As part of University of Michigan’s MLK symposium series, a colloquium is planned Friday on AI and possible discrimination in the way the tool evaluates accents.
Emerging AI model technologies help companies with things from changing accents in real-time to scoring interactions based on an employee’s voice.
Nicole Holliday is Acting Associate Professor of Linguistics at the University of California, Berkeley and this year’s speaker at U-M’s colloquium. She says since AI is in its infancy, these language models are pulling from limited data. The result is AI can label Black voices as irritated and condescending. This has a negative impact on Blacks in the workforce.
“Companies can say, ‘We don’t have a racial bias. We see that our AI gives lower scores to XYZ employee, and it happens that employee is Black.’ Why does the AI do that? Because the AI wasn’t trained on their voice.”
Holliday says only people have the ability to understand accents or dialects because, unlike AI, people can understand contextual information.
You can learn more and register for this lecture here.
Non-commercial, fact based reporting is made possible by your financial support. Make your donation to WEMU today to keep your community NPR station thriving.
Like 89.1 WEMU on Facebook and follow us on X (Twitter)
Contact WEMU News at 734.487.3363 or email us at studio@wemu.org
