Cyber security specialists in New Zealand have created Twitter robots that give a pretty good imitation of actual humans. Good enough at least that people are following them and having discussions with them. The Atlantic has the story.
The group started the Twitter bots as part of an experiment on “social engineering”, or how to trick people (as opposed to machines) into revealing info, and on how people can be manipulated online. Turns out, we can.
Right now this isn’t being used maliciously, and the researchers doing this aren’t going to. But it’s easy to see how similar techniques could be exploited for bad ends: if a spambot could seem more like a legitimate person we’d follow them and they could blast out affiliate links. Spam is already pretty bad on Twitter, but the more sophisticated spammers get, the more annoying it will be for the rest of us.