No Trolls Needed Robots Will Probably Develop Prejudice On Their Own

first_img It doesn’t take a high IQ to develop a prejudice toward others. In fact, bigotry doesn’t even require actual brain cells.New research from Cardiff University and MIT suggest artificially intelligent machines can demonstrate bias by identifying, copying, and learning the behavior from each other.Some computer systems have already become victims of sexist and racist views.Microsoft’s Tay chatbot, for example, was designed to interact with people, in hopes of helping the firm better understand conversational speech. Targeted at 18- to 24-year-olds, the algorithm turned out to be a huge hit with online miscreants, who taught her to repeat extremely offensive viewpoints. Within 24 hours of Tay’s launch, she was put to “sleep” by her creators.But this study focuses on the possibility of AI evolving prejudicial groups on their own—without the help of Internet trolls.As part of their analysis, researchers played a game of give and take: Each individual chooses whether they donate to someone in their own party or a separate group—based on the recipient’s reputation, and the benefactor’s donating strategy (which showcases their personal levels of prejudice toward outsiders).Participants begin to learn new strategies by copying others, either within their group or the entire population, who gain a higher short-term payoff.“By running these simulations thousands and thousands of times over, we begin to get an understanding of how prejudice evolves and the conditions that promote or impede it,” study co-author and Cardiff University professor Roger Whitaker said in a statement.The findings, published in the journal Scientific Reports, reveal that these decisions—sexism, nationalism, classism, racism, religious discrimination—don’t necessarily require advanced cognitive abilities.“It is feasible,” Whitaker explained, “that autonomous machines with the ability to identify with discrimination and copy others could in [the] future be susceptible to prejudicial phenomena that we see in the human population.“Many of the AI developments that we are seeing involve autonomy and self-control, meaning that the behavior of devices is also influenced by others around them,” he continued, citing self-driving vehicles and the Internet of Things. “Our study gives a theoretical insight where simulated agents periodically call upon others for some kind of resource.”The idea that an autonomous Uber might refuse service to a woman or your connected coffee maker could shut down when activated by a black person is a revolting step backward in the fight for equal rights.Racism, unfortunately, is not limited to human skin tone: A recent study suggests folks have similar automatic biases toward dark-colored robots as they do toward dark-colored people.More than half of adult Internet users believe, within 30 years, machines will be able to perform most activities currently done by humans. Stay up to date on all things bots here.Let us know what you like about Geek by taking our survey. Stay on target Evan Rachel Wood Just As Disturbed by Humanoid Sophia As Everyone ElseMIT’s Thread-Like Robot Slides Through Blood Vessels In the Brain last_img

Leave a Reply

Your email address will not be published. Required fields are marked *