Ratings For Everything

This page rated G, Suitable For All Audiences

Rating systems like the one used on this website inevitable provoke follow-up ideas like this: Why not have a rating site for website promotion, you enter an URL and people vote on it. You could learn from the positive and the negative feedback. Or like this: Why not let companies introduce their products on a website and let people vote on them. This way, both the companies benefit (from higher profits) and the users (from better products and more transparency).

Let us examine what these ideas have in common: There is a certain good, be it information, be it services, be it material products, and the quality of this product is supposed to be decided by the people that use it. We are talking about rating systems here. Is this something new? Yes, definitely. Let me explain why.

Only 10 years ago, there were two main outlets for information about products people are supposed to buy, books they are supposed to read, etc.: Other people and mass media. The problem with other people is the scope of normal inter-human communications. The information you can obtain is limited, you can only inquire one person at a time, and only one person at a time can give you a recommendation (unless you are in a room filled with people).

Then there are mass media. They are a truly great way of communicating an idea, but they are controlled by relatively few people. There are the big media companies ABC, MSNBC, FOX, Murdoch, Bertelsmann, and there are the corporations which pay for advertising. The goal of the people behind the media and the corporations is to collectively make you buy what they produce and deliver, regardless of its real quality. With enough effort, they can even make you buy/believe complete crap you'll never need. This is dangerous even for the companies themselves, since mass media controlled by few people have a tendency to cover up problems instead of examining them.

Like the Greenhouse Effect: Instead of reducing CO2 emissions, companies pay PR flacks to say that it doesn't exist, produce pseudo-scientific videos, bribe mainstream media to print/not to print certain stories, etc. etc. Or the effects of tobacco on health: Hundreds of pseudo-scientific studies are produced, mass media are used as an outlet for obvious lies, etc. In both cases, company CEOs and their children will inevitably become the victims of their own propaganda: They will live in a world with a depleted ozone layer, with global warming, they might get lung cancer from smoking. Do they believe in their own lies? Yes, of course. They have no other choice since they try to stifle the sources of all contradicting information.

Now, in the year 2000 (that's the time I write this), the Internet gives us the solution to this problem, but we, the people, have to build it. Using the Net, we can bring thousands, even millions of people together, all in one virtual room, and they can all listen to each other. Big corporations like AOL are not interested in letting people make their own, unbiased decisions. Instead, they want to use the Internet like they would use TV: Broadcasting messages and hoping that many people buy them. Broadcasting more "targeted" messages, but still, pre-fabricated euphemisms, lies and empty promises.

The alternative? We're using it. A system that allows users to vote on a particular subject, no matter what. You can vote on the quality of products, services, ideas, art, whatever. What we need is a universal rating system, decentralized, uncontrollable, uncensorable, unfloodable. Why decentralized, why not a site like Slashdot or kuro5hin or everything2|this one, just for more categories? Simple: Because companies will try to go after opinions they don't like. Nothing like "Microsoft sucks", that is harmless. But opinions like "Microsoft works with the NSA" or "Nestle is responsible for the death of children in Africa" or "Monsanto threatens Third World Countries' farmers' access to seeds" or whatever (these are just examples), especially when they are based on clear and obvious facts. These are the things that large corporations are afraid of and that they will try to censor.

That's why we need a decentralized system that uses the technology of the Net to bring as many people as possible together to decide on the quality of service|quality of the services offered in the marketplace. How could such a system be designed? It would have to be implemented into the ones that are being developed right now, like FreeNet and Gnutella. See my other post on "Distributed dynamic databases" for some of the basic concepts that are, IMHO, required. It would have to give users the opportunity to rate anything, including other ratings and the people that write them, to filter products according to ratings of a certain value, by all persons, or by certain persons. Or to see only new ratings.

It would have to be free, open-source, cross-platform and easy-to-use. It would have to be browsable by a multitude of possibly predefined categories. It would have to be as anonymous as possible.

The fact that this site works, that kuro5hin.org works, that other sites that offer ratings, like Amazon.com, work, shows that the need for them exists. Sooner or later a large corporation will come along and try to monopolize this technology. They will let companies buy a high rating for their products, and they will censor god is dead|unpopular opinions if "necessary". Some booksellers already do this with their book reviews. Since they will claim copyright to all ratings/reviews posted by their users, they will be able to go after every other service that offers them.

We have to be faster, better, cheaper (sorry, NASA). Everyone who is interested in participating in designing the future should join the FreeNet project, especially the mailing list(s), and bring in their ideas. Of course you may also start your own, but FreeNet already has the technology for anonymity, what it lacks are ratings and broadcasts (I will talk about broadcasts some other time). And, of course, micropayment, but that is too far away. You should also take a look at Filetopia (www.filetopia.com), which is the centralized equivalent: with encryption & sophisticated catalog management, anonymous searches and special proxy servers for anonymizing transfers.

Even if you disagree with the details, I think we all agree on one thing: The technology for letting people voice their opinions in an organized and effective manner (i.e. through ratings attached to certain chunks of information that describe a product, idea, service, ..) is a) necessary b) must be built and controlled by the people that use it.

This fits with two notions I have, one older, one more recent; the first, that the Wiki might use a rating system for people akin to the AdvogatoTrustMetric; the second, that a system for rating patterns according to various PatternityTests might strengthen the overall case for patterns.

But,

The core notion of "rating someone" touches a nerve deep within me; it stirs up a wariness I can't quite define but which definitely exists.

Part of it is I suppose the notion that whatever you're rating, you're always rating people. If you rate a pattern of mine you're indirectly rating me. There's the related, deep-seated anxiety I have had about the school system after angsting my way through it, as evidenced by the fact that 10+ years afterwards it still literally gives me nightmares.

There's the notion that "worth" (which is what we want a "rating system" to indicate) is for most non-trivial things a constellation of parameters that vary non-linearly with respect to each other, and thus can't theoretically be subject to "rating systems" which are essentially N-dimensional coordinate systems. (Case in point, the evil notion of IQ; see StephenJayGould's The Mismeasure of Man, ISBN -0393314251 .)

Therefore,

I would suggest that as with many uses of technology, "rating systems" might be good, but only subject to the proviso that there are restrictions on their use, either legal ones or adequately designed structural ones. These restrictions would be to prevent abuse of such systems in the was suggested above.

Reflection on these restrictions, the grounds for them, their effectiveness, and so on, possibly should take precedence on developing the technology itself. As far as I'm concerned, this would hold true even if it should take us much longer to get done thinking about the issues than it would to develop the technology.