Jim Giles, contributor
Here’s something for Twitter fans to think about. When you’re chatting with other users, are you sure you’re talking to a human and not a piece of software?
Earlier this year, researchers created software that can impersonate human Twitter users. New Scientist ran a story yesterday describing how these creations were surprisingly successful at infiltrating networks of human users.
While reporting that piece, I came across a curious example of another “socialbot”. This one was created by accident. Last year, Luca Aiello, a PhD student at the University of Torino, began analysing an online social network dedicated to the discussion of books. He created software, known as a crawler, that mapped the connections within the network. To give the crawler access to the network, Aiello had to create an account for it. To his surprise, users started noticing that the crawler was visiting their pages. They flocked to the crawler’s page. It now has over 70 followers and has received over 2,000 messages, says Aiello, making it one of the most popular members of the network.
He is now considering how to use this unexpected opportunity. “I want to see if this bot can be influential,” says Aiello. The crawler’s profile is currently empty, but Aiello could use it to recommend a book. How many members of the network, which numbers around 100,000 people, would then start reading that book, he wonders.
Jim Giles is a correspondent in New Scientist’s San Francisco office. His (human-powered) Twitter feed is at http://twitter.com/jimgiles.