
Words don't carry meaning. They point towards shared understanding.
Next Defrag: 11/4/2008, in Denver again. Get those absentee ballots in early!