Defrag 2008 Notes – Picasso, Information Day Trading, Stowe “The Flow” Boyd


One of the most consistently provocative conferences I attended last year — my own Money:Tech 2008 aside, of course — was Eric Norlin’s Defrag conference. Oodles of interesting people, lots of great conversation and all of it aimed at one of my favorite subjects: How we cope with the information tsunami.

Paul Kedrosky, Defrag 2008 Conference

I spent two days out in Denver earlier this week at Defrag 2008 with Connectbeam. As Kedrosky notes above, the conference is dedicated to managing the increasing amount of information we’re all exposed to. Now my conference experience is limited. I’ve been to five of them, all in 2008: Gartner Portals, BEA Participate, TechCrunch50, KMWorld, Defrag.

Defrag was my favorite by far. Both for the subject matter discussed and the attendees. The conference has an intimate feel to it, but a high wattage set of attendees.

In true information overflow style, I wanted to jot down some notes from the conference.

Professor William Duggan: He’s a professor at Columbia Business School. He gave the opening keynote: “Strategic Intuition”, which is the name of his book.  Duggan talked about how studies of the brain showed that we can over-attribute people’s actions as being left-brained or right-brained. Scientists are seeing that both sides of the brain are used in tackling problems.

He then got into the meat of his session – that people innovate by assembling unrelated data from their past experience. For example, he talked about how Picasso’s style emerged. Picasso’s original paintings were not like those for which he became famous. The spark? First, meeting with Henri Matisse, and admiring his style. In that meeting, Picasso happened to become fascinated with a piece of African sculpture. In one of those “aha!” moments, Picasso combined the styles of Matisse and African folk art to create his own distinctive style. He combined two unrelated influences to create his own style.

Duggan also described how all innovation is fundamentally someone “stealing” ideas from others. In “stealing”, he means that people assemble parts of what they’re exposed to. This is opposed to imitating, which to copy something in whole. That’s not innovation.

Re-imagining the metaphors behind collaborative tools: This session examined whether we need need ways of thinking about collaboration inside the enterprise. The premise here is that we need to come up with new metaphors that drive use cases and technology design. I’ll hold off on describing most of what was said. My favorite moment was when Jay Simons of Atlassian rebutted the whole notion of re-imagining the metaphors. He said the ones we have now are fine, e.g. “the water cooler”. What we need is to stop chasing new metaphors, and execute on the ones we have.

Rich Hoeg, Honeywell: Rich is a manager in Honeywell’s corporate IT group (and a Connectbeam customer). He talked about the adoption path of social software inside Honeywell, going from a departmental implementation to much wider implementation, and how his own career path mirrored that transition. He’s also a BarCamp guy. Cool to hear an honest-to-goodness geek making changes in the enterprise world.

Yatman Lai, Cisco: Yatman discussed Cisco’s initiatives around collaboration and tying together their various enterprise 2.0 apps. I think this is something we’ll see more of as time goes along. Companies are putting in place different social software apps, but they’re still siloed. Connecting these social computing apps will become more important in the future.

Stowe “The Flow”: Stowe Boyd apparently gave quite the interesting talk. I didn’t attend it, because Connectbeam had a presentation opposite his. But from what I gather, the most memorable claim Stowe made was that there’s no such thing as attention overload. That we all can be trained to watch a constant flow of information and activities go by, and get our work done. I think there will be a segment of the population that does indeed do this. If you can swing it, you’re going to be well-positioned to be in-the-know about the latest happenings and act on them.

But in talking with various people after the presentation, there was a sense that Stowe was overestimating the general population’s ability and desire to train their minds to handle both the work they need to do for their employers, and to take in the cascade of information flowing by (e.g. Twitter, FriendFeed). Realistically, we’ll asynchronously take in information, not in constant real-time.

We’re Becoming Day Traders in Information: I heard this quote a few times, not sure who said it (maybe someone from Sxipper or Workstreamr). It’s an intriguing idea. Each unit of information has value, and that value varies by person and circumstances. Things like Twitter are the trading platform. Of course, the problem with this analogy is that actual day traders work with stocks, cattle futures, options, etc. Someone has to actually produce something. If all we do is trade in information and conversations, who’s making stuff?

Mark Koenig: Mark is an analyst with Saugatuck Technology. He gave the closing keynote for Day 1, Social Computing and the Enterprise: Closing the Gaps. What are the gaps?

  1. Social network integration
  2. Information relevance
  3. Integration with enterprise applications
  4. The culture shift

Mark also believes in the enterprise market,  externally focused social computing will grow more than internally focused. Why? Easier ROI, more of a sales orientation.

Charlene Li: Former Forrester analyst Charlene Li led off Day 2 with her presentation, Harnessing the Implicit Valkue of the Social Graph. Now running her own strategic consulting firm, Altimeter Group, Charlene focused on how future application will weave “social” into everything they do. It will be a part of the experience, not a distinct, standalone social network thing. As she says, “social networks will be like air”. She ran the gamut of technologies in this presentation. You can see some tweets from the presentation here.

One thing she said was to “prepare for the demise of the org chart”. When I see things like that, I do laugh a bit. The org chart isn’t going anywhere. Enterprises will continue to have reporting structures for the next hundred years and beyond. What will change is the siloed way in which people only work with people within their reporting structures. Tearing down those walls will be an ongoing theme inside companies.

Neeraj Mathur, Sun Micro: Neeraj talked about Sun’s internal initiatives around social computing in his session, “Building Social Capital in an Enterprise”. Sun is pretty advanced in its internal efforts. One particular element stuck with me. It the rating that each employee receives based on their participation in the Sun social software. Called Community Equity, the personal rating is built on these elements (thanks for Lawrence Liu for tweeting them):

Contribution Q + Skills Q + Participation Q + Role Q = Personal Q

Sun’s approach is an implementation of an idea that Harvard Professor Andrew McAfee put out there, Should Knowledge Workers Have Enterprise 2.0 Ratings? It’s an interesting idea – companies can gain a lot of value from social computing, why not recognize those that do it well? Of course, it’s also got potential for unintended consequences, so it needs to be monitored.

Laura “Pistachio” Fitton: Twitter-ologist Laura Fitton led a panel called “Finding Serendipitous Content Through Context”. The session covered the value of serendipity, and the ways in which it happens. The panel included executives from Aggregate Knowledge and Zemanta, as well as Carla Thompson from Guidewire.

What interested me was the notions of what serendipity really is. For example, Zemanta does text matching on your blog post to find other blog posts that are related. So there’s an element of structured search to bring related articles.

So I asked this question: Does persistent keyword search, delivered as RSS or email, count as “serendipity”? Carla’s response was , no it doesn’t. Serendipity is based on randomness. It’s an interesting topic worth a future blog post potentially.

And of course, Laura encouraged people to tweet during the session, using the hash tag #serendip. The audience tweets are a good read.

Daniela Barbosa, Dow Jones, Daniela works for Dow Jones, with coverage of their Synaptica offering. She’s also an ardent supporter of data portability, serving as Chairperson of Her session was titled Pulling the Threads on User Data. She’s a librarian by training, but she kicks butt in leading edge thinking about data portability and organization. In her presentation, she says she’s just like you. She then pops up this picture of her computer at work:


Wow – now that’s some flow. Stowe Boyd would be proud.

Wrapping up: Those are some notes from what I heard there. I couldn’t get to everything, as I had booth duties for Connectbeam. Did plenty of demos for people. And got to meet many people in real life that I have followed and talked with online. Looking forward to Defrag 2009.


See this post on FriendFeed:

Greed, Competition, Investor Expectations: Some Things Social Media Will Never Change

Courtesy Bryan Maleszyk on Flickr

Courtesy Bryan Maleszyk on Flickr

In my previous post, I wrote about the Paul Kedrosky session at Defrag 2008, Around the Horn. It was a free form session in which he queried several panelists on a range of subjects. Lots of good discussions from that.

One topic that got some extended discussion both on the panel and in audience questions was this:

Could social media and better information awareness tools have prevented the financial meltdown?

The basis of the question, in the Defrag context, was that there were signs and data that pointed to the implosion. The argument on the table was that there was a failure of information systems, and of social media, to alert the world to what was happening. And the follow-up: how can we improve this?

You can’t. Don’t even bother.

Because the problem isn’t one of not seeing the warning signs. See Morgan Stanley analyst Mary Meeker’s slides about the various financial ratios at the time of the financial collapse. We were clearly running things in the red zone when you see that data. And the data was there to see.

The problem is that people will never change. They will ignore any system telling them that things might be getting out of hand. Why? Tragedy of the commons.

Financial Meltdown = Tragedy of the Commons

In economics and game theory, there is the notion of the commons dilemma. This is the idea that when there is a common good, people will act upon their own interests in consuming that common good:

  • Person 1: consumes proportional share of large resource
  • Person 2: takes an outsized portion of resource, which by itself doesn’t destroy the resource
  • Person 3: sees Person 2 take larger share, matches that or even increases amount consumed
  • And on and on…

The problem is that as people do this, they are not acting as stewards of the common good. The result is that the common good ends up entirely consumed as each person acted in their own self-interest.

With regard to the financial crisis, what was the common good that was over consumed? Home ownership. The grpahic below is from Mary Meeker’s presentation:


Consider the dotted line on the above graph to be the normal consumption rate for home ownership. From 2000 onward, an increasing number of people purchased their own homes. Turns out, this put a mighty strain on the financial system. A lot of people purchased homes who shouldn’t have.

Now I don’t blame people for wanting homes. But the effect was to drive the prices up incredibly, which caused more desire for home ownership. Home ownership is sustained by a number of factors: steady incomes, mortgage tax breaks, personal financial management, a robust collateralized mortgage market, mortgage insurance, etc. All of those are part of the common resources. They were undermined by too many people partaking in home ownership.

The Banker’s Life

So in the case of home ownership, mortgage bankers ended up destroying the various shared resources that make up the market for home ownership. Each banker acted independently to get as much as he could from the system: loosened lending rules overall, finance build-n-flip construction, push home ownership into markets with lower financial means (a.k.a. sub-prime).

I understand the mentality. Once upon a time, I was an investment banker with Bank of America, in the syndicated loan group. Syndicated loans are like bonds, with the risk spread across many lenders. Here are two examples of how things work in banking.

First, I was part of a deal team trying to win a mandate with HMO company Humana. We were trying to unseat Humana’s incumbent bank, Chase Manhattan. We went in aggressively, with some pretty cut-rate financing terms. Chase did want to lose the deal, and so they went even more aggressively. In the syndicated loan market, you need a bunch of lenders to participate. Which means you need realistic pricing to sell the deal, just like in the bond or stock markets. Turns out, in its effort to keep Humana, Chase went too low in its pricing. They couldn’t syndicate the deal. Chase got caught up in the competition to keep Humana.

Second, at a dinner with the head of our group, the subject of minding the bank’s credit position came up. As in, what was the role of the syndicated loans group in being stewards of the bank’s balance sheet? Should we use our superior market knowledge to alert the credit underwriters about risks we’re seeing, and deals from which we should walk? I was a brash young guy, and provocatively opined that we were all about winning and syndicating deals. We shouldn’t focus on the credit risks, as that was the job elsewhere in the bank. To which the head of the group replied to me, “For the sake of your career, I’m going to pretend I didn’t hear that.” He was right – we really did have a responsibility. But I was reflecting the behaviors I’d seen out there in the market.

I bring up these examples to give you a sense of what it can be like inside banking. There is money splashing around everywhere, and when everything is up-up-up, you really can’t turn off the motivations of people.

Social Media Will Never Change This

What was happening in the mortgage industry was that the motivations were all one way: make the deal! You see your colleague cranking on getting deals done, and getting recognition internally. Your compensation is based on how many mortgages you get done. Competitor banks were reporting higher revenues and earnings. If you don’t match the growth of your peers, Wall Street dings your stock. The real estate market just kept going up, up, up.

Social software wasn’t going to change these dynamics. The current financial crisis is just the latest in a string of such events. And there will be more. It’s just human nature.

Professor Andrew McAfee tweeted a musing about Enterprise 2.0 and the financial collapse. I responded with my own thoughts:

amcafee: Could E2.0 have saved Lehman and Merrill? No. 🙂

bhc3: @amcafee – I used to work in banking. E2.0 would have made the banks better at achieving their growth goals. But those goals hurt the banks.

Social software is a powerful tool for organizations to get better in terms of innovation, productivity and responsiveness. But companies are still run by humans, and we’ll never be rid of that, for both the good and the bad.


See this post on FriendFeed:

Matt Drudge Is Not Your Friend

Saw Matt Drudge post this note the other day on his site:


Drudge is bigger than ever. Sure, politics are a big driver, but his site was running in 1996, 2000 and 2004. The site’s popularity continues to grow.

I sat in on a Paul Kedrosky-moderated session today at Defrag 2008, called “Around the Horn“. An open discussion that was entertaining.

One area of discussion was the rise of social media as a source of information for people. This is something that we have seen written and discussed in many places. I’ve asked Who Is Your Information Filter? I continue to believe people will turn to trusted friends in social media for a primary source of news.

That being said, why does a one-man link-aggregation site continue to grow? After all, Matt Drudge is not your friend. Anyone got his Twitter or Facebook account info? I need to follow this guy.


See this post on FriendFeed: Gets Better with New Data…Are You Using It Yet?

Lately, I’ve been using for shortening the URLs I tweet, on the advice of Marshall Kirkpatrick at ReadWriteWeb. I started using it instead of, which had been my previous favorite.

Why? Because offers an array of useful data. Who knew that a simple URL shortener could open up so much interesting data?  I can’t believe people still use tinyurl and other services that “only” shorten URLs. The tracking of metadata around a posted URL – for free – makes really powerful.

Here’s what was offering before the latest data features…

  • Last 15 URLs: knows your last 15 shortened URLs, courtesy of a cookie.
  • Post to Twitter: Post shortened URLs from to your Twitter account
  • Archived web page: Yup, see that page anytime because there’s a cached version of it, even if the source link changes or disappears.
  • Traffic sources: See how much click action that URL got once you put it out there. And from what apps.
  • Conversations: Tracks which users on Twitter and FriendFeed put the URL out there. This is really cool, as you can see others who liked the same thing you did.
  • Browser bookmarklet: Easy way to create a shortened URL, stay on the page you’re reading.
  • Semantic metadata: According to Marshall’s July post, was going to add semantic analysis via Reuter’s OpenCalais API. Looks like it’s there. Cool to see per link, probably more interesting with a critical mass of URLs.

On October 30, announced several nice additions to their service.

  • Full referring domains: Not just the top-level domain.
  • Graph of click activity by time: The dates and times that a URL got clicked.
  • Clicks by Country: The countries of people who click on your URL. This is really fascinating.

Seriously, if you’re not using, why not?


See this post on FriendFeed: