Taking Down Pseudoscience, One Robo-Tweet At A Time
Posted by Michael_Byrne on Wednesday, Nov 03, 2010
The Bulletin of the American Scientician has found a new ally: Twitter chatbot @AI_AGW, a program whose mission is to chase around the Twitterverse hunting out the usual tired suspects of pseudoscience, anti-science, and dim “skepticism,” and automatically argue them down with actual facts.
So we don’t have to.
The general principle is pretty simple. The current pool of circulating bad science is limited; the arguments are all pretty “stock.” So, programmer Nigel Leck wrote a script that searches Twitter for hundreds of key phrases—say, “current global warming is just a natural process”— and @AI_AGW kicks back a zinger like "Climate Deniers still flogging those emails http://is.gd/cXthg. " Pretty neat.
According to a post on Technology Review, Leck’s ‘bot can get into days long arguments with anti-science nutbars. Which it’s been programmed to handle, tweeting back different responses to what invariably just repetitions of the same tired arguments.
The catch is that it doesn’t know from sarcasm. And, today in particular, Twitter is full of sarcasm. Like, “The people have spoken. Science is wrong, greenhouse gasses are great — more Medicare and tax cuts for everyone! #populism,” from @mattyglesias. Adorably, the bot responded to that with an earnest argument, probably confusing the hell out of @mattyglesias.
Reach this writer at michaelb@motherboard.tv.
Filed under:
About the author
Michael covers physics, climate science, the future of music, and assorted things fallen through cracks at Motherboard. A native of Colorado, Michigan, and Oregon, he currently resides in Baltimore...
