New Faculty Spotlight: Aylin Caliskan

What is your Research Focus?

My research interests lie in artificial intelligence (AI) ethics, AI bias, computer vision, natural language processing, and machine learning. I study implicit machine cognition to uncover the underpinning mechanisms of information transfer from human society to AI. I investigate the reasoning behind AI representations and decisions by developing computational methods that detect and quantify human-like associations and biases learned by machines. How do machines that automatically learn implicit associations impact humans and society? What are the implications for justice?

AI systems automate consequential decision-making processes that determine life’s outcomes and opportunities. AI models in language, speech, and vision are widely used in every domain of society. These AI systems are trained on large-scale sociocultural data, which embed human biases and associations. Therefore, these sociotechnical systems learn social group biases such as gender, race or ethnicity, social class, ability, age, and intersectional social group associations. Developing transparency-enhancing algorithms for AI bias evaluation enables studying bias in AI and society. Analyzing the acquisition and evolution of bias and associations is the first step toward mitigating bias in sociotechnical systems and society. Consequently, this line of research on trustworthy AI informs policy-making.

What opportunities at the UW excite you?

I am excited about working with brilliant students and researchers in our natural language processing, artificial intelligence, human-computer interaction, technology policy, and ethics communities.