UC Berkeley (Audio) cover logo

Excavating Ground Truth in AI: Epistemologies and Politics in Training Data with Kate Crawford

56m · UC Berkeley (Audio) · 27 Mar 21:00

The last decade has seen a dramatic capture of digital material for machine learning production. This data is the basis for sense-making in AI, not as classical representations of the world with individual meaning, but as mass collections: ground truth for machine abstractions and operations. What happens when data is seen as an aggregate, stripped of context, meaning, and specificity? In what ways does training data limit what and how machine learning systems interpret the world? And most importantly, what forms of power do these approaches enhance and enable? Professor Kate Crawford is a leading international scholar of the social and political implications of artificial intelligence. In this lecture, Crawford shares new work that reflects on what’s at stake in the architecture and contents of training sets, and how they are increasingly part of our urban, legal, logistical, and commercial infrastructures. Series: "Tanner Lectures on Human Values" [Humanities] [Science] [Show ID: 37729]

The episode Excavating Ground Truth in AI: Epistemologies and Politics in Training Data with Kate Crawford from the podcast UC Berkeley (Audio) has a duration of 56:57. It was first published 27 Mar 21:00. The cover art and the content belong to their respective owners.

More episodes from UC Berkeley (Audio)

For the People? Representative Government in America

It has become commonplace that democracy in the United States faces an existential threat. This belief has gained popular currency in the wake of Donald Trump’s presidency, nourished by his conduct in office, the attempt to overturn the 2020 election, and continuing efforts to subvert the electoral process. Whether this is true only time will tell. But a common narrative among scholars of American government holds that representative democracy is failing more systematically than the Trump phenomenon suggests. In this program, Charles Beitz, professor of politics at Princeton University, along with commentary by Martin Gilens, Pamela S. Karlan and Jane Mansbridge, talk about the current state of democratic dysfunction and what the future might hold. Series: "Tanner Lectures on Human Values" [Public Affairs] [Show ID: 38274]

For the People? Representative Government in America: Regulating Rivalry

It has become commonplace that democracy in the United States faces an existential threat. This belief has gained popular currency in the wake of Donald Trump’s presidency, nourished by his conduct in office, the attempt to overturn the 2020 election, and continuing efforts to subvert the electoral process. Whether this is true only time will tell. But a common narrative among scholars of American government holds that representative democracy is failing more systematically than the Trump phenomenon suggests. In this program, Charles Beitz, professor of politics at Princeton University, along with commentary by Pamela S. Karlan and Jane Mansbridge, talk about how to regulate rivalry in democratic representative government. Series: "Tanner Lectures on Human Values" [Public Affairs] [Show ID: 38273]

For the People? Representative Government in America: Intimations of Failure

It has become commonplace that democracy in the United States faces an existential threat. This belief has gained popular currency in the wake of Donald Trump’s presidency, nourished by his conduct in office, the attempt to overturn the 2020 election, and continuing efforts to subvert the electoral process. Whether this is true only time will tell. But a common narrative among scholars of American government holds that representative democracy is failing more systematically than the Trump phenomenon suggests. In this program, Charles Beitz, professor of politics at Princeton University, and Martin Gilens, professor of public policy at UCLA, address how to diagnose the problem of whether or not our system of government is failing. Series: "Tanner Lectures on Human Values" [Public Affairs] [Show ID: 38272]

21st Century Global Health Priorities with Christopher Murray

The world has lived through 2+ years of the COVID-19 pandemic, heightening the awareness of the links between health and other aspects of life including education and the economy. Future pandemics are a real risk but there are a number of other threats to human health and well-being as well. These include climate change, the rise of obesity, inverted population pyramids, inter-state conflict, rising inequalities, antimicrobial resistance. Counterbalancing these threats are the opportunities that may come through the health sector and broader innovation. Using a comprehensive future health scenario framework, Christopher Murray, professor and chair of Health Metrics Sciences, University of Washington and director of the Institute for Health Metrics and Evaluation, explores the range of future trajectories that may unfold in the 21st century. Series: "UC Berkeley Graduate Lectures" [Public Affairs] [Health and Medicine] [Business] [Show ID: 38271]

The Social Safety Net as an Investment in Children with Hilary Hoynes

A hallmark of every developed nation is the provision of a social safety net – a collection of public programs that deliver aid to the poor. Because of their higher rates of poverty, children are often a major beneficiary of safety net programs. Compared to other countries, the U.S. spends less on antipoverty programs and, consequently, has higher child poverty rates. Professor Hilary Hoynes discusses the emerging research that examines how the social safety net affects children’s life trajectories. The long run benefits are significant for the families, but also show that many programs prove to be excellent public investments. This has implications for current policy discussions such as the expanded Child Tax Credit. Series: "UC Berkeley Graduate Lectures" [Public Affairs] [Business] [Show ID: 38275]

Every Podcast » UC Berkeley (Audio) » Excavating Ground Truth in AI: Epistemologies and Politics in Training Data with Kate Crawford