Bill Gates has an op-ed in the Wall Street Journal today in which he discusses email spam and the measures that Microsoft is taking to fight it. (As a side note: I wonder whether or not Gates wrote it himself, and if so, how many checkers did it pass through afterwards? I have no reason whatsoever to doubt he wrote it, but I'm truly curious.)
Before discussing technological issues, Gates says that Microsoft has filed "15 lawsuits in the U.S. and U.K. against companies and individuals alleged to have sent billions of spam messages in violation of state and federal laws." That's pretty interesting. Although I'm skeptical of Microsoft's motives in general, unless there's more to this story it sounds like a good thing all around. Microsoft surely incurs costs associated with spam that runs through its corporate systems, plus Hotmail, and the lawsuits will end up benefiting end-users as well.
Gates then briefly mentions some "machine learning" "smart" systems that Microsoft is developing to help blocking software recognize spam before passing it through the system. Extracting meaning from language is one of the most difficult things a human can do, and thus one of the holy grails of artificial intelligence. I'm sure this is fascinating research; say what you like about Microsoft, but they spend more money on research than many universities.
Gates then talks a little about how Microsoft is working together with some other major companies to fight spam at different levels, and I think that's great. The only part of the op-ed that mildly distersses me is where Gates begins talking about government regulation.
A key to eliminating spam is establishing clear guidelines for legitimate commercial e-mail. With industry and consumer groups, we are developing best-practice guidelines to help responsible companies understand how to reach their customers without spamming them. Congress could help by providing a strong incentive for businesses to adopt e-mail best practices. Our proposal is to create a regulatory "safe harbor" status for senders who comply with e-mail guidelines confirmed by an FTC-approved self-regulatory body. Senders who do not comply would have to insert an "ADV:" label--standing for advertisement--in the subject line of all unsolicited commercial e-mail. This would enable computer users either to accept ADV-labeled mail or to have it deleted automatically.It would be very beneficial for companies and individuals to adopt voluntary standards, but I don't like the idea of Congress or the FTC enshrining those standards into law. Aside from free speech issues that might render such standards unconstitutional, the bureacracy is ponderous and technologically incompetent -- in the long run the internet would only be harmed by the inevitable heavy-handedness of the federal government.