Duke professor – a critic of ‘runaway train’ AI development – is named one of the field’s top women

DURHAM – Cynthia Rudin, a professor at Duke University, is one of the top 10 women in the field of artificial intelligence research and development, reports AI Magazine.

She is “known for her pioneering work in the fields of ML, applied ML, and causal inference. She has also held positions at Columbia, NYU, and MIT,” the magazine says.

Rudin recently was critical of the rapid development taking place in AI, preceding by weeks the issuing of a letter from more than 1,000 scientists, researchers and executives such as Elon Musk and Steve Wozniak calling for a pause in development.

“AI technology right now is like a runaway train and we are trying to chase it on foot. I feel like that because the technology is increasing at a very fast rate. It’s amazing what it can do now compared to even a year or two ago,” she told Eric Ferreri, Senior Writer at Duke Today.

“Misinformation can be generated very, very quickly. Also, recommender systems (that push content to people) in directions we don’t want them to be. And I feel the people haven’t yet had a chance to speak up about this. It’s really technology companies imposing it on us rather than the people getting a chance to decide themselves what they want.”

At Duke Rudin is Earl D. McLean, Jr. Professor of Computer Science, Electrical and Computer Engineering, Statistical Science, Mathematics, and Biostatistics & Bioinformatics PI, Interpretable Machine Learning Lab, according to her Duke bio.

“My research focuses on machine learning tools that help humans make better decisions, mainly interpretable machine learning and its applications,” she says.

“We work on decision trees, sparse linear models and scoring systems, variable importance measures, causal inference methods, interpretable deep learning, dimension reduction, and methods that can incorporate domain-based constraints and other types of domain knowledge into machine learning.

“These techniques are applied to critical societal problems in healthcare, criminal justice and energy grid reliability, as well as to materials science and computer vision. Many of our interpretable machine learning algorithms heavily rely on efficient discrete optimization techniques.”