Abstract: Should we be concerned about long-term risks from superintelligent AI?

If so, what can we do about it?  While some in the mainstream AI community dismiss these concerns, I will argue instead that a fundamental reorientation of the field is required. Instead of building systems that optimize arbitrary objectives, we need to learn how to build systems that will, in fact, be beneficial for us.  I will show that it is useful to imbue systems with explicit uncertainty concerning the true objectives of the humans they are designed to help.

About the Speaker: Stuart Russell received his B.A. with first-class honors 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 Professor (and formerly Chair) of Electrical Engineering and Computer Sciences and holder of the Smith-Zadeh Chair in Engineering. He is also an Adjunct Professor of Neurological Surgery at UC San Francisco and Vice-Chair of the World Economic Forum's Council on AI and Robotics. He is a recipient of the Presidential Young Investigator Award of the National Science Foundation, the IJCAI Computers and Thought Award, the World Technology Award (Policy category), the Mitchell Prize of the American Statistical Association and the International Society for Bayesian Analysis, and Outstanding Educator Awards from both ACM and AAAI. In 1998, he gave the Forsythe Memorial Lectures at Stanford University and from 2012 to 2014 he held the Chaire Blaise Pascal in Paris. 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 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, global seismic monitoring, and philosophical foundations. His books include The Use of Knowledge in Analogy and Induction, Do the Right Thing: Studies in Limited Rationality (with Eric Wefald), and Artificial Intelligence: A Modern Approach (with Peter Norvig).

Abstract: Current technologies and practices have created large stores of medical data, including electronic medical records, genomic data, and mobile-health measurements.  There is great promise for discovery and implementation of more efficient and effective health care, but there are also tensions between the sharing of data and the ability to make assurances about security and privacy to patients and study participants.  I will discuss these challenges in the setting of genomic research and medical record data mining.  In many cases, social mechanisms are likely to be the more reliable safeguards than technical mechanisms for privacy, security, and obfuscation.

About the Speaker: Russ Biagio Altman is a professor of bioengineering, genetics, medicine, and biomedical data science (and of computer science, by courtesy) and past chairman of the Bioengineering Department at Stanford University. His primary research interests are in the application of computing and informatics technologies to problems relevant to medicine. He is particularly interested in methods for understanding drug action at molecular, cellular, organism and population levels.  His lab studies how human genetic variation impacts drug response (e.g. http://www.pharmgkb.org/). Other work focuses on the analysis of biological molecules to understand the actions, interactions and adverse events of drugs (http://feature.stanford.edu/).  He helps lead an FDA-supported Center of Excellence in Regulatory Science & Innovation (https://pharm.ucsf.edu/cersi). Dr. Altman holds an A.B. from Harvard College, and M.D. from Stanford Medical School, and a Ph.D. in Medical Information Sciences from Stanford. He received the U.S. Presidential Early Career Award for Scientists and Engineers and a National Science Foundation CAREER Award. He is a fellow of the American College of Physicians (ACP), the American College of Medical Informatics (ACMI), the American Institute of Medical and Biological Engineering (AIMBE), and the American Association for the Advancement of Science (AAAS). He is a member of the National Academy of Medicine (formerly the Institute of Medicine, IOM) of the National Academies.  He is a past-President, founding board member, and a Fellow of the International Society for Computational Biology (ISCB), and a past-President of the American Society for Clinical Pharmacology & Therapeutics (ASCPT).  He has chaired the Science Board advising the FDA Commissioner, currently serves on the NIH Director’s Advisory Committee, and is Co-Chair of the IOM Drug Forum.  He is an organizer of the annual Pacific Symposium on Biocomputing (http://psb.stanford.edu/), and a founder of Personalis, Inc.  Dr. Altman is board certified in Internal Medicine and in Clinical Informatics. He received the Stanford Medical School graduate teaching award in 2000, and mentorship award in 2014.

Abstract: Supercomputing impacts everybody, everywhere, every day. The simulation capabilities have allowed advanced medicine, energy, aviation and manufacturing. Supercomputers allow us to explore fields such as global climate change, as well as tackle problems for which experiments are impractical, hazardous or prohibitively expensive. The Department of Energy is a leader in supercomputers as part of their national security mission. With the demise of underground testing, supercomputers are a key resource used to ensure the safety and reliability of the nuclear stockpile. This talk will explore the buildup to our current petaflop systems and the challenges to obtaining exascale systems in the future.

About the speaker: As Acting Associate Director for Computation at Lawrence Livermore National Laboratory (LLNL), Trish Damkroger leads the 1,000-employee workforce behind the Laboratory’s high performance computing efforts. The Computation team develops and deploys an integrated computing environment for petascale analytics and simulations such as understanding global climate warming, clean energy creation, biodefense, and nonproliferation. LLNL’s computing ecosystem includes high performance computers, scientific visualization facilities, high performance storage systems, network connectivity, multiresolution data analysis, mathematical models, scalable numerical algorithms, computer applications, and necessary services to enable LLNL mission goals and scientific discovery through simulation.

Abstract:  Institutions like LLNL are part of an enterprise established in the mid-twentieth century to enable teams of scientists and engineers to deliver technological capabilities to address challenges to U.S. national security.  The steadily increasing pace of technological change, the reduced proportion of U.S. government funding invested in research and development relative to private sector investments, and the accelerating resources and programs for research globally have dramatically changed the context for this enterprise. Operational practices established to meet the national security needs of the last century must be updated to ensure that the national security science and technology enterprise can continue to deliver high quality capabilities to meet future threats and innovation to enhance U.S. national and economic security. Possible approaches for updating research careers and the structure of institutions include revamping policies for domestic and international partnerships, more effectively managing dual use technologies, and updating enterprise elements to draw on cutting-edge developments in academia and industry.

About the Speaker:  Patricia Falcone is the Deputy Director for Science and Technology, and Chief Technology Officer, of the Lawrence Livermore National Laboratory (LLNL). From 2009 to 2015, she served in the White House Office of Science and Technology Policy, including as the presidentially appointed and Senate-confirmed Associate Director for National Security and International Affairs.  Earlier she worked at the Sandia National Laboratories in Livermore, CA.  She earned a Ph.D. working in the High Temperature Gasdynamics Laboratory in Stanford’s mechanical engineering department.

Abstract: Somebody once said, “What a damn fool can do for a dollar, an engineer can do for a nickel.” Thinking about cost as an engineering constraint brings new life to ideas. This is what makes the difference between an idea influencing a hundred people or a billion. With our planet literally teeming with problems (ecological, health and social), it’s time to take cost constraints into serious consideration. As physicists, we like to make stuff. We use these skills (and field work) to design solutions for extremely resource constrained settings, specially in the field of global health. I will discuss our current work from field diagnostics to high-throughput vector ecology and hands on science education and talk about it’s implication in a global context. I will also discuss outcomes, and lessons from a global experiment - Foldscope (a 50 cent origami microscope); where we shipped 50,000 origami microscopes around the world (130+) countries enabling curious users to discover and explore the microscopic world surrounding them. 

About the Speaker: Manu Prakash is an assistant professor in bioengineering. He leads a curiosity driven research group, focused on technological interventions in extreme resource-poor settings, tackling global public health problems. A physicist and a prolific inventor, his inventions include a 50 cent “print-and-fold” paper microscope, a $5 chemistry lab, a computer that works by moving water droplets in a magnetic fields, and Oscan, a 3-D printed smartphone add-on that helps diagnose oral carcinomas responsible for 40% of cancer-related deaths in India. Professor Prakash has been distinguished as a Frederick E. Terman Fellow (2011-2013), a Pew Scholar (2013-2017), a top innovator under 35 by MIT Technology Review (2014) and in the Brilliant 10 by Popular Science (2014). Born in Meerut, India, Prakash earned a BTech in computer science and engineering from the Indian Institute of Technology in Kanpur before moving to the United States. He did his master’s and PhD at MIT before founding the Prakash Lab at Stanford.

Abstract: Technological and social forces are allowing the growth of science outside a professionalized context.  The Internet, mobile devices, and public availability of big data are technological facilitators of “citizen science”, in part through enabling data sharing, analysis, and communication.  These technological changes now allow the entire scientific process, from funding and development of a research agenda, to conduct, analysis and dissemination and application of findings, to take place without the involvement of any science professionals or research-related institutions.  However, most ethical and regulatory frameworks for biomedical science arose from concepts of obligations of professionals and (largely not-for-profit) institutions.  We will discuss current examples of “citizen science” in biology and clinical research, and the ethical and policy implications.

About the Speaker: Mildred Cho is a Professor in the Division of Medical Genetics of the Department of Pediatrics at Stanford University, Associate Director of the Stanford Center for Biomedical Ethics, and Director of the Center for Integration of Research on Genetics and Ethics. She received her B.S. in Biology in 1984 from the Massachusetts Institute of Technology and her Ph.D. in 1992 from the Stanford University Department of Pharmacology.  Her post-doctoral training was in Health Policy as a Pew Fellow at the Institute for Health Policy Studies at the University of California, San Francisco and at the Palo Alto VA Center for Health Care Evaluation.  She is a member of international and national advisory boards, including for Genome Canada, the March of Dimes, and the Board of Reviewing Editors of Science magazine.  Her current research projects examine ethical and social issues in research on the human genome and microbiome, synthetic biology and genome editing, and the ethics at the intersection of clinical practice and research.  

Abstract: Faster evolving technologies, new peer adversaries, and the increased role of non-government entities changes how we think about decisions to develop and adopt new technology. Uncertainties about technology “shelf life,” adversary intentions, and dual uses of technology complicate these decisions. This seminar will discuss the use of mathematical models and optimization methods to provide insight on technology policy issues. These issues include: balancing risk and affordability during technology research and development; timing technology adoption; and understanding adversary responses to new technologies. Examples will be discussed from offensive cyber operations and synthetic biology. We will conclude by discussing implications for how policy analysts and policy makers think about technology and security.

About the Speaker: Philip Keller is a National Defense Science and Engineering Graduate Fellow at Stanford. He is completing his PhD in Management Science & Engineering. He studies technology policy problems posed by new technologies. His research is highly interdisciplinary, drawing on methods from engineering risk and decision analysis, game theory, and operations research. His professional experience includes conducting studies and analysis for the Department of Defense and the Department of Homeland Security at RAND and the Homeland Security Studies and Analysis Institute. Previous study topics include unmanned aircraft operations; nuclear terrorism; offensive cyber operations; and military force structure. Philip holds a BS in Mathematics and an MS in Defense and Strategic Studies.