Ok, so over the past day or so, I’ve managed to get rid of some 2000 spam comments, which I believe I’ve ignored for the greater part of 2 years lol. It pains me to see that despite all these, one thing was clearly obvious; they were completely and unequivocally automated. Now I realise for spammers out there, yes I’m talking to you guys, that spambots just make the task easier, as there’s bound to be some gullible fool out there that will take the bait. But the more prevalent it is, the more obvious it is. And these days newbies to the interwebz are fairly clued up from the rigorous tutelage of 12 yr old children and grandchildren. Keywords, lookups and tags help bots figure out what sort of response to give, making it more relevant to the site’s raison d’etre. But even they get misconstrued and screw up if the bot doesn’t understand colloqualisms.
So what am I really getting at? What’s my point? Well, that’s simple really. I want to see someone build a smarter spam bot. Something that in some way can emulate not only a proper human response, but also detect accurately the context of what the site is really about. Oh I know this may require neural nets abound to simulate human behaviour, but there are still many ways to trick. One such encounter was a simple bot that recorded thousands and thousands of irc conversations and strung chunks of them together to form strangely coherent speech. The more speech there was, the more coherent it could spit out sentences. It was a delight to play with!
So come on, spammers. Up your game, you’re not the salesmen, your bots are. Help them become better salesmen-people-things…
Yes, Joe must have his minions… Muahahaha!!!