×

Stuart Russell

Professor of EECS at UC Berkeley
Holder of the Smith Zadeh Chair in Engineering
Author of Artificial Intelligence: A Modern Approach

Stuart Russell received his B.A. with first-class honours in physics from Oxford University in 1982 and his Ph.D. in computer science from Stanford in 1986. He then joined the faculty of the University of California at Berkeley, where he is a Professor (and formerly Chair) of Electrical Engineering and Computer Sciences and holder of the Smith-Zadeh Chair in Engineering. He is a Fellow of the American Association for Artificial Intelligence, the Association for Computing Machinery, and the American Association for the Advancement of Science. His book Artificial Intelligence: A Modern Approach (with Peter Norvig) is the standard text in AI; it has been translated into 13 languages and is used in over 1300 universities in 118 countries. His research covers a wide range of topics in artificial intelligence including machine learning, probabilistic reasoning, knowledge representation, planning, real-time decision making, multitarget tracking, computer vision, computational physiology, and philosophical foundations. His current concerns include the threat of autonomous weapons and the long-term future of artificial intelligence and its relation to humanity. You can read more about Prof. Russell’s work and accomplishments at his website.

×

Pieter Abbeel

Professor of EECS at UC Berkeley

Pieter Abbeel has been a Professor at UC Berkeley (EECS, BAIR) since 2008 and was a Research Scientist at OpenAI during 2016-2017. Pieter has developed apprenticeship learning algorithms that have enabled advanced helicopter aerobatics, including maneuvers such as tic-tocs, chaos and auto-rotation, which only exceptional human pilots can perform. His group has enabled the first end-to-end completion of reliably picking up a crumpled laundry article and folding it and has pioneered deep reinforcement learning for robotics, including learning locomotion and visuomotor skills. His current research focuses on robotics and machine learning with particular focus on deep reinforcement learning, deep imitation learning, deep unsupervised learning, meta-learning, learning-to-learn, and AI safety. You can read more about Prof. Abbeel’s work and accomplishments at his website.

×

Anca Dragan

Assistant Professor of EECS at UC Berkeley
Founder, InterACT Lab

Anca Dragan got her B.Sc. in Computer Science from Jacobs University Bremen in Germany, after which she went to Carnegie Mellon for her Ph.D. in Robotics. She is now an Assistant Professor in the EECS Department at UC Berkeley. Her goal is to enable robots to work with, around, and in support of people. She runs the InterACT Lab, which focuses on algorithms for human-robot interaction—algorithms that move beyond the robot’s function in isolation, and generate robot behavior that also accounts for interaction and coordination with end-users. She also helped found and serves on the steering committee for the Berkeley AI Research (BAIR) Lab.

You can read more about Prof. Dragan’s work and accomplishments at her website.

×

Bart Selman

Professor of Computer Science and Engineering at Cornell University

Bart Selman is a Professor of Computer Science at Cornell University; previously, he worked at AT&T Bell Laboratories. His research interests include computational sustainability, efficient reasoning procedures, planning, knowledge representation, and connections between computer science and statistical physics. He has (co-)authored over 100 publications, including six which won best paper awards. He is a Fellow of the American Association for Artificial Intelligence and a Fellow of the American Association for the Advancement of Science.

You can read more about Prof. Selman’s work and accomplishments at his website.

×

Joseph Halpern

Professor of Computer Science and Engineering at Cornell University

Joseph Halpern is a Professor of Computer Science at Cornell University. His research focuses on the interface between game and decision theory and computer science, on reasoning about knowledge and uncertainty, and on causality. He has also done work on and continues to think actively about security, (fault tolerant) distributed computing, and modal logic. You can read more about Prof. Halpern’s work and accomplishments at his website.

×

Michael Wellman

Professor of Computer Science and Engineering at the University of Michigan

Michael P. Wellman is a Professor of Computer Science & Engineering at the University of Michigan. He received a Ph.D. from the Massachusetts Institute of Technology in 1988 for his work in qualitative probabilistic reasoning and decision-theoretic planning. From 1988 to 1992, Wellman conducted research in these areas at the USAF’s Wright Laboratory. For the past 25 years, his research has focused on computational market mechanisms and game-theoretic reasoning methods, with applications in electronic commerce, finance, and cyber-security. Wellman previously served as Chair of the ACM Special Interest Group on Electronic Commerce (SIGecom) and as Executive Editor of the Journal of Artificial Intelligence Research. He is a Fellow of the Association for the Advancement of Artificial Intelligence and the Association for Computing Machinery.

You can read more about Prof. Wellman’s work and accomplishments at his website.

×

Satinder Singh Baveja

Professor of Computer Science and Engineering at the University of Michigan
Director of the Artificial Intelligence Laboratory

Satinder Singh Baveja is a Professor in the Division of Computer Science and Engineering, Department of Electrical Engineering and Computer Science at the University of Michigan, Ann Arbor, as well as Director of the AI Lab there. His main research interest is in the old-fashioned goal of Artificial Intelligence (AI), that of building autonomous agents that can learn to be broadly competent in complex, dynamic, and uncertain environments. The field of reinforcement learning (RL) has focused on this goal and accordingly his deepest contributions are in RL. From time to time, he takes seriously the challenge of building agents that can interact with other agents and even humans in both artificial and natural environments. This has led to research in computational game theory / mechanism design, cognitive science, and human-computer interaction.

You can read more about Prof. Singh’s work and accomplishments at his website.

×

Tania Lombrozo

Professor of Psychology at UC Berkeley

Tania Lombrozo received her Ph.D. from Harvard University and is a Professor of Psychology at UC Berkeley. She will be joining the Department of Psychology at Princeton University in July 2018. Her research aims to address basic questions about learning, reasoning, and decision-making using the empirical tools of experimental psychology and the conceptual tools of analytic philosophy. You can read more about Prof. Lombrozo’s work here or here.

×

Tom Griffiths

Professor of Psychology and Cognitive Science at Princeton
Previously Director of the Computational Cognitive Science Lab
Previously Director of the Institute of Cognitive and Brain Sciences

Tom Griffiths is a Professor of Psychology and Cognitive Science at Princeton, and previously the Director of the Computational Cognitive Science Lab and the Institute of Cognitive and Brain Sciences at the University of California, Berkeley. He is interested in developing mathematical models of higher level cognition and understanding the formal principles that underlie our ability to solve the computational problems we face in everyday life. His current focus is on inductive problems, such as probabilistic reasoning, learning causal relationships, acquiring and using language, and inferring the structure of categories. He tries to analyze these aspects of human cognition by comparing human behavior to optimal or “rational” solutions to the underlying computational problems. For inductive problems, this usually means exploring how ideas from artificial intelligence, machine learning, and statistics (particularly Bayesian statistics) connect to human cognition.

You can read more about Prof. Griffiths’ work and accomplishments at his website.

×

Mark Nitzberg

Executive Director of CHAI
Head of Strategic Outreach for Berkeley AI Research

Dr. Mark Nitzberg is currently Executive Director of CHAI, as well as head of strategic outreach for Berkeley AI Research (BAIR). He began studying AI as a stowaway student at MIT in the early 1980’s and wrote his Ph.D. in Computer Vision and Human Perception in 1991 under David Mumford at Harvard University. Dr. Nitzberg has built companies and products in the areas of computer vision, machine learning, financial portfolio optimization, workflow efficiencies, online commerce, development aid data capture and analytics, and film and theatre. Most recently he was the Director of Computer Vision Products at A9/Amazon following their acquisition of The Blindsight Corporation, a maker of assistive technologies for low vision and active aging, where Mark was founding CEO. You can read more about Dr. Nitzberg’s work at his LinkedIn.

×

Martin Fukui

Assistant Director

Martin currently serves as the Assistant Director at CHAI. Prior to his current appointment, he served as the Interim Assistant Director and the Operations and Events Manager. Before joining CHAI, he worked at several different IP transactional law firms in the Bay Area and mostly dealt with contract drafting, contract negotiations, terms of use, terms of service, and privacy policies. He also worked for the California Bar Association by organizing events related to entertainment law and digital media. While he was in law school, he served as the President of the Association for Communication, Sports, and Entertainment Law (ACSEL) and the Co-Editor-In-Chief for the Hastings Communications and Entertainment Law Journal. After working in the IP transactional legal field for several years, he came across AI safety and decided to change his career path and use his skillset to help AI safety research organizations flourish and grow. He received his B.A. from University of California, Berkeley and J.D. from University of California, Hastings.

×

Caroline Jeanmaire

Director of Strategic Research and Partnerships

Caroline leads CHAI’s partnership and external relations strategy, focusing on building a research community around AI safety and relationships with key stakeholders. She also researches models of international coordination to ensure the safety and reliability of AI systems. Before working at CHAI, she was an AI Policy Researcher and Project Manager at The Future Society, a think-tank incubated at Harvard’s Kennedy School of Government. She notably supported the organization of the first and second Global Governance of AI Forums at the World Government Summit in Dubai, with over 200 attendees. In the 2019 edition, she managed the Geopolitics of AI and International Panel on AI research committees. Caroline is experienced in multi-party coordination and negotiation. She was a Youth Delegate to the United Nations for two years with the French delegation. She participated in numerous climate negotiations and technical intersessions (including COP21, COP22, COP23 and COP24). Caroline has a dual master’s degree in International Relations from Peking University and Sciences Po Paris and a bachelor’s degree in political sciences from Sciences Po Paris. She also studied at Tufts University and at the Graduate Fletcher School of Law and Diplomacy. Caroline speaks English, French, Spanish and Mandarin Chinese.

×

Julia Kerley

Assistant to Andrew Critch

Julia Kerley assists Andrew Critch at CHAI. She is an undergraduate student at UC Berkeley majoring in Art Practice.

×

Noor Brody

Operations Assistant

Noor is supported via the BERI-CHAI collaboration and works alongside Martin and Caroline to support CHAI researchers. Their work includes planning virtual AI safety conferences, publishing updates about CHAI, and managing administrative tasks. Noor graduated from UC Berkeley with a degree in Computer Science and has been working at CHAI since January 2020.

×

Andrew Critch

Research Scientist at CHAI
Founder & President of the Berkeley Existential Risk Initiative

Dr. Andrew Critch is currently a full-time Research Scientist at CHAI, and spends around a ½ day per week volunteering for the Berkeley Existetial Risk initiative. He earned his Ph.D. in mathematics at UC Berkeley studying applications of algebraic geometry to machine learning models. During that time, he cofounded the Center for Applied Rationality and SPARC. Dr. Critch has been offered university faculty and research positions in mathematics, mathematical biosciences, and philosophy, worked as an algorithmic stock trader at Jane Street Capital’s New York City office, and as a Research Fellow at the Machine Intelligence Research Institute. His current research interests include logical uncertainty, open source game theory, and mitigating race dynamics between companies and nations in AI development. You can read more about Dr. Critch’s work and accomplishments at his website.

×

Jonathan Stray

Visiting Scholar

Jonathan Stray is a Visiting Scholar at the Center for Human Compatible AI at UC Berkeley, where he works on the design of recommender systems for better personalized news and information. He previously taught the dual masters degree in computer science and journalism at Columbia University, built several pieces of software for investigative journalism, worked as an editor at the Associated Press, and developed graphics algorithms for Adobe. He holds an MSc in Computer Science from the University of Toronto and an MA in Journalism from the University of Hong Kong.

×

Karthika Mohan

Postdoctoral Fellow

Karthika Mohan is a postdoctoral scholar at CHAI mentored by Stuart Russell. Karthika received her PhD in Computer Science (Artificial Intelligence) from UCLA where she was advised by Judea Pearl. Her research is of an interdisciplinary nature and her areas of interest include Causal Inference, Graphical Models and AI Safety. She was awarded the Google Outstanding Graduate Research Award, 2017 which is a UCLA Commencement Award. Currently she serves on the editorial board of the Journal of Causal Inference. In addition she is active on program committees of leading AI/ML conferences and reviews for journals in varied disciplines such as Machine Learning, Psychology and Philosophy. For more details, please visit here

×

Kimin Lee

Postdoctoral Fellow

×

Cody Wild

Machine Learning Engineer

Cody Wild is a Research Engineer at CHAI. She holds a B.S. in Statistical Economics from Tulane University, and a M.S. in Analytics from the University of San Francisco. In the years after finishing her Masters program, Cody worked as an industry Data Scientist, first at LendUp, and then on an applied R&D team for Sophos Labs working on developing deep learning models to detect viruses and other malicious file types. Outside of formal employment, she’s worked to deepen her knowledge in Machine Learning by reading and synthesizing technical content from the field, which can be found online in both long and short form. She is passionate about precise understanding, clear arguments, and clean code, and is glad to be helping support a research team that’s thinking critically about the ways their work and AI writ large could negatively impact the world.

×

Steven Wang

Machine Learning Engineer

Steven is an ML research engineer at CHAI. He became interested in AI Safety after reading the 80000 Hours career guide.

During his undergraduate studies at UC Berkeley, Steven worked with Anca Dragan and Jaime Fisac on confidence-based human predictions. As an intern at CHAI, he worked with Dylan Hadfield-Menell on extensions to Inverse Reward Design.

You can check out what he’s been coding up lately on his GitHub.

×

Tom Lenaerts

Professor at the Université Libre de Bruxelles

Tom Lenaerts, PhD, is Professor at the Université Libre de Bruxelles (ULB) where he is co-heading the Machine Learning Group (MLG). MLG targets machine learning, AI and behavioural intelligence research focusing on time series analysis, causal and network inference, collective decision-making, social AI and behavioural analysis with applications in finance, medicine, cybersecurity and biology. He is currently the director of the Interuniversity Institute of Bioinformatics in Brussels and also holds a partial affiliation as research professor with the Artificial Intelligence Lab of the Vrije Universiteit Brussel, the Flemish counterpart of the ULB. He is vice-chair of the Benelux Association for Artificial Intelligence. He has worked in a variety of interdisciplinary domains and has co-authored many papers in AI in the different areas like evolutionary optimisation, collective intelligence, evolutionary game theory, computational biology and bioinformatics.

×

Joar Skalse

First Year PhD Student at Oxford University

Joar is a first-year PhD student at Oxford University, funded by the Future of Humanity Institute. He previously completed the BA and MCompPhil in Computer Science and Philosophy at Oxford, which he graduated from as top of the year. Joar's research is on machine learning, and how to make AI systems safe and reliable, but his broader interests include everything from cognitive science to formal epistemology. In the past he has done research in areas such as philosophical decision theory, constrained reinforcement learning, computational learning theory, inductive logic programming, and active learning. At CHAI, he is working with Adam Gleave to understand the advantages and disadvantages of different reward learning methodologies from a theoretical perspective.

×

Antoni Lorente Martinez

First Year PhD Student at Universitat Autònoma de Barcelona & King’s College London

Toni is a first-year PhD student, advised by Dr. Mark Coté (King’s College London) and Dr. Jordi Vallverdú (Universitat Autònoma de Barcelona). His work focuses on developing an experimental philosophical approach to the mind via AI and cognitive science, in order to explore the nature and limits of philosophical inquiry. In particular, he is currently researching the relationship between empirical findings and theory development. He is also interested in AI safety and the epistemological role of fiction.

Before starting his PhD, Toni graduated in Aerospace Engineering (2017), and obtained an MA in Political Philosophy at Universitat Pompeu Fabra (2019), and an MSc in Philosophy of Science at the London School of Economics (2020). You can read one of his last articles here.

×

Oliver Richardson

PhD Student at Cornell University

×

Dorsa Sadigh

Assistant Professor of Computer Science and Electrical Engineering at Stanford University

Dorsa Sadigh is an Assistant Professor in Computer Science and Electrical Engineering at Stanford University. Her research interests lie in the intersection of robotics, learning, and control theory. Specifically, she is interested in developing algorithms for safe and adaptive human-robot interaction. Dorsa has received her doctoral degree in Electrical Engineering and Computer Sciences (EECS) from UC Berkeley in 2017, and has received her bachelor’s degree in EECS from UC Berkeley in 2012. She is awarded the NSF CAREER award, the AFOSR Young Investigator Program Award, the IEEE TCCPS early career award, the Google Faculty Award, and the Amazon Faculty Research Award. Learn more at her website.

×

Dylan Hadfield-Menell

Assistant Professor, MIT

Dylan will be starting as an Assistant Professor at the Massachusetts Institute of Technology in July 2021. While at CHAI, Dylan was a Ph.D. student at UC Berkeley advised by Anca Dragan, Pieter Abbeel, and Stuart Russell. His research focused on algorithms that facilitate human-compatible artificial intelligence. In particular, he tried to develop frameworks that account for uncertainty about the objective being optimized. Before coming to Berkeley, Dylan did a Master’s of Engineering with Leslie Kaelbling and Tomás Lozano-Pérez at MIT. At Berkeley, Dylan’s research has taken a turn to focus more on AI safety and, thinking longer term, AI value alignment. In 2016, he and his advisors formally described a cooperative inverse reinforcement learning problem (paper). The problem serves as a tool to help researchers consider how robots could learn humans’ values via cooperative instruction. While robots could learn humans’ values by observing humans, cooperative instruction is likely to be significantly faster. In 2017, Dylan and his advisors described an “off-switch game,” a simplified problem describing scenarios in which a human would like to turn off a robot, but the robot is able to disable its off-switch (paper). They showed that a robot who is uncertain about the utility derived from various outcomes in the game is more likely to allow a human to turn it off. You can learn more about Dylan’s other professional work at his website and follow him on Twitter here.

×

Jaime Fernandez Fisac

Assistant Professor of Electrical Engineering at Princeton University

Jaime is currently an Assistant Professor in Electrical Engineering at Princeton University. Prior to this appointment, he worked at Waymo as a Research Scientist While Jaime was at CHAI as a PhD student, he worked on autonomous robots in both academia and industry, with a particular focus on collision avoidance and multi-agent systems. Broadly, his research focused on safely introducing robotics into society. Under the guidance of CHAI Professors Anca Dragan and Tom Griffiths, and along with fellow CHAI graduate students Vael Gates and Dylan Hadfield-Menell, Jaime worked on solving the cooperative inverse reinforcement learning (CIRL) dynamic game using well-established models of human inference, decision making, and theory of mind from the cognitive science literature. Previous solutions have relied on modelling both the human and robot as perfectly rational and able to coordinate in advance, which are nontrivial assumptions in the real world. Instead, Jaime and his colleagues’ work models the human as pedagogic (i.e., her behaviour will aim to be instructive) and the robot as pragmatic (i.e., it knows the human is not perfectly rational but is still trying to teach it). Results suggest this formulation produces robots that are more competent collaborators (paper). In the past, Jaime has also researched how we can have robots choose their course of action in a way that will be easy for a human observer to anticipate (paper) and how incorporating uncertainty into a safety framework for robotic systems that works in conjunction with their learning process can provide meaningful safety guarantees (paper). At Princeton, Jaime intends to look into how AI systems can utilize models of human cognition and behavior to ensure safe interaction with people. He believes that the safer robots will be those that engage their users and procure their cooperation, rather than try to protect against their indifference. He hopes that designing safe human-centered robotic systems in the short term will give us key insights to tackle the broader, long-term AI safety problem. Also, he has done research on having AI treat their models as fallible to protect against overconfidence. You can explore more about this topic by watching this video. You can learn more about Jaime at his website.

×

Rohin Shah

Research Scientist, DeepMind

Rohin currently works as a Research Scientist on the technical AGI safety team at DeepMind. His general interests in CS are very broad, including AI, machine learning, programming languages, complexity theory, algorithms, and security, and so he started his PhD working on program synthesis. Rohin joined CHAI in his fourth year after he became convinced of the importance of building safe, aligned AI. He now thinks about how to provide specifications of good behavior in ways other than reward functions, especially ones that do not require much human effort. In his PhD, he worked on building AI systems that can learn to assist a human user, even if they don't initially know what the user wants.

Rohin writes the Alignment Newsletter, a weekly publication with recent content relevant to AI alignment that has over 600 subscribers. You can learn more about his work at his website and by following him on Twitter.

×

Thomas Krendl Gilbert

Postdoc at Cornell Tech

Tom graduated in 2021 with his PhD in Machine Ethics and Epistemology. This fall he will join the Digital Life Initiative at Cornell Tech as a postdoc. With prior training in philosophy, sociology, and political theory, Tom researches the ethical and political predicaments that emerge when technical practitioners use machine learning to reshape organizational decision-making. His recent paper “Hard Choices in Artificial Intelligence” argues that advanced AI systems should be developed democratically so that relevant preferences and norms are modeled in a way that is simultaneously accountable, contextually appropriate, and compatible between stakeholders. This work has concrete implications for the design of AI systems that are fair for distinct subpopulations, safe when enmeshed with institutional practices, and accountable to the public interest, including medium-term applications like automated vehicles. You can learn more about his work at his website.

×

Vael Gates

Postdoc at Stanford HAI / CISAC

In 2021 Vael graduated with their PhD in Neuroscience with a focus on computational cognitive science. This fall Vael will start as a postdoc at Stanford HAI / CISAC, where they intend to work on an ethnography of AI researchers.

At UC Berkeley, Vael was advised by Professor Tom Griffiths (a PI at CHAI). Vael’s research is broadly aimed at developing computational models of social cognition. Vael is interested in: How people can infer beliefs and intentions in others, especially by observing others’ actions and employing recursive theory-of-mind (e.g. inverse reinforcement learning, social psychology); Group-level equilibria when agents are collaborating or competing (e.g. game theory, agent-based modeling); and Mechanism design and other ways quantitative characterizations of a phenomenon can be used to predict and shape behavior.

Recently, Vael has worked on two projects related to CHAI’s mission. They have assisted with the paper on solving the cooperative inverse reinforcement learning (CIRL) dynamic game and is working with Professors Anca D. Dragan, Tom L. Griffiths, and Anant Sahai on preference aggregation across agents. More specifically, in the latter project Vael and colleagues have set up a study in which participants are presented with a problem that requires mediating between the preferences of multiple agents. Vael and colleagues take participants’ responses and attempt to explain them using a quantitative model. Their hope is to create a baseline standard of “fair” reactions to the problem—a standard to which the behavior of future AIs can be compared.

Vael wants to continue to approach social cognition and social inference problems from a computational perspective, using probabilistic models and large-scale web-based crowdsourcing to investigate the computational goals and algorithms driving the social mind. By understanding the complex inferences made by human minds, they hope to contribute to the development of artificial intelligence that can collaborate and is compatible with human behavior.

You can learn more at Vael’s website.

×

Alison Gopnik

Professor of Psychology at UC Berkeley
Affiliate Professor of Philosophy at UC Berkeley

Alison Gopnik is a professor of psychology and affiliate professor of philosophy at the University of California at Berkeley. She received her BA from McGill University and her PhD. from Oxford University. She is an internationally recognized leader in the study of cognitive science and of children’s learning and development and was one of the founders of the field of “theory of mind”, an originator of the “theory theory” of children’s development and more recently introduced the idea that probabilistic models and Bayesian inference could be applied to children’s learning. She has held a Center for Advanced Studies in the Behavioral Sciences Fellowship, the Moore Distinguished Scholar fellowship at the California Institute of Technology, the All Souls College Distinguished Visiting Fellowship at Oxford, and King’s College Distinguished Visiting Fellowship at Cambridge. She is an elected member of the Society of Experimental Psychologists, and the American Academy of Arts and Sciences and a fellow of the Cognitive Science Society and the American Association for the Advancement of Science. She has been continuously supported by the NSF and was PI on a 2.5 million dollar interdisciplinary collaborative grant on causal learning from the McDonnell Foundation.

She is the author or coauthor of over 100 journal articles and several books including “Words, thoughts and theories” MIT Press, 1997, and the bestselling and critically acclaimed popular books “The Scientist in the Crib” William Morrow, 1999, “The Philosophical Baby; What children’s minds tell us about love, truth and the meaning of life”, and “The Gardener and the Carpenter”, Farrar, Strauss and Giroux, the latter two won the Cognitive Development Society Best Book Prize in 2009 and 2016. She has also written widely about cognitive science and psychology for The New York Times, The Atlantic, The New Yorker, Science, Scientific American, The Times Literary Supplement, The New York Review of Books, New Scientist and Slate, among others. Her TED talk on her work has been viewed more than 3 million times. And she has frequently appeared on TV and radio including “The Charlie Rose Show” and “The Colbert Report”. Since 2013 she has written the Mind and Matter column for the Wall Street Journal. She lives in Berkeley with her husband Alvy Ray Smith, and has three children and three grandchildren.

×

Brian Christian

Author of The Most Human Human, Algorithms to Live By, and The Alignment Problem

Brian Christian is the author of The Most Human Human, which was named a Wall Street Journal bestseller, a New York Times Editors’ Choice, and a New Yorker favorite book of the year. He is the author, with Tom Griffiths, of Algorithms to Live By, a #1 Audible bestseller, Amazon best science book of the year and MIT Technology Review best book of the year.

His third book, The Alignment Problem, was released in October of 2020.

Christian’s writing has been translated into nineteen languages, and has appeared in The New Yorker, The Atlantic, Wired, The Wall Street Journal, The Guardian, The Paris Review, and in scientific journals such as Cognitive Science. Christian has been featured on The Daily Show with Jon Stewart, Radiolab, and The Charlie Rose Show, and has lectured at Google, Facebook, Microsoft, the Santa Fe Institute, and the London School of Economics. His work has won several awards, including fellowships at Yaddo and the MacDowell Colony, publication in Best American Science & Nature Writing, and an award from the Academy of American Poets.

Born in Wilmington, Delaware, Christian holds degrees in philosophy, computer science, and poetry from Brown University and the University of Washington. A Visiting Scholar at the University of California, Berkeley, he lives in San Francisco.

×

Charis Thompson

Chancellor's Professor of Gender & Women's Studies at UC Berkeley

Charis Thompson is Chancellor’s Professor of Gender & Women’s Studies, and a former founding director of the Science, Technology, and Society Center at the University of California, Berkeley; and Professor of Sociology, London School of Economics and Political Science. Her current book in progress, Getting Ahead, focuses on minds, bodies, and emotions in an age of populism and technology elites. Getting Ahead is the third in her book series on technology and democracy. The first, Making Parents (2005), won the Rachel Carson Prize from the Society for the Social Study of Science, and examined gender, race, and agency in relation to reproductive technologies. The second book, Good Science (2013), examined the geopolitics and bioethics of pro-cures innovation economies. Thompson served on the Nuffield Council on Bioethics Working Group on Genome Editing, and currently serves on the World Economic Forum’s Global Technology Council on Technology, Values, and Policy, as well as UC Berkeley’s Stem Cell Research Oversight Committee, and the faculty advisory board of the Center for Race and Gender. She is a recipient of the Social Science Division Distinguished Teaching Award, and in 2017 received an Honorary Doctorate from the Norwegian Science and Technology University.

×

Dawn Song

Professor of EECS at UC Berkeley

Dawn Song is a Professor in the Department of Electrical Engineering and Computer Science at UC Berkeley. Her research interest lies in deep learning and security. She has studied diverse security and privacy issues in computer systems and networks, including areas ranging from software security, networking security, distributed systems security, applied cryptography, blockchain and smart contracts, to the intersection of machine learning and security. She is the recipient of various awards including the MacArthur Fellowship, the Guggenheim Fellowship, the NSF CAREER Award, the Alfred P. Sloan Research Fellowship, the MIT Technology Review TR-35 Award, the Faculty Research Award from IBM, Google and other major tech companies, and Best Paper Awards from top conferences in Computer Security and Deep Learning. She is a winner of the AMiner Most Influential Scholar Award, as the most cited scholar in Computer Security. She obtained her Ph.D. degree from UC Berkeley. Prior to joining UC Berkeley as a faculty, she was a faculty at Carnegie Mellon University from 2002 to 2007.

You can read more about Prof. Song’s work and accomplishments here.

×

Demian Pouzo

Associate Professor of Economics at UC Berkeley

Demian Pouzo is an associated professor in the Department of Economics at UC Berkeley. Demian joined the faculty at Berkeley in 2009 as an assistant professor after receiving his PhD in Economics from NYU. He also holds an MA and BA in Economics from Universidad Torcuato Di Tella (Argentina). Pouzo’s research interests include econometrics as well as other fields such as economic theory and macroeconomics.

×

Jacob Steinhardt

Assistant Professor of Statistics at UC Berkeley

Professor Steinhardt’s goal is to make the conceptual advances necessary for machine learning systems to be reliable and aligned with human values. This includes the following directions:

Robustness: How can we build models robust to distributional shift, to adversaries, to model mis-specification, and to approximations imposed by computational constraints? What is the right way to evaluate such models?

Reward specification and reward hacking: Human values are too complex to be specified by hand. How can we infer complex value functions from data? How should an agent make decisions when its value function is approximate due to noise in the data or inadequacies in the model? How can we prevent reward hacking–degenerate policies that exploit differences between the inferred and true reward?

Scalable alignment: Modern ML systems are often too large, and deployed too broadly, for any single person to reason about in detail, posing challenges to both design and monitoring. How can we design ML systems that conform to interpretable abstractions? How do we enable meaningful human oversight at training and deployment time despite the large scale? How will these large-scale systems affect societal equilibria?

These challenges require rethinking both the theoretical and empirical paradigms of ML. Theories of statistical generalization do not account for the extreme types of generalization considered above, and decision theory does not account for cases where the reward function is only approximate. Meanwhile, measuring empirical test accuracy on a fixed distribution is insufficient to analyze phenomena such as robustness to distributional shift.

For more information about Professor Steinhardt, please visit: https://www.stat.berkeley.edu/~jsteinhardt/

×

Jakob Foerster

Assistant Professor of Computer Science at the University of Toronto and the Vector Institute

Jakob Foerster received a CIFAR AI chair in 2019 and is starting as an Assistant Professor at the University of Toronto and the Vector Institute in the academic year 20/21. During his PhD at the University of Oxford, he helped bring deep multi-agent reinforcement learning to the forefront of AI research and interned at Google Brain, OpenAI, and DeepMind. He has since been working as a research scientist at Facebook AI Research in California, where he will continue advancing the field up to his move to Toronto. He was the lead organizer of the first Emergent Communication (EmeCom) workshop at NeurIPS in 2017, which he has helped organize ever since. Learn more at his website.

×

John Zysman

Professor Emeritus in Political Science at UC Berkeley

John Zysman is Professor Emeritus in the Department of Political Science. He received his B.A. from Harvard University and his Ph.D. from the Massachusetts Institute of Technology. He has written extensively on European and Japanese political economy, policy and corporate strategy.

His most recent articles include:

The Next Phase in the Digital Revolution: Abundant Computing, Platforms, Growth and Employment in February 2018 CACM
Intelligent Tools and Digital Platforms: Implications for Work and Employment in Intereconomics Review of European Economic Policy (Issue 6 2017)
The Rise of the Platform Economy In NAS Issues in Science and Technoogy, Spring 2016
Entrepreneurial Finance in the Era of Intelligent Tools and Digital Platforms: Implications and Consequences for Work

Recent books include:

The Third Globalization: Can Wealthy Countries Stay Rich Ed. with Dan Brezniz (Oxford University Press 2013)
Can Green Sustain Growth: From the Religion to the Reality To Sustainable Prosperity Ed. with Mark Huberty (Stanford Unversity Press. 2013)

Earlier books include:

The Highest Stakes: The Economic Foundations of the Next Security System (Oxford University Press, 1992)
Manufacturing Matters: The Myth of the Post-Industrial Economy (Basic Books, 1987)
Governments, Markets, and Growth: Finance and the Politics of Industrial Change (Cornell University Press, 1983)

×

Juliana Schroeder

Assistant Professor of Haas School of Business at UC Berkeley

Juliana Schroeder researches how people navigate their social worlds: first, how people form inferences about others’ mental states and mental capacities and, second, how these inferences influence their interactions.

Juliana’s research has been published in journals such as Journal of Personality and Social Psychology, Journal of Experimental Psychology, and Psychological Science. It has been featured by outlets such as the New York Times, Newsweek, NBC, and the Today Show, and has been funded by the National Science Foundation.

Juliana received her B.A. in psychology and economics from The University of Virginia. She received a M.A. in social psychology and advanced methods from the University of Chicago, and an M.B.A. from The Chicago Booth School of Business. Her Ph.D. is in Psychology and Business from the University of Chicago. Juliana is currently an Assistant Professor at the Berkeley Haas School of Business.

You can read more about Prof. Schroeder’s work and accomplishments here.

×

Ken Goldberg

Professor and Chair of the Industrial Engineering and Operations Research Department at UC Berkeley

Ken Goldberg is an artist, inventor, and UC Berkeley Professor focusing on robotics. He was appointed the William S. Floyd Jr Distinguished Chair in Engineering and serves as Chair of the Industrial Engineering and Operations Research Department. He has secondary appointments in EECS, Art Practice, the School of Information, and Radiation Oncology at the UCSF Medical School. Ken is Director of the CITRIS “People and Robots” Initiative and the UC Berkeley AUTOLAB where he and his students pursue research in machine learning for robotics and automation in warehouses, homes, and operating rooms. Ken developed the first provably complete algorithms for part feeding and part fixturing and the first robot on the Internet. Despite agonizingly slow progress, he persists in trying to make robots less clumsy. He has over 250 peer-reviewed publications and 8 U.S. Patents. He co-founded and served as Editor-in-Chief of the IEEE Transactions on Automation Science and Engineering. Ken’s artwork has appeared in 70 exhibits including the Whitney Biennial and films he has co-written have been selected for Sundance and nominated for an Emmy Award. Ken was awarded the NSF PECASE (Presidential Faculty Fellowship) from President Bill Clinton in 1995, elected IEEE Fellow in 2005 and selected by the IEEE Robotics and Automation Society for the George Saridis Leadership Award in 2016. He lives in the Bay Area and is madly in love with his wife, filmmaker and Webby Awards founder Tiffany Shlain, and their two daughters.

You can read more about Prof. Goldberg’s work and accomplishments here.

×

Lara Buchak

Professor of Philosophy at Princeton University
Previously Associate Professor of Philosophy at UC Berkeley

Lara Buchak is a Professor in the Philosophy Department at Princeton University. Her research interests include decision theory, social choice theory, epistemology, ethics, and the philosophy of religion. Her book Risk and Rationality (2013) concerns how an individual ought to take risk into account when making decisions. Her research following the book has focused on applications of her view to ethics, arguing that we ought to defer to individuals’ risk-attitudes in biomedical research; that we ought to weight worse scenarios very heavily in setting climate policy; and that we ought to care a great deal about the interests of the worse-off when acting ethically.

Other topics she has written on include the nature and rationality of faith; group decision-making; the relationship between assigning probability to a hypothesis and believing that hypothesis outright; and the nature of free will. Professor Buchak received her Ph.D. from Princeton in 2009. She spent 12 years in the Philosophy Department at UC Berkeley before returning to Princeton.

×

Mariano Florentino Cuéllar

Chair, Board of the Directors at the Center for Advanced Study in the Behavioral Sciences
Former Stanley Morrison Professor of Law at Stanford University

Justice Mariano-Florentino Cuéllar was nominated by Governor Jerry Brown and began serving on the California Supreme Court in January 2015. A naturalized U.S. citizen born in Northern Mexico, Cuéllar received a B.A. from Harvard magna cum laude, a J.D. from Yale Law School, and a Ph.D. in Political Science from Stanford. Before serving on the Court, he was the Stanley Morrison Professor of Law and Professor (by courtesy) of Political Science at Stanford University. A member of the Stanford faculty since 2001, Cuéllar has written books and articles on administrative law and legislation, criminal law, international law, cyberlaw, immigration, public health law, and the history of institutions. Between 2004 and 2015, Cuéllar also held leadership positions at Stanford’s Freeman Spogli Institute for International Studies. During his tenure leading the Institute and, earlier, its Center for International Security and Cooperation, Cuéllar grew the Institute’s faculty, expanded Stanford’s role in nuclear security research, launched university-wide initiatives on global poverty and cyber security, and broadened opportunities for student and faculty research abroad. You can read more about Justice Cuéllar’s work and accomplishments here or here.

×

Marion Fourcade

Professor of Sociology at UC Berkeley

Marion Fourcade received her PhD from Harvard University (2000) and taught at New York University and Princeton University before joining the Berkeley sociology department in 2003. A comparative sociologist by training and taste, she is interested in variations in economic and political knowledge and practice across nations. Her first book, Economists and Societies (Princeton University Press 2009), explored the distinctive character of the discipline and profession of economics in three countries. A second book, The Ordinal Society (with Kieran Healy), is under contract. This book investigates new forms of social stratification and morality in the digital economy. Other recent research focuses on the valuation of nature in comparative perspective; the moral regulation of states; the comparative study of political organization (with Evan Schofer and Brian Lande); the microsociology of courtroom exchanges (with Roi Livne); the sociology of economics, with Etienne Ollion and Yann Algan, and with Rakesh Khurana; the politics of wine classifications in France and the United States (with Rebecca Elliott and Olivier Jacquet). A final book-length project, Measure for Measure: Social Ontologies of Classification, will examine the cultural and institutional logic of what we may call “national classificatory styles” across a range of empirical domains.

Fourcade is also an Associate Fellow of the Max Planck-Sciences Po Center on Coping with Instability in Market Societies (Maxpo), and a past President of the Society for the Advancement of Socio-Economics (2016).

You can read more about Prof. Fourcade’s work and accomplishments here.

×

Michael Littman

Professor of Computer Science at Brown University

Michael Lederman Littman works mainly in reinforcement learning, but has done work in machine learning, game theory, computer networking, partially observable Markov decision process solving, computer solving of analogy problems and other areas. He is currently a professor of computer science at Brown University.

Before graduate school, Michael worked with Thomas Landauer at Bellcore and was granted a patent for one of the earliest systems for Cross-language information retrieval. Michael received his Ph.D. in computer science from Brown University in 1996. From 1996 to 1999, he was a professor at Duke University. During his time at Duke, he worked on an automated crossword solver PROVERB, which won an Outstanding Paper Award in 1999 from AAAI and competed in the American Crossword Puzzle Tournament. From 2000 to 2002, he worked at AT&T. From 2002 to 2012, he was a professor at Rutgers University; he chaired the department from 2009-12. In Summer 2012 he returned to Brown University as a full professor.

×

Moritz Hardt

Assistant Professor of EECS at UC Berkeley

Moritz Hardt is an Assistant Professor in the Department of Electrical Engineering and Computer Sciences at the University of California, Berkeley. Hardt investigates algorithms and machine learning with a focus on reliability, validity, and societal impact. After obtaining a PhD in Computer Science from Princeton University, he held positions at IBM Research Almaden, Google Research and Google Brain. Hardt is a co-founder of the Workshop on Fairness, Accountability, and Transparency in Machine Learning (FAT/ML) and a co-author of the forthcoming textbook “Fairness and Machine Learning”. He has received an NSF CAREER award, a Sloan fellowship, and best paper awards at ICML 2018 and ICLR 2017. You can read more about Prof. Hardt’s work and accomplishments here.

×

Nika Haghtalab

Assistant Professor in the Department of Computer Science at Cornell University

Nika Haghtalab an Assistant Professor in the Department of Computer Science at Cornell University. She works broadly on the theoretical aspects of machine learning and algorithmic economics. She especially cares about developing a theory for machine learning that accounts for its interactions with people and organizations, and the wide range of social and economic limitations, aspiration, and behavior they demonstrate.   Prior to Cornell, Nika was a postdoctoral researcher at Microsoft Research, New England, in 2018-2019. She received her Ph.D. from the Computer Science Department of Carnegie Mellon University, co-advised by Avrim Blum and Ariel Procaccia. Her thesis titled Foundation of Machine Learning, by the People, for the People received the CMU School of Computer Science Dissertation Award (2018) and a SIGecom Dissertation Honorable Mention Award (2019).

×

Niko Kolodny

Professor of Philosophy at UC Berkeley

Niko Kolodny is Professor of Philosophy at UC Berkeley. He works in moral and political philosophy. He has written on the ethical significance of personal relationships and the nature of rationality, especially instrumental rationality. More recent papers focus on the value of democracy, the justification of the state, and the future of humanity. He has designed a new course for both philosophy and data science majors, “Moral Questions of Data Science,” which explores, among other things, the implications of replacing human judgment in social decision-making with statistical inference and automated algorithms.

×

Rediet Abebe

Junior Fellow at the Harvard Society of Fellows
Assistant Professor of Computer Science at the University of California, Berkeley

Rediet Abebe is a Junior Fellow at the Harvard Society of Fellows and an incoming Assistant Professor of Computer Science at the University of California, Berkeley. Abebe holds a Ph.D. in computer science from Cornell University as well as graduate degrees from Harvard University and the University of Cambridge.

Her research is in the fields of artificial intelligence and algorithms, with a focus on equity and justice concerns. She co-founded and co-organizes Mechanism Design for Social Good (MD4SG), a multi-institutional, interdisciplinary research initiative working to improve access to opportunity for historically disadvantaged communities. Abebe’s research has informed policy and practice at the National Institute of Health (NIH) and the Ethiopian Ministry of Education. Abebe has been honored in the MIT Technology Reviews’ 35 Innovators Under 35, ELLE, and the Bloomberg 50 list as a “one to watch.” She has presented her research in venues including National Academy of Sciences, the United Nations, and the Museum of Modern Art. Abebe co-founded Black in AI, a non-profit organization tackling representation and inclusion issues in AI. Her research is deeply influenced by her upbringing in her hometown of Addis Ababa, Ethiopia.

×

Rosie Campbell

Strategy and Operations Advisor

Rosie Campbell is the Strategy and Operations Advisor for CHAI. She is also currently working at Partnerships on AI as a researcher and used to be the Assistant Director of CHAI. Her academic background is in Computer Science (MSc) and Physics (BSc) and includes some Philosophy and Machine Learning. She is motivated by the long-term and short-term challenges of aligning AI with human interests, but optimistic about the benefits of friendly AI. Before joining CHAI, Rosie worked as a Research Engineer at BBC R&D in Manchester, UK. Here, she worked on a variety of technical research and development projects, including a leading role on a team exploring artificially intelligent production systems. While in Manchester Rosie co-founded the BBC’s Machine Learning Special Interest Group, as well as Manchester Futurists, a thriving community group exploring the impact of emerging technology on society and the future.

×

Siddharth Srivastava

Assistant Professor of Computer Science at Arizona State University

Siddharth Srivastava is an Assistant Professor of Computer Science at the School of Computing, Informatics, and Decision Systems Engineering at Arizona State University. He received his Ph.D. in Computer Science from the University of Massachusetts Amherst. Previously, Prof. Srivastava was a Staff Scientist at the United Technologies Research Center in Berkeley. Prior to that, he was a postdoctoral researcher in the RUGS group at the University of California Berkeley. His research objective is to develop intelligent robots and software agents that assist humans in their daily lives. Towards this objective, his research focuses on developing formal frameworks, algorithms and implementations that allow autonomous agents to reason and act efficiently under uncertainty. His dissertation work received a Best Paper award at the International Conference on Automated Planning and Scheduling (ICAPS) and an Outstanding Dissertation award from the Department of Computer Science at UMass Amherst.

You can read more about Prof. Srivastava’s work and accomplishments at his website.

×

Vincent Corruble

Associate Professor at Sorbonne Université

Vincent Corruble received degrees in Engineering (Ecole Centrale de Lille, France), a M.S. in Systems Engineering (University of Virginia, 1992), a Ph.D. in Artificial Intelligence (UPMC, Paris 6, 1996). After postdocs in South Africa, in the US, and in the UK, he joined the faculty of UPMC (now Sorbonne Université), LIP6 laboratory.

He has contributed to the field of Machine Learning and Pattern Recognition, Machine discovery and Data Science, with applications to history of medicine and medical research. In the early 2000’s, he turned his attention to intelligent agents, especially learning agents and multi-agent systems, with a special interest for multi-agent reinforcement learning in complex systems. This found significant applications in the area of Game AI and later in city modeling and urban simulation. This also led to a focus on virtual characters and agent architectures, with emphasis on affective computing to model emotions and their interplay with personality and social interactions.

Lately Vincent Corruble has developed interest in AI safety, and more specifically on how agents and humans can develop safe cooperation and trust over time. In this context he visited CHAI during the summer of 2019. He has been member of the Multi-Agent Systems group at LIP6 (Sorbonne Université, Paris) since 2007.

×

Wesley Holliday

Associate Professor of Philosophy at UC Berkeley

Wesley Holliday is an Associate Professor of Philosophy and Faculty Member of the Group in Logic and the Methodology of Science at the University of California, Berkeley. He currently serves as the Chair of the Group in Logic and co-organizer of the Berkeley-Stanford Circle in Logic and Philosophy.

Prof. Holliday has worked mainly in formal philosophy and logic, especially modal logic, intuitionistic logic, epistemic logic and epistemology, logic and natural language, logic and probability, and logic and social choice theory. His recent research includes voting theory and computational social choice. Read more on his website.

×

Adam Gleave

Fourth Year PhD student

Adam is a fourth year Ph.D. student in EECS at UC Berkeley advised by Stuart Russell. His research focuses on reward specification for reinforcement learning. Most recently, Adam developed the EPIC distance to compare reward functions. Previously, he has worked on reward modeling techniques with a number of collaborators, including inverse reinforcement learning (IRL) from vision and multi-task IRL. Adam is also interested in the robustness of reinforcement learning, and has demonstrated the existence of adversarial policies in multi-agent environments: policies that cause an opponent to fail despite behaving seemingly randomly. Prior to joining Berkeley, Adam did a M.Phil. with Christian Steinruecken and Zoubin Ghahramani in the Machine Learning Group at the University of Cambridge. You can learn more about Adam’s work at his website and by following him on Twitter.

×

Alex Gunning

First Year PhD Student

Alex Gunning is a first-year PhD student advised by Stuart Russell. He graduated with the degree of Master of Mathematics (MMath) from the University of Cambridge in 2020. Alex was a four-time member of Australia’s International Mathematical Olympiad team, winning three Gold and one Bronze medals and placing first overall in the 2014 competition. Alex interned at Jane Street in the summer of 2019.

×

Alyssa Li Dayan

First Year PhD Student

Alyssa Li Dayan is a first year Computer Science PhD student at UC Berkeley. She received her BS in mathematics with computer science from Massachusetts Institute of Technology in 2018. During undergrad she did research in systems neuroscience and computational cognitive science, before spending a couple years in industry working on prediction and simulation for autonomous vehicles.

×

Arnaud Fickinger

Second Year PhD Student

Arnaud began his PhD at UC Berkeley in 2019 and is supervised by Stuart Russell. He is interested in multiagent systems and is currently working on adding a social choice flavor to assistance games and creating complex behaviors via multiagent interaction.

Arnaud received his MS in Artificial Intelligence and Advanced Visual Computing and his BS in Applied Mathematics and Computer Science from the École Polytechnique in Palaiseau, France. He is currently a Visiting Researcher at Facebook AI and has been a Visiting Research Fellow at Harvard University as well as a Software Engineering Intern at Leia Inc. in Menlo Park.

×

Cassidy Laidlaw

First Year PhD Student

Cassidy Laidlaw is a first-year PhD student interested in human-AI cooperation, learning human preferences, and robustness in machine learning. He received his BS in computer science and mathematics from the University of Maryland in 2018.

Previously Cassidy worked as a freelance software developer and consultant with clients in business and government. He has also worked on computer vision research at the Max Planck Institute for Intelligent Systems. Cassidy is a recipient of a National Defense Science and Engineering Graduate (NDSEG) Fellowship. Learn more at his website.

×

Dan Hendryks

Third Year PhD student

Dan Hendrycks is a third-year PhD student at UC Berkeley, advised by Jacob Steinhardt and Dawn Song.

His research aims to disentangle and concretize the components necessary for safe AI. This leads him to work on quantifying and improving the performance of models in unforeseen out-of-distribution scenarios. He also works on measuring a model’s alignment with human values. Dan received his BS from the University of Chicago.

You can find out more about his research at his website.

×

Daniel Filan

Fifth Year PhD student

Daniel Filan is a 5th-year Ph.D. student of EECS at UC Berkeley, supervised by Stuart Russell. He’s interested in effective altruism and wants to ensure that future artificial intelligences who may be much more strategically intelligent than us behave in a safe way.

In 2016, Daniel worked with Tom Everitt, Mayank Daswani, and Marcus Hutter, considering the problem that a sufficiently advanced AI could choose to modify its source code in order to have easily achievable goals, and such modifications may not be to humans’ liking (paper). They determined that an agent will not self-modify if and only if the value function of the agent anticipates the consequences of self-modification and uses the agent’s current utility function when evaluating the future.

Daniel is currently thinking about mechanistic transparency, the problem of how to understand the workings of trained models.

You can learn more about Daniel’s work at his personal website.

×

Jessy Lin

First Year PhD Student

Jessy Lin is a first year PhD student at Berkeley AI Research. She is interested in building machine learning systems that work well in the dynamic, interactive settings of the real world.

Jessy graduated with a double-major in computer science / electrical engineering and philosophy from MIT. Previously Jessy worked on human-inspired AI with the Computational Cognitive Science Group and on human-in-the-loop machine translation at Lilt. At MIT she organized many seasons of HackMIT with the hackathon community. Jessy also spent a summer with the Natural Language Understanding group at Google Research NY, advised by David Weiss. Learn more at her website.

×

Lawrence Chan

Third Year PhD student

Lawrence is a third-year PhD in EECS at UC Berkeley Advised by Anca Dragan and Stuart Russell.

His research focuses on understanding the assumptions needed to apply preference inference techniques (such as inverse reinforcement learning) to recover good specifications for AI systems. Most recently, he studied the problem of assisting people that are themselves still learning about what they want. Lawrence’s work is helpful to our understanding of how a machine can learn what a human wants the machine to do.

×

Meir Friedenberg

Fourth Year PhD Student at Cornell University

Meir is a Ph.D. student in computer science at Cornell, advised by Joe Halpern. Meir’s work explores how we ascribe responsibility to agents when things go wrong, which will be critical to understand if we are to effectively integrate AI systems into society at large. For example, in his paper “Blameworthiness in Multi-Agent Settings”, he and his advisor introduced a new approach to making a formal notion of blameworthiness work in multi-agent settings. A hallmark of Meir’s work is that it often crosses disciplinary boundaries between computer science and philosophy, law, and cognitive science.

Before coming to Cornell, Meir worked with Philip Resnik and Hal Daumé III on mental health analytics from social media data, an area in which he still maintains active interests.

You can learn more about Meir’s work at his website.

×

Micah Carroll

First Year PhD Student

Micah recently graduated in Statistics from UC Berkeley, and is now pursuing a PhD at UC Berkeley advised by Anca Dragan. His main interest is that of reducing the negative societal effects caused by widespread usage of AI systems. Previously, Micah has worked with Rohin Shah on improving AI agents’ abilities to collaborate with humans in general settings. In their paper, they showed that even leveraging little information about human coordination behavior, the trained AI agents’ performance can be significantly improved. He is now also working on the problem of recommender system alignment. You can find Micah on his website or on Twitter.

×

Michael Dennis

Fifth Year PhD student

Michael is a fifth year graduate student advised by Stuart Russell. Before joining CHAI, Michael worked in theoretical computer science. He is working to close the gap between game theoretic principles and current approaches to multi-agent learning in order to provide better assurances of both performance and stability in these systems. AI is already being used to make decisions in social media, law, and finance. Michael believes that, as AI systems become more integrated into our society, it will no longer be sufficient to model each system in isolation. The failures of these aggregate systems will become less predictable and more impactful such that we will need new tools in order to understand how failures could occur and spread. You can learn more about Michael’s work at his website and by following him on Twitter.

×

Paria Rashidinejad

Sixth Year PhD Student

Paria is a sixth year graduate student working with Stuart Russell.

Read more on her LinkedIn.

×

Rachel Freedman

Second Year PhD Student

Rachel Freedman is a second-year Ph.D. student advised by Professor Stuart Russell. Her research focuses on reinforcement learning and reward modeling as paths toward value aligned AI. She’s particularly interested in integrating methodologies and expertise from computer science, statistics, cognitive science and economics to resolve the inherently interdisciplinary challenge of ensuring that AI is beneficial for humanity.

In her undergraduate thesis, Rachel adapted a kidney exchange algorithm to align with societal values by isolating the underlying moral frameworks in human responses to moral dilemmas. The resulting paper was awarded “Outstanding Student Paper Honorable Mention” at AAAI 2018, and will be published in the AI Journal in 2020. Rachel’s past work on this morally vital and complex topic makes her a great fit for CHAI’s long-term goal of benefiting society.

Prior to joining CHAI, Rachel studied computer science, neuroscience, and philosophy at Duke University through the Robertson Scholars Leadership Program, and at Oxford University as a Registered Visiting Student. At Oxford she cofounded and ran an interdisciplinary existential risk discussion society that met at the Future of Humanity Institute. More details about her work are available at her website

×

Sam Toyer

Third Year PhD Student

Sam is a third-year PhD student in EECS at UC Berkeley, advised by Stuart Russell. As an undergraduate at the Australian National University, he worked on a range of topics, including human pose estimation from video, establishing best practices for serving large-scale satellite data on the web, and accelerating classical and probabilistic planning with deep learning. Currently, he is interested in how to make intelligent agents that can infer and act in accordance with the preferences of their owners. More details about his research are available on his website.

×

Scott Emmons

Second Year PhD Student

Scott is a second-year graduate student advised by Stuart Russell. His research involves analyzing the stability of coordination in value-aligned games with theoretical and empirical results that apply, for example, to assistance games. He developed an algorithm bridging traditional planning methods with model-free reinforcement learning that promotes safety by enabling users to query an agent’s abstract plan before its execution.

In 2019, Scott received a BS in Mathematics and a BA in Computer Science from the University of North Carolina at Chapel Hill. As an undergraduate, he worked with Peter Mucha in the Carolina Center for Interdisciplinary Applied Mathematics, Kris Hauser in the Intelligent Motion Laboratory, and Carol Arnosti in the Arnosti Lab. In high school, Scott worked with Katy Börner and the Cyberinfrastructure for Network Science Center of Indiana University in collaboration with Stephen Kobourov of the University of Arizona. His graduate work is supported by a fellowship from the Department of Energy Computational Science.

Learn more at his website.

×

Smitha Milli

Fourth Year PhD student

Smitha Milli is a fourth-year PhD student at UC Berkeley advised by Anca Dragan and Moritz Hardt. Her research aims to make machine learning systems that are value-aligned, despite the fact that, in the real world, these systems must interact with humans that are both strategic and imperfect. To this end, some of her recent work has studied the robustness of human modeling, the impacts of strategic classification, and the potential for model leakage from model explanations. You can learn more about Smitha’s research on her website and follow her on twitter here.

×

Yuxi Liu

First Year PhD Student

Yuxi Liu is a first year PhD student with interests in mathematics, theoretical physics, philosophy, and AI. Before coming to Berkeley, Yuxi graduated from Australian National University with degrees in mathematics and theoretical physics. Learn more at yuxiliu1995.github.io.

×

Pulkit Verma

Third Year PhD Student at Arizona State University

Pulkit Verma is a Computer Science Ph.D. student working with Prof. Siddharth Srivastava at Arizona State University. His research focuses on the safe and reliable behavior of AI agents. He is currently investigating the minimal set of interfaces in an AI system that would enable a user to assess and understand the limits of its safe operability. As a starting point, his work shows that a primitive interface that lets an AI system execute high-level instruction sequences in simulators is sufficient to efficiently derive a user-interpretable model of the system in fully observable and deterministic settings. You can learn more about Pulkit’s work at his website.

×

Alex Turner

Intern

Alex is a fourth-year PhD student at Oregon State University interested in making significant progress on AI alignment; in particular, better understanding and quantifying when, why, and how people would be affected by deployment of proposed AI systems. Last summer, he introduced an approach that aims implicitly to account for the difficulty of specifying a formal goal for an AI: have the agent complete its task without overly changing its ability to complete other tasks. The ability to complete one task is often tightly correlated with the ability to complete another. The key insight is that we can have the agent remain equally able to complete tasks in general, hopefully thereby remaining able to do the right thing - even without specifying what that right thing is.

×

Alexis Wan

Intern

Alexis is a senior at UC Berkeley and an undergraduate intern at CHAI. She works with graduate student Alyssa Li Dayan on the problem of balancing the trade-off between diversity and accuracy for recommendation systems. At Berkeley she is majoring in Computer Science and Applied Math.

×

Carlo Attubato

Intern

Carlo Attubato is a PhD student in Computer Science at Linacre College, University of Oxford. He is supervised by Professor Georg Gottlob and works in ​conjunction with the RAISON DATA project on the topic of integrating automated symbolic reasoning with statistical machine learning. He receives full funding from the Royal Society. In 2020, Carlo received his Masters in Mathematics and Philosophy at Hertford College, University of Oxford. He currently serves as Senior Consultant to OSG Digital Consultancy and Contributing Editor to the Oxford Public Philosophy journal.

At CHAI, Carlo works with Michael Dennis on a project combining language-acquisition, learning, and decision-making into a single end-to-end pipeline for RL agents to specify their goals and beliefs. Their goal is to create RL agents which are more interpretable, aligned, and better at generalising.

×

Cynthia Chen

Intern

Cynthia is a rising senior at the University of Hong Kong. She is deeply passionate about developing AI systems that are beneficial to society and align with human values. Broadly, she is interested in improving robustness and reliability in decision-making algorithms.

Cynthia’s research at CHAI primarily focuses on improving imitation learning agents with representation learning methods. Prior to CHAI, Cynthia has conducted research in Automatic Machine Learning (AutoML) and had her paper published at a major top conference. She has also spent her undergrad years at Stanford University and Columbia University as a Visiting Student. Currently, Cynthia is among the Organizing Committee of SafeAI workshop at AAAI and AISafety workshop at IJCAI, where she hopes to help with growing the AI alignment research field. You can find out more about Cynthia at her website.

×

Erik Jenner

Intern

Erik completed his physics undergraduate degree in 2020 and is now a Master's student in artificial intelligence at the University of Amsterdam. He is passionate about Effective Altruism and wants to use his career to help ensure that AI has a positive impact. At CHAI, he will be working with Adam Gleave on improving the interpretability of reward functions. This could help evaluate learned reward functions when no ground truth is available. Before his CHAI internship, Erik did research on graph-based segmentation, causality, and on equivariant deep learning.

×

Joe Benton

Intern

Joe completed his Masters in mathematics at Trinity College, Cambridge in 2021. His primary research interests are in reinforcement learning and AI robustness. At CHAI, he will focus on encouraging robustness in RL agents by developing methods for agents to request additional information where necessary.

Previously, Joe has worked as a quantitative trading intern at Jane Street, and as a research assistant in the algebraic geometry group in the University of Cambridge's maths department. After interning with CHAI, he will begin a PhD in statistical machine learning at the University of Oxford.

×

Lauro Langosco di Langosco

Intern

Lauro graduated with a Master's in applied mathematics from ETH Zurich. He is primarily interested in the problems of AI corrigibility and alignment. Currently he is researching unusual ways in which deep learning systems can fail to generalize (for example by competently pursuing an objective different from the one they were trained on).

Prior to joining CHAI, Lauro worked on topics in probabilistic deep learning and the theory of stochastic optimization. He also co-organized Effective Altruism Zurich, helping run regular events, a conference on AI governance, and a reading group on AI alignment.

×

Matthew Farrugia-Roberts

Intern

Matthew is an aspiring academic researcher with an emerging research interest in the theory, capabilities, limitations, and beneficence of intelligent systems. He has completed coursework mainly in mathematics, computer science, and artificial intelligence at the University of Melbourne and ETH Zürich. Matthew is completing a minor thesis on understanding the remarkable generalization ability of deep learning systems using tools from pure mathematics and theoretical computer science. As an intern at CHAI in 2021, Matthew will work with Adam Gleave and Joar Skalse to investigate the theoretical foundations of the learning of reward functions.

×

Nathan Miller

Intern

Nathan Miller is an undergraduate studying Computer Science at UC Berkeley. He is a researcher with Berkeley Artificial Intelligence Research and currently is a CHAI intern working with Micah Carroll.

Read more about Nathan on his LinkedIn.

×

Neel Alex

Intern

Neel is a former UC Berkeley Computer Science undergrad. He hopes to make AI safety a larger part of AI research as a whole. With how important AI will be and the risks it poses to potential futures mankind might have, Neel hopes to have AI researchers think about safety side-by-side with their own research. At CHAI, he currently works with Rohin Shah on a benchmarking project intended to show flaws in current approaches to reinforcement learning, and propose solutions that will bring us closer to alignable AI.

×

Neel Nanda

Intern

Neel studied mathematics at the University of Cambridge, graduating in 2020. He has spent the past year working with a range of AI Alignment labs. He worked with Michael Cohen at the Future of Humanity Institute on theoretical Bayesian Reinforcement Learning, and with Jonathan Uesato at DeepMind on learning from noisy and biased labels. He hopes to better understand how neural networks work on the inside, and to use this understanding to help create networks that robustly do what we want them to. At CHAI, he works with Daniel Filan on neural network interpretability and clusterability.

×

Pavel Czempin

Intern

Pavel is a Master's student at the Technical University of Munich, focusing on Artificial Intelligence. His research interests include Deep Reinforcement Learning and making AI safe and robust. At CHAI he works with Adam Gleave to find defenses for adversarial policies.

He previously worked as a student research assistant at Fraunhofer AISEC on cognitive security technologies. For more information, please visit his website.

×

Rafael Albert

Intern

Rafael completed his undergraduate education in March 2021, graduating with two bachelor's degrees in computer science and mathematics from RWTH Aachen University, a leading German institution. Prior to working on AI, he conducted research in mathematical logic that resulted in a journal paper and his university's Best Thesis Award. At CHAI, Rafael works with Michael Dennis on reinforcement learning agents that are robust to external perturbations. After his internship with CHAI, he will intern as a quantitative trader at Jane Street before commencing graduate study in machine learning in the fall of 2021.

In addition to research, Rafael is passionate about Effective Altruism: He has served as Director of his local EA chapter and organized an official university course on the topic at his undergraduate institution with more than 200 participants.

×

Thomas Woodside

Intern

Thomas is a rising junior undergraduate at Yale University, studying computer science. Previously, Thomas has completed internships in machine learning research and engineering at Kiva, NASA, and a security startup. At CHAI, he will be working with Scott Emmons, researching safe exploration and natural language constraints for reinforcement learning. Thomas is interested in the question of how to make AI more beneficial for humanity, and hopes that his internship will help him develop skills and knowledge to help answer that question. At Yale, Thomas has been involved with running AI Safety reading groups to raise awareness of the issue, and also helps lead Yale Effective Altruism.

×

Yawen Duan

Intern

Yawen Duan received a B.S. in Decision Analytics at the University of Hong Kong in 2020. His research interests include out-of-distribution robustness and generalizability for decision-making algorithms. At CHAI, Yawen works with Adam Gleave on research on the benefits of reward function transfer: to what extent do reward learning methods actually transfer, and how can we exploit transfer to improve the reliability of RL systems?

Prior to CHAI, Yawen spent a year as a full-time research intern at Huawei Noah’s Ark Lab, where he worked on Neural Architecture Search (NAS) and published two academic papers at the Conference on Computer Vision and Pattern Recognition (CVPR) and European Conference on Computer Vision (ECCV). In addition to research, he is deeply passionate about doing the most good. He was a Tianxia Fellow and the director of Effective Altruism Hong Kong.

×

Yulong Lin

Intern

Yulong studies Computer Science at the University of Cambridge, where he has obtained a starred first. For his dissertation, he worked with Professor Michael Bronstein and Professor Pietro Liò to benchmark dynamic graph neural networks and develop new models. Previously, he has interned at Amazon Web Services (AWS) and the National University of Singapore (NUS). He is interested in ensuring that AI is developed safely. At CHAI, he works with Scott Emmons to develop robust models in response to unrestricted adversarial examples.

Read more about Yulong on his LinkedIn.

×

Beth Barnes

Former Intern

Beth is a researcher at OpenAI. Previously she worked as Research Assistant to Chief Scientist at DeepMind. Beth participated in CHAI’s first internship cohort in 2017, where she implemented and proved properties of IRL/IRD algorithms. Beth earned a degree in Computer Science at the University of Cambridge. At Cambridge she cofounded Future of Sentience (FuSe), a student society affiliated with CSER focused on improving the long-run future. FuSe created an All-Party Parliamentary Group advocating for representation of future generations in the UK Parliament. View Beth’s LinkedIn profile here.

×

Charlotte Roman

Former Intern

Charlotte is a PhD student at the University of Warwick at the center for Mathematics for Real-world Systems. In her PhD research, she uses game theory to study how people respond to intelligent traffic systems such as route planning apps and traffic lights. She is also interested in cooperative multiagent reinforcement learning. At CHAI, Charlotte is working on finding safe beliefs that encourage cooperation in sequential social dilemmas. Her broad interests include AI safety, fairness, and bounded rationality.

×

Chris Cundy

Former Intern

Chris is a PhD Student at Stanford University studying Machine Learning. His research is concerned with ensuring that sophisticated AI systems will robustly and reliably carry out the tasks that we want them to. At CHAI, Chris worked with Stuart Russell and Daniel Filan. He has also worked at the Future of Humanity Institute at Oxford University, collaborating with Owain Evans on scalable human supervision of complex AI tasks. Chris studied Physics as an undergraduate at Cambridge University before switching to complete a computer science master’s degree. During his master’s, he worked with Carl E. Rasmussen, developing variational methods for Gaussian Process State-Space Models. His website is cundy.me.

×

David Lindner

Former Intern

David is a masters student at ETH Zurich working on reinforcement learning. Broadly, his research is concerned with building reliable intelligent systems that interact with the world. Recently he has been interested in learning robust reward functions from human feedback and ways to reason about the uncertainty an RL agent has about its reward function. At CHAI, David is working with Rohin Shah on inferring human preferences from the current state of the world in complex environments. In the past, he has done work in condensed matter physics, network optimization, computational social science and bias in machine learning. His website is davidlindner.me.

×

Dmitrii Krasheninnikov

Former Intern

Dmitrii is a second-year masters student at the University of Amsterdam, where he studies artificial intelligence. At CHAI, Dmitrii investigates ways to combine misspecified and conflicting reward functions obtained via different specification processes (e.g. hand-specification and inverse reinforcement learning). This is Dmitrii’s third time working at CHAI. Previously he worked with Rohin Shah on an algorithm that can learn human preferences from the current state of the world, and studied properties of the maximum causal entropy inverse RL algorithm with Andrew Critch.

Before CHAI Dmitrii worked on stability of deep learning algorithms as an intern at a small startup in Finland, and designed efficient algorithms for processing next-generation sequencing data at the Bioinformatics Institute in Russia.

You can learn more about Dmitrii at his personal website.

×

Eric Michaud

Former Intern

Eric is a fourth-year undergraduate at UC Berkeley studying mathematics. He is broadly interested in understanding why deep learning works so well, and in investigating the general principles that give rise to intelligence in physical systems.

As a CHAI intern, Eric is working with Adam Gleave to develop interpretability techniques for reward functions. He hopes that this work will help to detect when learned reward functions fail to reflect human preferences, without having to train a policy and without knowing a ground-truth reward.

Before working at CHAI, Eric interned with the Berkeley SETI Research Center and the Lawrence Livermore National Laboratory. You can learn more about him at his personal website.

×

Harry Giles

Former Intern

Harry is masters graduate in mathematics from Cambridge, UK. He is interested in statistics, and has experience working in finance. In particular, he is interested in reinforcement learning, and other topics of decision making under uncertainty and/or Constraints.

At CHAI, Harry has been working with Lawrence Chan on models for suboptimal RL policies, in particular Boltzmann rationality, for which he has also been working on the IRL problem. You can contact him through his website.

×

Jess Reidel

Former Intern

Jess Riedel is a postdoc at the Perimeter Institute for Theoretical Physics specializing in quantum information, decoherence, and the quantum-classical transition. In particular, he has worked on dark-matter detection with matter interferometers and on rigorous definitions for wavefunction branches in a many-body context. As a visiting scholar at CHAI, he has been pursuing interest in the physical embedding of agents and in the formal verification of self-modifying hardware. Riedel was previously a postdoc at IBM Research with Charlie Bennett. He earned his PhD in Physics from UC Santa Barbara in 2012, advised by Wojciech Zurek at Los Alamos National Lab. His website is jessriedel.com.

×

Johannes Treutlein

Former Intern

Johannes is an incoming MSc student in computer science at the University of Toronto. Previously, he studied mathematics and computer science at the Technical University of Berlin and the University of Oxford. His research interests include multi-agent learning, game theory, bargaining and decision theory. At CHAI, he is working with Jakob Foerster and Michael Dennis on whether and how cheap talk can help with zero-shot coordination in cooperative multi-agent RL. Before that, Johannes was a research assistant at the machine learning lab at his university. His website is johannestreutlein.com.

×

Matthew Rahtz

Former Intern

Matthew is a Research Engineer at DeepMind. He earned an MSc in Neural Systems and Computation at ETH Zurich and a bachelor’s in Electronic Engineering at the University of York. Matthew interned at CHAI in 2018. Read his blog at amid.fish.

×

Michael McDonald

Former Intern

Michael completed his undergraduate studies at UC Berkeley with a double major in computer science and applied mathematics. Broadly, his interests cover value alignment in multi-step/multi-objective task scenarios and how robots can learn to efficiently aid humans in such scenarios.

As an undergraduate student, Michael researched under Professor Pieter Abbeel with Dylan Hadfield-Menell on task and motion planning algorithms - initially for a laundry-robot exhibit for the Victoria and Albert Museum in London meant to examine the feasibility and challenges of integrating robotic agents into everyday life. That work then served as foundation for designing imitation-learning algorithms for accomplishing multi-objective plans with normally intractable cost functions.

This coming autumn Michael will begin his master’s here at Berkeley in the EECS 5th year M.S. program, advised by Professor Anca Dragan.

×

Pedro Freire

Former Intern

Pedro is a M.Sc. student in Mathematics and Computer Science at Ecole Polytechnique, France. Some of his AI interests include (i) how to make agents better at processing higher-level feedback from humans, and (ii) the development of tools for RL interpretability - in particular to better understand how we can reduce variance and increase sample-efficiency in learning, and have agents with better guarantees of generalization.Other of his interests include interactive theorem proving, software tooling, game development, virtual reality, and human-computer interaction more broadly.

Previously, he worked at Google developing tools for monitoring and performing statistical analysis on the behavior of YouTube video classifiers, and worked at Facebook on the development of Augmented Reality software. He is currently working on benchmarks specialized in evaluating properties of Inverse Reinforcement Learning algorithms.

×

Sergei Volodin

Former Intern

Sergei is a Master student at the Swiss Federal Institute of Technology in Lausanne (EPFL) studying Computer Science and Neuroscience. Sergei did an internship at Google which resulted in a paper on causal models in reinforcement learning. Sergei co-founded the local AI safety group in Lausanne which does research and discussions. At CHAI, Sergei is working on adversarial policies with Adam Gleave.

×

Shlomi Hod

Former Intern

Shlomi is a data scientist and educator. He works on developing knowledge, practices, and tools to audit and mitigate the biases of real-world AI systems. He is the developer of Responsibly an open-source toolkit that brings techniques from AI bias and fairness research to practitioners. In CHAI, Shlomi works on developing a new method for increasing the interpretability of neural networks under the supervision of Daniel Filan. Besides that, Shlomi teaches various computing classes on Python programming and machine learning. His focus is on teaching a methodical and thoughtful process of programming problem-solving. He is one of the co-founders of the Israeli Cyber Education Center, where he led the design of national computing programs for kids and teenagers. For example, he initiated and managed the development of the cybersecurity Bagrut (the Israeli matriculation exam, the equivalent of A-level), and co-authored a Computer Network textbook in tutorial approach in Hebrew Before that, Shlomi worked as an algorithmic researcher and a research team leader in cybersecurity. He holds a B.Sc.. in pure Mathematics from Bar-Ilan University. Currently, he studies for an M.Sc. in Cognitive Systems at the University of Potsdam in Germany. In January 2020, Shlomi will start a Ph.D. in Computer Science in Boston Univesity.

×

Sören Mindermann

Former Intern

Sören is a DPhil student in CS in the Applied and Theoretical Machine Learning Group at Oxford University (OATML). His interests in machine learning include its economic and political properties, how it scales, and safety. Before joining OATML, Sören worked on reward inference and machine learning for game theory with David Duvenaud and Roger Grosse at Toronto’s Vector Institute, at the Future of Humanity Institute at Oxford, and with CHAI. He has degrees in machine learning (UCL), maths (Amsterdam) and Future Planet Studies (Amsterdam). His website is oatml.cs.ox.ac.uk/members/soren_mindermann/.

×

Stephen Casper

Former Intern

Stephen Casper is a rising senior at Harvard College pursuing a degree in statistics and a minor in mathematics. In addition to interning at CHAI under Daniel Filan, he is affiliated with the MIT/Harvard Center for Brains, Minds, and Machines and the Harvard Kreiman Lab. Broadly, his research interests align with the goals of making AI systems safe and aligned. More specifically, areas that he is interested in include understanding compressible features and generalization in deep networks, robust reinforcement learning, and decision theory.

Stephen’s work with CHAI is focusing on mechanistic transparency via modular interpretations of neural networks. However, he is also thinking about adversaries in contexts of deep reinforcement learning and decision theory.

In addition to being fascinated by questions in machine intelligence, Stephen’s primary motivations for working on AI alignment involve using it as a strategy for shaping the future impactfully for the better. He is excited about the Effective Altruism community and the paradigm of maximizing the good that one can do in the world.

For more information and contact info, please see his website.