Our weekly SRI Seminar Series welcomes Seth Lazar, professor in the School of Philosophy at the Australian National University, distinguished research fellow of the University of Oxford Institute for Ethics in AI, and general co-chair for the ACM Fairness, Accountability, and Transparency conference 2022. Lazar’s central interests are in the political philosophy of data and AI, evaluating how advanced technologies create new modalities for the exercise and concentration of power, and asking whether and how those new power relations can be legitimated and justified. Lazar’s research is also concerned with how to design intelligent systems so that they take moral or political values into account, and are calibrated to work with humans as we are, not as we might be.
In this talk, Lazar explores how the increasing mediation of algorithms throughout society might enable new kinds of human flourishing, and potentially support transformative changes to social structures that are otherwise resistant to progress. To consider these potentials, Lazar proposes that we must first understand and diagnose the potential capacities of algorithmic power, paying close attention to the conditions under which such tools can—or cannot—be justified.
Talk title:
“The nature and justification of algorithmic power”
Abstract:
Algorithmic intermediaries increasingly mediate and govern our social relations, across commerce, politics, and sociality more broadly. In doing so, they exercise a distinct kind of intermediary power: they exercise power over us; they shape power relations between us; and they shape the social structures that those social relations constitute. Sometimes, when new or intensified forms of power emerge, our task is simply to eliminate them—to re-establish our independence from domination. But algorithmic intermediaries can enable new kinds of human flourishing, and could support transformative change to ossified social structures that are otherwise resistant to progress. Our task, then, is to understand and diagnose algorithmic power, and determine whether and how it can be justified. This paper uses political philosophy to advance that project—and uses algorithmic intermediary power to advance political philosophy. It offers an empirically-grounded theory of algorithmic power, then sets out the conditions for its justification, paying particular attention to the conditions under which private algorithmic power either can, or must not, be tolerated.
Recommended reading:
Seth Lazar, “The nature and justification of algorithmic power” (PDF), MINT Lab.
About Seth Lazar
Seth Lazar is a professor in the School of Philosophy at the Australian National University, a distinguished research fellow of the University of Oxford Institute for Ethics in AI, and general co-chair for the ACM Fairness, Accountability, and Transparency conference 2022. He is also director of a Templeton World Charity Foundation project on ‘Moral Skill and Artificial Intelligence’, lead CI of an Australian Research Council (ARC) Discovery Project on Ethics and Risk, and, in 2022, will take up an ARC Future Fellowship on ‘Automatic Authorities: Charting a Course for Legitimate AI’. Lazar is a member of a 15 person study committee of the National Academies of Science, Engineering and Medicine, reporting to the US Congress on the ethics and governance of responsible computing research. He was also founding lead of the Humanising Machine Intelligence grand challenge, a multidisciplinary project on the morality, law, and politics of data and AI. In 2022, Lazar will give the Tanner Lecture on AI and Human Values at Stanford University, and the Kamm Lecture in Ethics at Harvard.
Lazar writes on topics in political philosophy, and normative and applied ethics. In recent years, he has begun to focus exclusively on the morality, law and politics of data and AI (broadly construed).
His central interests are in the political philosophy of data and AI, evaluating how big data, AI, and related technologies create new modalities for the exercise and concentration of power, and asking whether and how those new power relations can be legitimated and justified. This involves rethinking how analytic philosophers conceptualize power and legitimacy, as well as integrating empirical and technical research into AI and its impacts on society to enable both a critical evaluation of where we are now, and a productive attempt to articulate promising avenues for the future. This work lays the foundations for a Future Fellowship grant from the Australian Research Council from 2022-26.
As well as thinking about how to reshape society in order to legitimate (or eliminate) the new power relations enabled by intelligent systems, Lazar also thinks about how to design those systems themselves, so that they take our moral or political values into account. This involves answering several distinct kinds of question. How would we decide which values to incorporate? Who should decide? Once that decision is made, how would we operationalize it? This involves transitioning from questions in political epistemology to practical questions in algorithm design, in collaboration with computer scientists. This project is associated with the HMI Grand Challenge, an investment by the ANU in the multidisciplinary pursuit of democratically legitimate intelligent systems.
Lastly, Lazar is interested in thinking about how to design intelligent systems that are calibrated to work with humans as we are, not as we might be: that take into account the predictable and fallible biases and heuristics that we use to navigate the world, as well as attending to how our interaction with intelligent systems will itself change us.
Lazar has published papers in many top journals, including Ethics, Philosophy & Public Affairs, Australasian Journal of Philosophy, Nous, Canadian Journal of Philosophy, Synthese, Philosophical Quarterly, Philosophical Studies, Oxford Studies in Political Philosophy, and Oxford Studies in Normative Ethics. He is an editor of Philosophers' Imprint, and on the editorial board of Oxford Studies in Political Philosophy. Recently he has taken on leadership roles for several computer science-led conferences on AI and society, including serving as program co-chair for the ACM/AAAI AI, Ethics and Society Conference 2021, and program committees for IJCAI, NeurIPS, and EAAMO.
In 2019, Lazar was awarded the ANU Vice Chancellor’s award for excellence in research. He has also received an Early Career Researcher Commendation from the Academy of Social Sciences of Australia in 2016, and won the American Philosophical Association’s Frank Chapman Sharp Prize in 2011.
About the SRI Seminar Series
The SRI Seminar Series brings together the Schwartz Reisman community and beyond for a robust exchange of ideas that advance scholarship at the intersection of technology and society. Seminars are led by a leading or emerging scholar and feature extensive discussion.
Each week, a featured speaker will present for 45 minutes, followed by 45 minutes of discussion. Registered attendees will be emailed a Zoom link approximately one hour before the event begins. The event will be recorded and posted online.
Seth Lazar