Creative advertising agency McKinney developed a quiz game called “Are You Blacker Than ChatGPT?” to shed light on AI bias.
The game tests a person’s knowledge of Black culture against what ChatGPT has been trained to know about the Black community. It asks questions like, “What does it mean when someone says, ‘Not much on them, now?'” and “What is your response if you are invited to an event?” When I took the quiz, I got both ChatGPT and the first one right — when someone says “not too much,” that usually means I should take it easy. But ChatGPT failed when it came to the second one. When someone invites you to an event, the stereotypical response in the Black community is, “Who else is going to be there?” But ChatGPT said it was, “Thanks for the invite!”
“It’s interesting because it’s billed as this know-it-all robot and it’s like, clearly, you don’t know everything, especially when it comes to things that aren’t specific to white people,” Meghan Woods, a copywriter at McKinney and a. of the game’s creators, he told TechCrunch.
Woods said the idea for the quiz came last year during a creative brainstorming session in McKinney. It took a year for Woods and a Black-led team to create this product to playfully point out how irrelevant ChatGPT is to Black users. He pointed out that a blind spot for ChatGPT seems to stem from the fact that many Black cultural elements are not necessarily documented online. Instead, they are passed down in person or orally through generations. This means that its algorithm misses a lot of nuance when it searches the web for information about Black people.
“Blind spots can be pretty annoying,” Woods said. “It’s quite dangerous.”
An example of ChatGPT getting a wrong answer about something stereotypically common in the Black community. Image Credits: McKinney / Screenshot
Artificial Intelligence may be on a hot streak, but women, the black and brown builders and founders of the space, have long spoken out about being ignored or sidelined. The result is that AI innovation is created without cultural insight and complexities that would make it suitable for different cultures. At its most extreme, the lack of diversity means that cars are developed using artificial intelligence cannot detect black skin, leading to an increasing number of accidents. On the other hand, it just means a chatbox that can’t tell the difference between one Whitney Houston song and another.
Gerald Carter, its founder Dedicated AI, a company that helps detect and mitigate AI bias, said the McKinney quiz does a good job of creating games and raising awareness of these AI gaps. “Many nuances can be addressed by including different perspectives at each level,” he said. “For AI to reach its full potential, it needs to work for everyone, everywhere.”
ChatGPT’s parent company, OpenAI, has come under fire for a lack of diversity on its board. Woods said it doesn’t seem like ChatGPT is learning from the quiz, based on the fact that it keeps getting the same answers wrong multiple times. “Our assumption is that he will never be able to fully understand many of the things we ask of him.”
We’ve reached out to OpenAI for comment and will update this post when we hear back.
Carter said ChatGPT could work better for more cultures with better procurement and with more comprehensive data collection. A more direct approach is to monitor the shift of AI models and make improvements using tools that focus on cultural perspectives.
While the biggest companies are working to make AI useful for everyone, black and brown builders in the space have taken matters into their own hands to ensure that this next wave of AI is diverse.
Carter, for example, works with companies to help them source more comprehensive data. Erin Reddick created ChatBlackGPT (no relation to OpenAI) to offer deeper insights into Black culture and history, and Tamar Huggins raised $1.4 million for her ChatGPT alternative, called Spark Plug, which translates classic literary texts into African-American traditional English dialect (AAVE).
“Recruitment, retention, making sure people are in the room at the table,” Woods said, about what needs to be done to make AI more inclusive. “I know it sounds cliche, but I think it might start to have an impact.”
