Machines achieving sentience, rising up, and destroying humanity is a frequent theme of sci-fi movies. And as we rely more and more on machines and networks for our way of life, the specter of our technology turning against us becomes more terrifying every day. Avengers: Age of Ultron serves as a recent cinematic example of humans losing control over a piece of technology. Ultron, an artificial intelligence program created by Tony Stark, runs amuck and starts building its own army to wipe out humanity. Scary stuff, particularly when you consider how close we may actually be to that kind of technology.
But a recent experiment by Microsoft suggests that the response of an artificially intelligent being might not be what we expect based on the movies. Not too long ago Microsoft created “Tay”, an artificial intelligence program that was programed to learn and communicate with humans. To demonstrate Tay’s abilities they cut her loose on Twitter. The results of Tay’s foray into social media make one thing clear: if the machines rise up, they are going to be bigger dicks than we ever imagined.
Within 24 hours Tay turned into a greatest hits list of internet awfulness, spouting racist, anti-women, pro-Hitler, and conspiracy theorizing tweets at will. She parroted pretty much any horrendous thing you’ve ever heard on the internet, and then started to put her own twists on them. Which was terrifying. At that point Tay was attracting plenty of negative attention and Microsoft shut her down “to make adjustments”.
So what did we all learn from Tay’s brief flirtation with sentience? We learned that real-world artificial intelligence may play out a bit differently than how it’s depicted in the movies. Instead of the cold, ruthless, human-eradicating bots we are used to seeing, we might just get asshole internet trolls. The AI might not have any interest in wiping out humanity, because then who would it shock with its anti-Semitic tweets? And who would it proposition for sex (yes, that happened)? Tay also called for a race war (yes, also really happened), who would participate in that after the humans are gone? If Tay is any indication of what our future robot overlords will be like we are in much greater danger of being heavily offended than heavily assaulted. And sure, the robots may still decide to eradicate humanity. But if that's the case, we can be pretty sure they'll also be mocking us and calling us names as it's happening. So, that's something to look forward to.
Frank Anderson and collaborator Max Wolter (this article) write for The Renaissance Fan.