The Center for Research on Computation and Society continues its weekly lunch seminar:

CRCS Privacy and Security Lunch Seminar
Date: Wednesday, 12 December 2007
Time: 12:00pm-1:30 pm
Place: MD 221

Topic: Cooperation and Human Systems Design

Speaker: Yochai Benkler, Jack N. and Lillian R. Berkman Professor for Entrepreneurial Legal Studies at Harvard Law

Abstract:

Globalization and rapid innovation cycles make the social and economic environment more complex and harder to characterize for planning or pricing. In response, we see adoption of loosely-bound, permeable human systems— technical platforms, business processes, and institutional devices—that enable pervasive experimentation and learning through decentralization of practical capacity and authority to act. Providing such practical freedom for human agency creates new challenges in design for cooperation. Doing so requires attention to work in social and biological sciences, political science and business management, that diverges from dominant interpretations of human action as selfishly motivated, and developes a more cooperative view of human nature, human interaction, or both. Observed heterogeneity of motivational profiles and practices of sustained cooperation suggests the potential for design aimed not at aligning individual selfish incentives, but at enabling the dynamics of self-reinforcing, cooperative social-psycological processes.

(more…)

The Center for Research on Computation and Society continues its weekly lunch seminar:

CRCS Privacy and Security Lunch Seminar
Date: Wednesday, 5 December 2007
Time: 12:00pm-1:30pm
Place: Maxwell Dworkin119

Topic: Protecting the Privacy of the Users of Wireless Devices

Speaker: David Wetherall, Intel Research and University of Washington

Abstract:

We have started a new research effort to build wireless systems that provide a high degree of privacy for the users of mobile devices. Existing wireless protocols such as 802.11 transmit unique identifiers, e.g., MAC addresses, that allow users to be tracked and profiled by any nearby observer; they do not provide security models that work well in unmanaged environments. This is becoming problematic as wireless devices become more ubiquitous and more personal (with the proliferation of mobile phones, personal fitness and medical devices, headsets, and consumer electronics) and security problems become more pressing (with the rise in identity theft and unintended disclosures).

We are developing techniques that selectively disclose addresses and other distinguishing information that maps to high-level identities, and which restrict connectivity to intended service regions. The former is challenging because addresses play a basic role in protocols such that they cannot be concealed without impact; traditional encryption methods such as WPA2, IPSEC and SSL do not prevent tracking and profiling. The latter is challenging because wireless signals propagate in unpredictable ways and leak across boundaries in the physical world.

(more…)