AI and Power: From Bias to Justice

Photo by Seth Lazar

AI and Power: From Bias to Justice

Kate Crawford is an ANU alumna, a Distinguished Research Professor at NYU and a Principal Researcher at Microsoft Research, as well as the co-founder and director the AI Now Institute, arguably the world’s best centre for empirical research into the social impacts of AI in society, now. She is a leading scholar of the social implications of data systems, machine learning, and artificial intelligence.

Prof Crawford gave a public lecture on ‘AI and Power: From Bias to Justice’ on 17th December 2019 at the Manning Clarke Theatre, ANU. This event was a joint production of the Humanising Machine Intelligence Grand Challenge, the 3A Institute, and the Coral Bell School of Asia Pacific Affairs, all at the Australian National University.
Prof Crawford’s lecture focused on the need for a shift in our thinking about AI, from our obsession with identifying biases to finding paths towards justice. Machine learning systems now play a much bigger role in many of our social institutions, from education to healthcare to criminal justice. AI is rearranging power and who can what, where and how. The control of these power structures is dominated by 5 to 7 companies. But many scholars have shown the way these systems are built on data that result in the reproduction of structural bias and discrimination. But parity is not justice, a lack of bias is not due process. Equal surveillance is not equality. These systems are dangerous when they fail and harmful when they work.

In this talk, Professor Crawford opened the substrates of training data to uncover the historical origins, labor practices, infrastructures, and epistemological assumptions that go into the production of artificial intelligence. Rather than a focus on technically correcting biases, she argues for a recentering of justice and the enforcement of limits on centralized power. We must stop thinking about ‘taming’ technology but instead start asking if a technology helps us bring about the world we want to live in, or even a world we want to live in.