Huberman on Communities of Practices and Forecasting

Posted on Published by Eugene Eric Kim

Bernardo Huberman, the director of the Information Dynamics Lab at Hewlett-Packard, gave a talk at Stanford on January 8 entitled, “Information Dynamics in the Networked World.” Huberman covered two somewhat disparate topics: Automatically discovering Communities Of Practice by analyzing email Social Networks, and a general forecasting technique based on markets.    (TX)

The first part of his talk was a summary of his recent papers. I’ve alluded to this work here many times, mostly in reference to Josh Tyler and SHOCK. The premise of the work is similar to that posed by David Gilmour in his recent Harvard Business Review article.    (TY)

Huberman described a technique for discovering clusters of social networks within an organization by analyzing email. The key innovation is an algorithm for discovering these clusters in linear time. The algorithm, inspired by circuit analysis, treats edges between nodes as resistors. By solving Kirchoff’s equations (which gives “voltage” values for each node), the algorithm determines to which cluster a node belongs.    (TZ)

The second part of Huberman’s talk was enthralling: Predicting the near-term future using markets. This is not a new idea. A very topical example of a similar effort is the Iowa Electronic Markets for predicting the outcome of presidential elections.    (U0)

The methodology Huberman described (developed by Kay-Yut Chen and, Leslie Fine) aggregates the predictions of a small group of individuals. It works in two stages. The first provides behavioral information about the participants, specifically their risk aversion. Huberman remarked that risk aversion is like a fingerprint; an individual’s level of risk aversion is generally constant. In the second stage, participants make predictions using small amounts of real money. The bets are anonymous. Those predictions are adjusted based on risk aversion levels, then aggregated.    (U1)

Huberman’s group set up a toy experiment to test the methodology. There were several marbles in an urn, and each marble was one of ten different colors. Participants were allowed to observe random draws from the urn, and then were asked to bet on the likelihood that a certain color would be drawn. In other words, they were guessing the breakdown of colors in the urn.    (U2)

Although some individuals did a good job of figuring out the general shape of the distribution curve, none came close to predicting the actual distribution. However, the aggregated prediction was almost perfect.    (U3)

The group then tried the methodology on a real-life scenario: predicting HP’s quarterly revenues. They identified a set of managers who were involved in the official forecast, and asked them to make bets. The official forecast was significantly higher than the actual numbers for that quarter. The aggregated prediction, however, was right on target. Huberman noted that the anonymity of the bets was probably the reason for the discrepancy.    (U4)

The discussion afterwards was lively. Not surprisingly, someone asked about the Policy Analysis Market, the much maligned (and subsequently axed) brainchild of John Poindexter’s Information Awareness Office. Huberman responded that the proposal was flawed, not surprisingly suggesting that this technique would have been the right way to implement the market. It was also poorly framed and, because of the vagaries of politics, will most likely never be considered in any form for the foreseeable future.    (U5)

I see many applications for this technique, and I wonder whether the Institute for the Future and similar organizations have explored it.    (U6)

As an aside, I love how multidisciplinary the work at HP’s Information Discovery Lab is. I’ve been following the lab for about a year now, and am consistently impressed by the quality and scope of the research there.    (U7)

“Low-Focus Thought” in Knowledge Management Systems

Posted on Published by Eugene Eric Kim

David Gelertner wrote an essay called “The Logic of Dreams” (a chapter in Denning and Metcalfe’s Beyond Calculation: The Next Fifty Years of Computing), where he discussed the creative process. Gelertner suggested that there are two kinds of thought: high-focus (analytical, logical) and low-focus (free association). The former we understand well (according to the Gelertner); the latter we barely comprehend.    (TI)

Low-focus thought is the story of weak ties, not just in the context of Social Networks but of ideas in general. It’s a story that is told over and over again. A poet smells a rose, and is reminded of his lover. Friedrich Kekule dreams about a snake biting its tail, wakes up, and solves the structure of benzene. Grace Hopper remembers an old play from her college basketball days and figures out a memory-efficient algorithm for her A-0 compiler.    (TJ)

Gelertner was interested in implementing low-focus thought in Artificial Intelligence software. I’m interested in facilitating low-focus thought via Knowledge Management systems.    (TK)

In the past year, my tools and processes have revealed a number of unexpected connections. For example, last August, I blogged two interesting articles about Marc Smith and Josh Tyler. The following morning, I happened to be rifling through some old articles, and discovered papers written by Smith and Tyler that I had previously archived.    (TL)

Old-fashioned tools and a little bit of karma led to these discoveries. I wanted to eliminate one of the stacks of papers on my floor, which was how I accidentally came across the Smith article. Later that morning, I was searching for an email that a friend had sent me earlier, and it just so happened that the same email contained the reference to Tyler’s paper.    (TM)

These discoveries were largely due to luck, although the fact that I keep archives in the first place and that I review them on occasion also played a role. I don’t want to oversell this point, but I don’t want to undersell it either. Many people don’t archive their email, for example. Many groups don’t archive their mailing lists, a phenomenon that baffles me. More importantly, many people never review their old notes or archives, which is about the same as not keeping them in the first place. All of that knowledge is, for all intents and purposes, lost.    (TN)

Good Knowledge Management tools facilitates the discovery of these weak connections, and make us less reliant on luck. Blogging is great, because it encourages people to link, which encourages bloggers to search through old entries — both of others and their own. This is an example of tools facilitating a pattern, and it’s one reason why blogs are a powerful Knowledge Management tool.    (TO)

I’m excited about the work we’ve done integrating blogs and Wikis using Backlinks and WikiWords, because I believe these tools will further facilitate low-focus thought, which will ultimately lead to bigger and better things.    (TP)

Blog Backlinks Enabled on PurpleWiki

Posted on Published by Eugene Eric Kim

If you view the Backlinks on any of my Wiki pages, it will now display Backlinks from both the Wiki and also this blog. For example, if you view the backlinks to “DougEngelbart”, you will see a list of all of my Wiki pages and blog entries that mention Doug.    (SG)

The beautiful thing about this feature is that it maintains context for all of the different concepts described on my Wiki. I list several Patterns on my Wiki, with some level of detail on each page. But when you look at the Backlinks to those Patterns, you see a list of all the stories where the Patterns are mentioned. I tell the stories as I have before, and the tool explicitly ties the concept to the stories that describe the context. That’s augmentation! As Chris Dent said, it “makes the universe bigger.”    (SH)

My essay, Wikis As Topic Maps, describes this phenomenon in further (and slightly more technical) detail.    (SI)

Open Source At Work    (SJ)

How this feature finally became implemented is a wonderful example of what makes Open Source so great. We’ve wanted it for a while, but didn’t have time to implement it. Last month, I started thinking more seriously about implementing the feature, because I wanted to demonstrate it to some potential clients. Unfortunately, I was swamped, and didn’t have time to do it myself.    (SK)

David Fannin to the rescue. David had installed PurpleWiki and the MovableType plugin, and liked it. However, he also wanted the Backlink feature. So, he wrote it, and contributed it back to us. Neither Chris nor I nor anyone else in the small PurpleWiki community knew David beforehand, but as you can imagine, we welcomed his contribution.    (SL)

David’s patch was just a hack. Chris had some ideas for refactoring the PurpleWiki code to better integrate this feature. So, he implemented them, and released a preview of the code. Chris’s refactoring made it very easy for me to write a similar plugin for blosxom. Suddenly, we had the feature I had been pining for.    (SM)

As an aside, I had grander plans for how to implement this feature, and those plans haven’t gone away. (See my notes on TPVortex for a preview.) The important thing is, David and Chris’s approach worked. It may not do all of the whiz bang things I eventually want it to do, but it does what I want it to do right now. More importantly, it may very well inspire others to implement some of the grander ideas. Release Early And Often is an extremely important pattern of Open Source development, because it enables collaboration, which accelerates the implementation and dissemination of ideas.    (SN)

Precedence    (SO)

Ours is not the first integrated Wiki and blog. Notable precedents include Kwiki and Bill Seitz‘s Wiki Log. These tools all had the integrated Backlinks feature before we did.    (SP)

The key difference between these tools and ours is that they require you to use a single tool. You have to use Kwiki as both your blogging tool and Wiki to get all of the features. Our approach integrates PurpleWiki with MovableType, blosxom, and conceivably any other blogging tool. This is consistent with our overall philosophy of improving interoperability between tools using Doug Engelbart‘s ideas as a unifying framework.    (SQ)

We’ve only taken baby steps so far. We plan bigger and better things. More importantly, we want to encourage other tool developers to adopt a similar approach, and to collaborate with each other to do so.    (SR)

Battling Group Think

Posted on Published by Eugene Eric Kim

Geoff Cohen asked:    (PM)

As we build different kinds of groupware/social software, what’s the role of consensus, and how powerful is it? Does software make reaching consensus easier or harder? For purely message-driven systems like email lists or USENET, consensus is much harder to reach than it would be in a real-life meeting. But once consensus is reached, breaking that consensus often brings down the flames of wrath. All of this is somehow invisibly coded in the interstices of the software architecture and human nature.    (PN)

…    (PO)

Could we architect social software that fought groupthink? Or does it just make the gravitational attraction of consensus, even flawed consensus, ever so much more irresistable?    (PP)

Seb Paquet responded:    (PQ)

I think the key to avoiding unhealthy levels of groupthink has to do with designing spaces that consistently exert pull upon outsiders (or social hackers or community straddlers), so as to keep the air fresh.    (PR)

…    (PS)

I think the blogosphere exhibits this kind of “outsider pull” much more than topic-focused forums.    (PT)

…    (PU)

But what about action? A diverse group has fewer blind spots, but on the other hand, agreement in such a group can be harder to establish, so there is a real possibility that the group will go nowhere beyond conversation. Is a core of agreed-upon ideas necessary for group action to take place? I think so. Does this mean that group action requires groupthink? Not necessarily, because some people are able to act upon ideas without believing in them so strongly they can no longer challenge them.    (PV)

Ross Mayfield added:    (PW)

He [Seb] is right that groupthink is avoided by a social network structure that allows a dynamic and diverse periphery to provide new ideas, but the core of the network needs to be tightly bound to be able to take action.    (PX)

That’s the main point of Building Sustainable Communities through Network Building by Valdis Krebs and June Holley.    (PY)

…    (PZ)

The ideal core/periphery structure affords a densely linked core and a dynamic periphery. One pattern for social software that supports this is an intimacy gradient (privacy/openness), to allow the core some privacy for backchanneling. But this requires ridiculously easy group forming, as the more hardened the space the more hard-nosed its occupants become.    (Q0)

Finally, Bill Seitz commented:    (Q1)

I think a shared mission is necessary. Whether that amounts to groupthink is a fair question.    (Q2)

There are a goldmine of ideas here, and the discussion is highly relevant to issues currently faced by the Collaboration Collaboratory. I’ll address them one at a time.    (Q3)

Group Think Versus Group Action    (Q4)

Bill’s comment points to the crux of the matter. What qualifies as Group Think or Group Action? We’ve discussed this question a lot at Blue Oxen Associates. In our upcoming research report, we draw a distinction between bounded and unbounded goals, and individual and collective goals. Generally, having shared unbounded goals is enough to constitute group alignment, but having shared bounded goals is required before you can call an effort “collaboration.”    (Q5)

The larger the group, the harder it is to define a shared, bounded goal that every group member will endorse. A good example of where this happens are elections. In the case of Howard Dean supporters, for example, the community is defined by a universally shared, bounded goal — voting Dean for president in 2004. As we’ve seen in the Dean case, having that universally shared, bounded goal was a galvanizing force for a previously unseen community of progressives in this country.    (Q6)

For large groups, I don’t think it’s necessary to have universally shared, bounded goals, although it’s nice when it happens. It’s enough to have small subgroups sharing different bounded goals, as long as they do not conflict with the unbounded goals, which must be universally shared.    (Q7)

The Intimacy Gradient Pattern    (Q8)

An aside on terminology: Intimacy Gradient is an excellent name for the phenomenon I first tried to describe in a previous blog entry, where I introduced the Think Out Loud and Whine In Private patterns. The problem I had in describing the Whine In Private pattern was that some spaces — blogs being the best example — felt like private forums, but were actually public. So people whining on their blogs are not actually Whining In Private; they just feel like they are.    (Q9)

Ross also used the term Backchannel, which I had also recently noted in my Wiki as a good name to describe this mostly private, but partially public space.    (QA)

Community Boundaries    (QB)

One of the founding principles of the Blue Oxen Collaboratories is that the products of the discussion and interaction should all be freely available to everyone. This is why the mailing list archives are publically available, even if participation is restricted to members.    (QC)

There is an Intimacy Gradient pattern involved here. There is a small barrier to entry to participate in tight-knit discussions, which makes the environment more conducive to parlor-style conversations. On the other hand, anyone can benefit from the resulting knowledge, which is our ultimate goal. Our hope is that the collaboratories act as a substrate for a much larger conversation.    (QD)

This has already begun to happen, and blogs play a key role. Bill Seitz, Chris Dent, Danny Ayers, and I have all blogged about discussions on the Collaboration Collaboratory, which expands the conversation to a larger group. The side effects include countering Group Think, as Seb suggests, and also attracting new members who want to participate more directly in the lower-level interactions. Similarly, we mention these blogs on the mailing lists, so the collaboratory members are aware of the larger conversation, thus completing the circle.    (QE)

Are there hidden costs to these Intimacy Gradients? Absolutely. Examples of blogs being read by the “wrong” audiences abound. Gregory Rawlins became a victim when he made some choice comments about another programmer’s software on a private, but publically archived list. (Sorry, Greg, but I always get a good laugh when I reread this.)    (QF)

Nevertheless, I think the benefits outweigh the downsides. I recently joined Howard Rheingold‘s Brainstorms community, and have wanted to link to some of the discussions there, but couldn’t. It’s unfortunate, because those linkages are lost, but it’s a tradeoff I understand. Finding the right balance is tricky.    (QG)

How Open Should Wikis Be?    (QH)

Our original intention with the Wikis on the Blue Oxen Collaboratories was to treat them the same as the mailing lists — restrict writing to members, but allow anyone to read the content. However, we did not configure our Wikis that way, mainly because we couldn’t — UseModWiki doesn’t have this feature — and it was low on list of things to hack. (See PurpleWiki:RoadMap.)    (QI)

Based on our experiences with this configuration and further examination of other Wikis, I’m reluctant to change this model now. One potential compromise is to require registration to write to the Wikis, but to make registration free. The difference between this and simply allowing anyone to click on “Edit This” is subtle, but significant. I’m still a bit undecided on this issue, although I seem to be leaning in favor of extreme openness. The reason for this is simply that we’ve had some interesting contributions and comments to the Wikis that probably would not have been made if there were even the slightest barriers to entry. Again, it’s a good safeguard against Group Think.    (QJ)

This issue recently cropped up again, because both the PurpleWiki and Collaboration Collaboratory Wikis were vandalized for the first time. Chris Dent discovered the act first and quickly fixed it, noting, “In a way this is sort of a good sign. Infamy is almost as good as fame….” My reaction was, “Good catch, by the way. A good sign of a healthy Wiki is how quickly the community fixes vandalism.” Notable in both of our reactions was that we simply fixed the problem and moved on, instead of rushing to implement access control.    (QK)

John Sechrest, however, suggested that access control was exactly what the Wikis needed, which led to some interesting philosophical debate about the openness of Wikis. My response to John wasn’t very deep, but it does sum up my feelings on the matter: “Wikis are successful because the cost to contribute are zero. There are downsides, but there are also upsides. Get rid of one, you also lose the other.”    (QL)