This is a blog post that originates more in questions than answers. I find myself fascinated by the ai_agw account on Twitter, a automated chatbot that uses keyword analysis to engage climate change denialists. (I recently found a similar Twitterbot, counterwording, that is devoted to challenging right-wing framing of the proposed Park 51 community center in Manhattan.)
The MetaFilter thread on this reminded me of the recent xkcd that imagined future spammers who are forced to create bots that generate constructive and helpful comments to get through CAPTCHA. In a way, it's already real.
If the "bad guys" were using Twitterbots like this, naturally I'd think it was completely terrible -- but here, it's the folks I consider the good guys, and I have to admit I think it's pretty cool.