Thomas Lord [12.13.07 06:17 PM]

You would probably also like Foucault's writings about "actuarial practices" for their relevance, though others can more accurately point you towards them then I can off the top of my head.

Roughly speaking, it was during the 17th and 18th centuries that we developed the bulk of what counts as modern finance, modern actuarial theory, accounting, contract theory, etc.

These new techniques and their particular manner of displacement in commerce, law, the formation of the state etc. are some of the main organizing forces of how individual lives have come to be shaped. From the schoolroom and testing, to the adult's relation to the health care system, to treatment by law enforcement and on forever, our built environment (both physical and social) and in so many important details the emergent effect of those basic advances in measuring, counting, and making statistical predictions, combined with a thought about how to fit that data into commerce.

Considerable good comes of it. The trains run. There's food on the grocery store shelves, around here, mostly.

Some not so good things. The same "actuarial" world view, applied to humans, suggested a commodity. So, for example, the slave trade in its particular forms. I mean: who would have gone to such lengths were it not for economies of scale, ROI projections, available investment capital, and low overhead and transaction costs to be spent in measuring the product out in bulk?

What I see in Web 2.0 is a few things.

Technologically, it's a nifty and thorough yet, nonetheless, simple exploration of a particular "easy" design space on top of the LAMP stack and its cousins. These are fairly simple database schema and simple business logic rules -- just a particular focus on engaging users interactively, often in order to stalk them in one way or another, and/or in order to create a proprietary aggregate work from their efforts.

But, businesswise, it's a lot deeper: it's a new advance in "actuarial" technology. It's a new way to measure populations, in the interests of business processes that aim to manipulate populations.

So, for the individual, I think Web 2.0's contributions to identity, trust, reputation, presence, and privacy have been, well, not welcome at least. "Don't go around calling people names," my Mother once said to me.

It seems so needless. That "easy" space of plays is over-mined anyway. One of the best things the investment / industry leadership folks can do is find some better vein to tap, preferably one that's very supportive of individual liberty.

-t

  bex [12.13.07 06:45 PM]

Interesting dilemma... in the past I've said that the louder you speak, the less control you should have over your own reputation:

http://bexhuff.com/2007/09/spam-identity-and-scary-frickin-clowns

Similar to libel laws in the US: once you become a "public figure," the rules about libel change drastically. Perhaps in the future, your Technorati rank will determine what legal recourses you have?

  Michael H [12.14.07 05:24 AM]

Something like del.icio.us for reputation is an interesting idea. Maybe LinkedIn or a similar site will come up with something.

However, I think it would be too easy to game the system, either to trash someone's reputation or make them look like an expert when they are not. I can certainly see spammers trying to "enhance" their reputations.

  Thomas Lord [12.14.07 08:28 AM]

And, Michael, those are just the "first order" effects. You're talking about the games that arise directly among the users -- but, it's much, much deeper than that.

For example, in the general Web 2.0 project to re-define identity, trust, reputation, etc. part of the ticket of admission for ideas that will spread are things that improve opportunities for the suveillance of users by large organizations. Often the debate is obscured by a focus of discussion on "privacy features" that, within the context of greater surveillance, give the user some asthetic choices but the general program of increasing surveillance and selling off the results to investors as a form of entrepreneurialism goes unchallenged.

Aside from tuning the lottery game known as on-line advertising, what other uses do you think people are finding and will find for this population measurement data? We're being "sold out," by these businesses.

-t

  Brian Behlendorf [12.14.07 09:14 AM]

"This happens to be the same business that CollabNet tried and failed at many years ago" is not an accurate statement. One of the three experiments in open source business models that CollabNet set up was a marketplace named SourceExchange, which unlike the SourceForge marketplace was not about selling support services but rather a system for open source developers to bid on RFPs for enhancements to open source projects, EBay-style. It had a reputation system built in, as well as certified Peer Reviewers who helped ensure quality. It was a success at one level - about $200K worth of projects were funded and completed through the system over the course of 3 years. However, another one of CollabNet's experiments was proving to be more fruitful (the business CollabNet is in today), and given the state of the industry in 2001-2002, focusing on one great idea rather than three was the right thing to do.

Back to the point of the article - I think it's very important that any cross-site system for reputation recognize that reputation is not a scalar variable, but a multivariable vector; instead just a number, it's a compilation of many different possible attributes. A given person may be rated highly by one person for their promptness in business while rated poorly by another for their writing skills. Frankly, I don't find it a problem that assessing someone's reputation involves something more sophisticated than a DNS query.

  Michael H [12.14.07 09:50 AM]

Thomas

I guess that's part of the problem. I can see individual sites coming up with reputation systems. Slashdot's karma system is more-or-less like that, though it is a general idea where positive karma = a good reputation. When people with positive karma post, the posts are given bonus points. In a sense, they are "modded up" on the basis of the user's karma/reputation. The practical aspect of this is people can set a filter so they won't see posts below a certain level of points. This helps cuts down on the signal to noise ratio for the user.

However, I'm not sure about the portability aspect. I visit a wide variety of sites; some for work, others for classes I'm taking, and still others because they're relevant to interests of mine. Having Excellent karma on /. isn't really relevant to a site on French culture.

Some websites probably need a reputation system more than others. Auction sites like eBay use ranking system based on feedback from parties in transactions. While I don't go to B2B sites, I think knowing the reputation of a supplier would be beneficial before entering into a contract with them. Wikipedia needs one badly. A blogger posting pics of their dogs, probably not so much...unless they are a dog breeder or professional dog handler.

Maybe reputation is more about the context of how it is earned, as it is good/bad.

  thacker [12.14.07 10:13 AM]

Reputation is hardly linear and shouldn't be confused, let alone attempts made to interpret it via the Internet, with things such as 'authority' or 'expertise'.

  Ramses [02.15.08 12:27 PM]

I think it would be too easy to game the system, either to trash someone's reputation or make them look like an expert when they are not. I can certainly see spammers trying to "enhance" their reputations.