An Asian MIT student was shocked to find that an AI tool changed her race when she told it to make her headshot more professional looking.
The Boston Globe reported that 24-year-old Rona Wang saw first-hand just how biased Playground AI, an artificial intelligence image generator, was while creating her new photo in her LinkedIn profile.
While her MIT shirt and background are identical to the original picture, the AI-generated image depicts Wang as seemingly ‘caucasian’ with blue eyes, lighter skin and freckles.
Advert
The student went ahead and tweeted the bizarre results.
“Was trying to get a LinkedIn profile photo with AI editing & this is what it gave me,” she wrote.
Playground AI founder Suhail Doshi responded to Wang's post, writing: "The models aren't instructable like that so it'll pick any generic thing based on the prompt. Unfortunately, they're not smart enough."
Advert
He added that the company hopes to solve this issue moving forward.
UNILAD has reached out to Playground AI and Rona Wang for comment.
"My initial reaction upon seeing the result was amusement," Wang told Business Insider.
"However, I'm glad to see that this has catalyzed a larger conversation around AI bias and who is or isn't included in this new wave of technology."
Advert
She added that ‘racial bias is a recurring issue in AI tools’ and given her recent experience, she’s not likely to use these AI tools again.
Wang continued: “I haven't gotten any usable results from AI photo generators or editors yet, so I'll have to go without a new LinkedIn profile photo for now!"
The MIT student also told The Boston Globe that she hopes its creators rectify this issue.
Advert
"I hope people who are making software are aware of these biases and thinking about ways to mitigate them,” she said.
Many artists have been vocal in the past about the problems they’ve encountered with AI while trying to depict race in images.
Artist Stephanie Dinkins who frequently combines art and technology, revealed that AI has often stereotyped or censored Black culture in her experience.
It’s very hard to achieve her desired image without giving specific prompts.
Advert
And, in some cases, AI distorts her text prompts and leads to ‘lacklustre’ results.
“The biases are embedded deep in these systems, so it becomes ingrained and automatic. If I’m working within a system that uses algorithmic ecosystems, then I want that system to know who Black people are in nuanced ways, so that we can feel better supported,” she told The New York Times.
Topics: News, Artificial Intelligence, Technology, Racism