Register on eventbrite: https://www.eventbrite.co.uk/e/book-talk-programmed-inequality-tickets-34469086967

Marie Hicks is a historian of technology, gender, and modern Europe. She researches how gender and sexuality change what we think we know about technological progress and the global "computer revolution." Her new book is Programmed Inequality: How Britain Discarded Women Technologists and Lost Its Edge In Computing published by MIT Press in January 2017. It investigates why the proportion of women declined as electronic computing matured, and how this labor situation had grave effects on the technological aspirations of that waning superpower. It shows what lessons this holds for other nations, especially the United States.

Registration required: http://www.sbs.ox.ac.uk/faculty-research/entrepreneurship-centre/events

June 1st 2017 — AI + Management: when your boss is a bot

As more activities in society run via algorithms, what is the role of humans to “manage” or “govern” the systems? How do we enshrine our goals and values in the technology, and ensure that ethics and transparency exist (especially as AI systems cloud causality in return for efficiency)? Privacy, antitrust, accountability, open data and the role of government to create markets are essential public and private sector interests. What recourse should individuals and companies have in the face of adverse decisions?

“Salvo” talk 1: Jonathan Trevor (of Oxford’s Saïd Business School) on adapting organizations for AI-augmented management.

Pre-readings:

Andrew McAfee and Erik Brynjolfsson. “Human Work in the Robotic Future: Policy for the Age of Automation,” Foreign Affairs. July/August 2016. https://www.foreignaffairs.com/articles/2016-06-13/human-work-robotic-future

Sarah O’Connor. “When your boss is an algorithm.”Financial Times, September 8, 2016. https://www.ft.com/content/88fdc58e-754f-11e6-b60a-de4532d5ea35

Also recommended:

Thomas H. Davenport and Julia Kirby. “Beyond Automation,” Harvard Business Review. June 2015. https://hbr.org/2015/06/beyond-automation

Min Kyung Lee, Daniel Kusbit, Evan Metsky, Laura Dabbish. “Working with Machines: The Impact of Algorithmic and Data-Driven Management on Human Workers,” Proceedings of the ACM/SIGCHI Conference on Human Factors in Computing Systems (CHI 2015), 1603–1612. https://www.cs.cmu.edu/~mklee/materials/Publication/2015-CHI_algorithmic_management.pdf

Kroll, Joshua A. and Huey, Joanna and Barocas, Solon and Felten, Edward W. and Reidenberg, Joel R. and Robinson, David G. and Yu, Harlan. “Accountable Algorithms,” University of Pennsylvania Law Review, Vol. 165, 2017, March 2, 2016. Fordham Law Legal Studies Research Paper №2765268. https://ssrn.com/abstract=2765268

Please register on Eventbrite: https://www.eventbrite.co.uk/e/oxcrowd-from-tags-to-tools-tickets-33456487254

Direct Anonymous Attestation (DAA) is a protocol that allows a security chip embedded in a platform such as laptop to authenticate itself as a genuine chip. Different authentications are not linkeable, thus the protocol protects the privacy of the platform. The first DAA protocol was proposed by Brickell, Chen, and Camenisch and was standardized in 2004 by the Trusted Computing Group (TCG). Implementations of this protocols were rather slow because it is based on RSA. Later, alternative and faster protocols were proposed based on elliptic curves. Recently the specification by the TCG was updated to allow for DAA protocols based on elliptic curves. Unfortunately, the new standard does not allow for provably secure DAA protocols. In this talk, we will review some of the history of DAA and then discuss the latest protocols, security models, and finally a provably secure realization of DAA based on elliptic curves.

https://www.epsrc.ac.uk/funding/calls/tipsworkshopeoi/

Andrew Martin, Marina Jirotka attending

Please let max.van.kleek@cs.ox.ac.uk know if you'd like to meet Ben while he's here.

Abstract
Event Analytics is rapidly emerging as a new topic to extract insights from the growing set of temporal event sequences that come from medical histories, e-commerce patterns, social media log analysis, cybersecurity threats, sensor nets, online education, sports, etc. This talk reviews our decade of research on visualizing and exploring temporal event sequences to view compact summaries of thousands of patient histories represented as time-stamped events, such as strokes, vaccinations, or admission to an emergency room. Our current work on EventFlow supports point events, such as heart attacks or vaccinations and interval events such as medication episodes or long hospitalizations. Demonstrations cover visual interfaces to support hospital quality control analysts who ensure that required procedures were carried out and clinical researchers who study treatment patterns that lead to successful outcomes. I show how domain-specific knowledge and problem-specific insights can lead to sharpening the analytic focus so as to enable more successful pattern and anomaly detection.

Speaker bio
Ben Shneiderman is a Distinguished University Professor in the Department of Computer Science, Founding Director (1983-2000) of the Human-Computer Interaction Laboratory, and a Member of the UM Institute for Advanced Computer Studies (UMIACS) at the University of Maryland. He is a Fellow of the AAAS, ACM, IEEE, and NAI, and a Member of the National Academy of Engineering, in recognition of his pioneering contributions to human-computer interaction and information visualization. His contributions include the direct manipulation concept, clickable highlighted web-links, touchscreen keyboards, dynamic query sliders for Spotfire, development of treemaps, novel network visualizations for NodeXL, and temporal event sequence analysis for electronic health records.

Ben is the co-author with Catherine Plaisant of Designing the User Interface: Strategies for Effective Human-Computer Interaction (6th ed., 2016). With Stu Card and Jock Mackinlay, he co-authored Readings in Information Visualization: Using Vision to Think (1999). His book Leonardo’s Laptop (MIT Press) won the IEEE book award for Distinguished Literary Contribution. He co-authored Analyzing Social Media Networks with NodeXL (2010) with Derek Hansen and Marc Smith. Shneiderman’s latest book is The New ABCs of Research: Achieving Breakthrough Collaborations (Oxford, April 2016).

Convenor: sarah.green@st-hildas.ox.ac.uk

‘Smart contracts’ entail autonomous drafting, execution and verification of contractual agreements. This has important implications for: (1) Private law: what input will private lawyers have to such contracts? How relevant is the traditional ‘reasonable person’ heuristic for contractual interpretation for an agreement that is executed without human agency? (2) Regulation: how can and should regulators oversee the deployment of smart contracts to avoid the production of socially harmful consequences?

Register: http://www.bsg.ox.ac.uk/node/3800

Understanding deterrence and dissuasion in cyberspace is often difficult because our minds are captured by Cold War images of massive retaliation to a nuclear attack by nuclear means. The analogy to nuclear deterrence is misleading, however, because many aspects of cyber behavior are more like other behaviors such as crime that states try (imperfectly) to deter. Preventing harm in cyberspace involves four complex mechanisms: threats of punishment, denial, entanglement, and norms. Even when punishment is used, deterrent threats need not be limited to cyber responses, and they may address general behavior as well as specific acts. Cyber threats are plentiful, often ambiguous, and difficult to attribute. Problems of attribution are said to limit deterrence and dissuasion in the cyber domain, but three of the major means – denial by defense, entanglement; and normative taboos – are not strongly hindered by the attribution problem. The effectiveness of different mechanisms depends on context, and the question of whether deterrence works in cyberspace depends on “who and what.” Not all cyberattacks are of equal importance; not all can be deterred; and not all rise to the level of significant national security threats. The lesson for policymakers is to focus on the most important attacks and to understand the full range of mechanisms and contexts in which they can be prevented.

This event is free and open to all.

Registration required: jonathan.lusthaus@sociology.ox.ac.uk

This conference serves to highlight the social dimensions of cybercrime. With the focus often on the more technical aspects of cybersecurity, the human actor can sometimes be forgotten. Organised by the Human Cybercriminal Project of the Extra-Legal Governance Institute, this event aims to bring together leading thinkers in this area from across academia, journalism, government and the private sector.

Many approaches investigating cybercrime thus far have focused on the technological elements, but there is now a great need to understand the “human” side of online crime: Who are the cybercriminals? What motivates them? How do they operate? What can be done to address the threat? Drawing together experts from across academia, law enforcement and the private sector, the aim of the conference is to shed light on these vital, but not widely studied, issues. It will build on some initial insights developed in the first Decoding the Human Conference held in 2015.

We welcome contributions from a wide field of disciplines including cyber security, socio-legal studies, anthropology, psychology, criminology as well as political science, economics and sociology. This conference will be held under the Chatham House Rule.

Abstract:

Long before self-driving cars arrived on our roads, humans have worried about robots and other human-created intelligent beings. The Sorcerer’s Apprentice, Frankenstein’s monster, and many other fictional scenarios warn of intelligent creatures that follow our instructions literally, failing to understand moral and ethical constraints, with catastrophic results.

To understand how to design moral and ethical constraints into robots, we need to understand the roles that these social norms play in human society. Society depends on cooperation among its individual members. Cooperation requires trust, that partners will not exploit each other’s vulnerability. Society imposes social norms to lead individuals away from attractive selfish decisions, toward better outcomes for everyone.

The long-term welfare of society may be the ultimate purpose of morality and ethics, but it is not a practical criterion for an individual, human or robot, to decide how to act in the current situation. Moral and ethical reasoning takes place at multiple time-scales: some situations require immediate action; deliberative reflection takes longer but can improve future decisions; children learn and develop morally as well as cognitively from instruction and experience; and the moral standards of society itself evolve over historical time.

To design representations for moral and ethical knowledge, and algorithms for its learning and use, we can draw on useful ideas from philosophical theories of ethics (utilitarianism, deontology, virtue ethics) and corresponding computational theories in AI (game theory, rule- and constraint-based systems, and case-based reasoning). No one theory is adequate, but we can begin to sketch a hybrid design.

For robots to be acceptable participants in human society, they will need to understand and follow human social norms. They also must be able to communicate that they are trustworthy in the many large and small collaborations that make up human society.

Speaker biography:

Benjamin Kuipers is a Professor of Computer Science and Engineering at the University of Michigan. He received his B.A. from Swarthmore College, and his Ph.D. from MIT. He served as Department Chair at the University of Texas at Austin, and is a Fellow of AAAI, IEEE, and AAAS. He investigates the representation of commonsense and expert knowledge, with particular emphasis on the effective use of incomplete knowledge. His research accomplishments include developing the QSIM algorithm for qualitative simulation, the Spatial Semantic Hierarchy models of knowledge for robot exploration and mapping, and methods whereby an agent without prior knowledge of its sensors, effectors, or environment can learn its own sensorimotor structure, the spatial structure of its environment, and its own object and action abstractions for higher-level interactions with its world.

Registration: Please email events@oii.ox.ac.uk to register

The design and implementation of social media platforms has put several advanced democracies into a kind of democratic deficit. First, social algorithms allow fake news stories from untrustworthy sources to spread like wildfire over networks of family and friends. Second, social media algorithms provide very real structure to what political scientists often call “elective affinity” or “selective exposure”. We prefer to strengthen our ties to the people and organizations we already know and like. Third, technology companies, including Facebook and Twitter, have been given a moral pass on the normative obligations for democratic discourse that we hold journalists and civil society groups to. Using evidence from the ERC-funded Consolidator Award on Computational Propaganda (COMPROP, www.politicalbots.org), I discuss the ways in which social media platforms have become they key infrastructures for political discourse, identify how these technological affordances have put us into a democratic deficit, and conclude with some ideas about ways in which social media platforms could be a better infrastructure for deliberative democracy.

(followed by a drinks reception in the Masters Dining Room)

ABSTRACT: Technology is rapidly evolving and enriching the modern world with numerous of distributed cyber physical networks and so the applications. Many of the applications are integrated with extremely low-cost networked computers that add eye and ear (i.e., sensor) and arm, leg and voice (i.e., actuator) to perform their intended tasks efficiently. Inevitably, much of sensor nodes information are highly sensitive and critical, and thus security is a critical concern. In this seminar, I will provide an overview of my recent and ongoing research, i.e., design and analysis of security schemes in distributed cyber physical networks and applications.

Registration required: claire.bunce@nuffield.ox.ac.uk

ABSTRACT: The proliferation of algorithms and AI systems is accelerating across the public (e.g. healthcare and criminal justice) and the private (e.g. finance and insurance) sectors. Automated decision-making systems are being more and more frequently used to make assessments or predictions about people. This new trend could lead to less biased and more accurate decisions, but also lead to inadvertent discrimination and privacy violations. Further, the decision-making process and effects of algorithmic decisions are often unclear, owing to the inherent opacity and inaccessibility of systems. There is increasing effort from the public and the private sectors to make these technologies more explainable and ultimately accountable. This talk will first consider the ethical challenges facing complex algorithmic systems concerning opacity, discrimination, group privacy, and moral responsibility for failures and harms. Ethical considerations often ground legal and regulatory action; recognising this, the second half of the talk will examine how these challenges are being addressed in European and American regulation, in particular concerning new provisions the European General Data Protection Regulation that pressures designers and users to explain how complex algorithmic systems function.

Convenor: john.armour@law.ox.ac.uk

How might smart technology impact corporate governance? (1) ‘Tech assets’ are opaque: firms keep them secret, and compiled code is hard to reverse engineer. Asset opacity lends itself to fraud, necessitating special monitoring efforts. (2) ‘Data assets’ create existential risks for firms. What determines the consequences of data breaches, and how can firms guard against them? (3) Automated systems tend to reduce human agency. While this increases productive efficiency, it also underlines the need for legitimation of organisational goals.