You are currently browsing the monthly archive for November, 2008.
I was in the middle of a run this past weekend with a friend who works for a large tech company. He was commenting on the fact that there are going to be layoffs coming up for his company and that despite all of this the CEO might still be making in the range of 10 million dollars. We did some quick calculations, but figured that the average R&D/ General & Administrative employee probably made in the range of 100-150k. That means that the CEO was making in the range of 100x to 66x more than the average US college educated fairly experienced employee. What could you do with 66-100 employees? You could crowdsource the job of CEO…
It’s still an idea, but you’d have to put some controls around it.
A) The group would have to be sufficiently diverse. This way the CEO crowd would have more overall experience than any one man would have. Being diverse, they’d sidestep some of the groupthink issues as well.
B) It would have to be sufficiently large, but in order to pitch this to boards, you’d have to show some improvement in cost savings. Is 25 enough, 30?
C) There would have to be some thought into how to enable some forms of communication between the CEOs, and how to disable some others. My concern here is that we’d turn the CEO job into a mini-Senate/House with factions and partisanship, which destroys the “independent” nature of the crowd.
D) A couple of those CEOs might have to be in charge of maintaining the communication process, ensuring no funny business is going on. They would assemble the information and pass it up to board, or down to the operating divisions. This would be an overhead cost, but somebody’s got to do it.
I admit that this was dreamt up while in a somewhat oxygen deprived state, but I think going through why this might or might not work would be valuable into understanding what can be done in an informal open communication-based system vs. what could be done in the traditional way of doing things.
http://www.nytimes.com/2008/11/12/technology/internet/12flu.html
The New York Times has this article detailing the CDC/Google partnership to provide data on searches about the flu. This is particularly interesting to me due to some of the research I did in my Systems Engineering project work in a previous life. The point of the research was to look at what a public health group could do to poll data from hospitals. I looked at the ways hospitals stored data, and the possible ways of creating a standard database system, or translation layer to allow a large city like NYC to be able to know there was an outbreak. This type of system would be better than relying on individual hospitals to signal an outbreak, since it would be able to tell if 20 people went to 20 different hospitals with the same symptoms. However, the sheer cost and effort makes this system pretty difficult to implement. There is some interest from the homeland security side, especially in bioterrorism prevention and containment. There were some interesting thoughts on using a peer-to-peer system (the rage at the time), but I wonder if any of our new Web 2.0 tools would help in this…
What Google has done is go to the other side of the equation… They are not polling the hospitals, because that data is pretty tough to translate, and very difficult to aggregate. Plus, the timeliness of the data varies by the hospital and their practices. In this way, Google is going to the user of their search and looking for key terms. These terms and IP addresses would help them to figure out where (in the general sense) there might be a rising trend. It’s not without its faults however, as it relies on the user to know how to search for flu-like symptoms. “Flu”, or “Flu symptoms” could be easily tracked, but “muscle soreness” or “fever” have ambiguous causes. I think this could be valuable in cases where people know what they have. In cases where it requires a doctor to know, this might not work as well.
Light reading for the interested:
International Society for Disease Surveillance
CDC’s website for Syndromic Surveillance
http://profkane.wordpress.com/2008/10/31/the-parameters-of-wisdom/
I think Professor Kane’s post is thought provoking in two ways:
1. This class has created a conflict between traditional learning styles to which most students have become accustomed and a collaborative learning environment this course promotes.
2. Is popularity equal to quality?
1. I’ve been personally struggling with the style of presentation in the class, and for good reason. It’s different that what we’ve had in the past in virtually every class we’ve taken. Is different better? Is different worse? Well, in this class we rely heavily on the contributions of students in the class. We create the content, Prof. Kane creates the structure, facilitates the conversations and makes the backbone powerpoints. While in some cases it may just be a saying, in this class, we make the class good or bad, it’s up to us. In a class of relative newbies, and a Professor with expertise, we rely heavily on the class to contribute… which is much different than a normal lecture and problem set class. In adoption of this format, we’ve seen some unintended consequences. The Web 2.0 environment tries to be self-policing, but generally that requires experts who have ownership. So this raises the questions, who owns the class? Who feels responsible for the content? Is the class sufficiently large to be a “crowd”? How much can we rely on learners to be experts as well? I don’t have the answers to these questions…but it seems to be food for thought.
2. As Jerry’s post says, we reward students for popular content, which in this case is a proxy for valuable content. Jerry’s stated that the combination of presentation in class and the blog posts/weekly assignments have created an issue that he may correct. However, maybe this shows a point in the web 2.0 environment that we hadn’t really challenged in class. We’ve accepted that in general the crowd can filter the “junk” and figure out the gems.
Is that what we’re doing? Is that what the crowd does in practice?
It really depends on the purpose of the crowd. We’ve got 30-40 MBA students participating in a collaborative class in which most are taking this class primarily because it’s required. We have varying levels of participation, varying levels of commitment to the class, and varying expertise in the field… all good for our crowd experiment (diversity is not quite random…it is an MBA class). This is just hearsay, because I haven’t crunched the numbers, but I would say we as a crowd tend to vote for blog posts based on entertainment value rather than educational value (a post can be both, they are not exclusive).
This isn’t necessarily an anomaly, as entertainment is an important piece of the value of web 2.0. Why? Because we all know what’s entertaining, and in that realm we’re probably pretty homogenous. So in this case, we’re rewarded implicitly for educational posts (Jerry’s mention in class), but explicitly for entertaining posts (bonus points based on vote). Can we achieve both? Sure, but it’s easier to post a youtube video of a singing guy on the virtues of facebook.
Our class should value those posts that are both entertaining (because we don’t want to be bored to death) and educational (we’re paying for the class). My attitude tends to be more towards the educational, but I can see why entertaining stuff should be valued as well.
Recent Comments