I mean this in a very precise and technical sense. The other day I was complaining that there is no spam detection software that could distinguish random nonsense — quite a lot of the spam I now get has a plain text body that consists of utter gibberish. Presumably there’s a Javascript/HTML payload, but I never see those. It’s obvious to the human eye what this is, but not to the otherwise excellent spamassassin.
But suppose for a moment that someone came up with the kind of sophisticated text analysis necessary to catch categories like ‘gibberish’. Think what else it could do. Think of all the spiders crawling around the web, looking for stuff that could be classified more interestingly than by google. Anything that can reliably identify spam can probably reliably find the subversive thoughts in anything you’ve posted, anywhere, ever. Have a look at the latest US immigration forms, and ask yourself what the FBI could do with a web spider and a technology that could reliably identify all spam. But don’t write your answer on the web.
I had always thought it would be the rise of spam that might close down the public spaces of the internet. Now I see that the rise of anti-spamming technologies might help to drive us into private, encrypted hideaways.
Makes me wonder if spam filters can cope with edward lear.
I wouldn’t worry unduly. It’ll take strong AI to get that sort of thing right, and at that point we’ll have far more to worry about than whether Joe Q Picklegruber wants us to take our underpants off in his airport.
I’m rather puzzled by a whole class of encryption that hasn’t happened yet. A DVD filled with gigabytes of random noise, duplicated and Fedex’d to your pal would make a most excellent one time pad for simple voice over IP telephony (*). That’s the sort of thing you could do with a handful of cheap components hanging off your PC these days. So why hasn’t anyone?
R