As the importance of artificial intelligence grows in the coming years, many experts have been raising red flags after finding that AI bots often quickly learn to be racist.
Writing at Entrepreneur.com, Inbenta founder Jordi Torras says that the problem isn’t as simple as weeding out racist programmers. In fact, he says that even a completely unbiased AI can become racist simply through the basic processes of machine learning.
“AI can perpetuate and reinforce a bias that already exists,” he writes. “For example, if an organization has traditionally hired male CEOs, a bot trained to find future CEOs would look at the past for likely candidates, based upon real data that indicates that previous CEOs were male. The machine would use male candidates as a predictor of someone who is qualified for a job.”
In other words, any machine that has the ability to learn and analyze the real world will inevitably take on and internalize the biases that exist within that world. To Torras, this means that humans must never rely on artificial intelligence as a universal fix for society’s ills, or else it will simply permanently reinforce popular prejudices.
“At a minimum, AI developers can take a more active role to ensure biases are not inadvertently created,” he writes. “Human oversight is still needed at all levels of development, and to ensure there is a healthy mix of opposing views to encourage diversity. Additionally, this must be supported by an equal opportunity for everyone to develop these systems.”