Got a hundred friends on Facebook?
What if 10 of them were actually robots?
Researchers at Gartner Inc., a respected technology analyst firm, predict that in four years, 10 percent of all our online “friends” will be nonhuman. In at least some cases, we likely won’t know which ones.
“By 2015, efforts to systematize and automate social engagement will result in the rise of social bots — automated software agents that can handle, to varying degrees, interaction with communities of users in a manner personalized to each individual,” Gartner said in a recent report.
The expected proliferation of bots won’t necessarily be a bad thing. They may well be used to provide information you want:
Bot (disguised as human): Aren’t Chevy Volts cool? I’d love to have an electric car. Gas prices are driving me crazy.
You: Me, too. I’ve heard Volts are in short supply though, and it’s hard to get one.
Bot: I’ve heard Valley Chevrolet is getting a half-dozen next week. I can get you a number to call if you want one.
I’m not sure who gets the commission on the sale, but the bot surely should get credit for an assist.
There are social bots out there right now, and unfortunately many are being used for sinister means. An example is so-called phishing bots, with their purpose being to get you to give up data like your user ID and password for financial sites.
Nor is the concept of social bots anything new. ELIZA, one of the first “chatterbots,” dates back to the mid-1960s. Designed to parody a Rogerian psychotherapist, it would reply to statements like “My mother hates me” with simple restatements turned into questions: “Why do you think your mother hates you?”
ELIZA’s popularity has endured through the decades in various forms and using various coding techniques, and can still be found on various Web sites, as well as more sophisticated chatterbot variations like Jabberwacky or Elbot.
Earlier this year, The Web Ecology Project set up a programming contest to see who could gather the most Twitter followers by using a chatterbot. Each team would have two weeks to see how many followers could be tricked into following their Twitter account.
The winner, using the Twitter account @JamesMTitus, was incessant and cheerful, never sleeping. “How many pets do you have?” he’d ask. “Any cats?”
Tim Hwang, director of the project in Berkeley, noted in an interview on National Public Radio’s “On the Media” that such bots are cheap to run. What if a swarm of 10,000 of them were released to provide the social scaffolding for an online community based around a specific subject?
Or, what if they were released with the purpose of disrupting a specific online conversation? So, for instance, what if they targeted Twitter feeds during the Iranian election protests or the current “Arabian spring” democracy movements in places like Syria and Bahrain?
NBC’s “Meet the Press” uses a giant TweetDeck display to keep tabs on what’s being discussed online each week. But are those discussion trends really human-spurred or are they synthetic—or possibly hybrids of both?
When you consider how important Twitter and Facebook have become in political movements, and how bots could be used to shift the entire online conversation, it’s easy to see that social bots are something to take seriously.
Not just for the future, but now.