Wired Magazine ran an article recently about Greg Marra, a Google Plus Product Manager who created a Twitter bot (@Trackgirl) that people mistook for a real user. Targeting runners, the bot followed people back who followed it and repeated messages based upon predetermined keywords. When one of those messages described an ankle injury, people actually messaged get well wishes to the bot account. In a talk, Marra marveled “People were sympathizing with a Python script.”
I knew a guy back in high school who seemed pretty nice. When I saw him again at our reunion, we chatted a bit and found we had some things in common. I didn’t really think two things about it until he friended me on Facebook and I saw him frequently posting negative things about Barack Obama. I don’t care for politics, but since I thought he was a pretty nice guy that I had some stuff in common with I just ignored his posts. Then his posts started to be overtly racist. And that’s all I know because I sent him a note, unfriended him, and blocked him. I was sympathizing with an ignorant dirtbag.
Point being, the barrier for entry for acquaintance is rather low. I don’t think it’s novel or impressive that people can be deceived be it by a bot on Twitter or in real life. If we scrutinized everyone before we had any contact with them I don’t think we’d associate with many people. Gaming people in social isn’t particularly novel or impressive – it’s just kind of dumb.
The Wired article proposes scenarios where a legion of bots could game trending topics on Twitter, which may be true. Whatever hashtag the bot Army embellishes Twitter with, I can say with certainty that it won’t be the stupidest thing that has ever trended.
One of the things I like to do is to use people’s names when I tweet. I think it’s a small, personalized touch that hopefully shows people that there is genuine care behind the mundane message that follows it. Once in awhile I’ll get a note from someone chiding me saying “Nobody calls me Michael, everyone that knows me calls me Mike.” Yet their Twitter profile says “Michael.” And while I’ll generally just apologize, what I want to ask is if it’s really my fault that I didn’t know their name, or if it’s theirs for putting the wrong name on their profile? I feel the same about intuitive bots – are these people really dupes because they showed some kindness to a bot, or is building a bot to engage people just a deceptive thing to do?
In any case, Marra posted his source code for “@trackgirl” here.