ai-bias
(4)
Kieran Snyder and the trouble with a single word
A linguist noticed what managers wrote about women. The question took Kieran Snyder from performance reviews to Textio, and eventually back to Microsoft with a different way of thinking about AI.
The Machine Got the Culture Wrong. So Better Together Built Its Own.
A Washington communications shop encountered a small machine error with a large human meaning. The mistake became a research program, then a product - and a useful lesson in what AI still cannot be trusted to understand alone.
Mainak Mazumdar
Mainak Mazumdar is Chief Business Officer at Fiddler AI, the AI observability and trust company, where he leads sales, customer success and partnerships. A data scientist by training with a Ph.D. from Brown, he spent two decades remaking how the media and advertising industries measure audiences - as Chief Data Officer at Nielsen (where he launched NielsenOne), Chief Research and Analytics Officer at Fox, and a leader at DoubleClick/Google, GfK and Simulmedia. His 2020 TED Talk, 'How bad data keeps us from good AI,' argued that biased data, not algorithms, is the root of unfair AI. He holds more than 15 U.S. patents and has served on the U.S. Census Scientific Advisory Committee.
Deniz A. Johnson
COO of Stratyfy, adjunct faculty at Northeastern University, and one of the most credible voices in the fight against algorithmic discrimination in financial services. Born in Istanbul, educated in engineering and business, Deniz has spent decades bridging the gap between AI technology and human consequences — building interpretable, bias-mitigating AI systems used by 20+ lenders through the Underwriting for Racial Justice programme. She is a serial board member, community convener, and published author who believes fairness and performance are not opposites but the same equation solved correctly.