Complexity Symposium March 2004
Co-Sponsored by the Santa Fe Institute, 25-26 March 2004 hosted by the
Thursday 25th March 2004 8.30 - 9:00 Registration and Continental Breakfast 09:00 - 13:00 Session I -- Prospects for Complexity Science in Helping Solve Societal Issues Welcome from LSE, the European Commission and SFI. Lord Robert May, Royal Society: Overview Lecture There is an increasing appreciation of the centrality of complex cascades of interactions among various (psychological, social, physical, and technological) elements of human systems and our environment. Dr Barrett will introduce a perspective on foundations of interaction-based systems and interaction-based computing as the natural basis for the simulation and analysis of in large socio-technical systems. The approach is interesting theoretically and also useful in many important ways, and he will emphasize one of them: scaling practical simulations of operating civil infrastructures that involve interactions among very large numbers (106 to 1012) of heterogeneous, interacting entities. Dr. Charles F. Sabel, Professor of Law and Social Science, Columbia Law School Error detection and correction disciplines such as benchmarking, root-cause and near-miss analysis induce new organizations to routinely search for solutions outside the space defined by their routines. The search networks that result appear to be more efficient and more robust--resistant to catastrophic disruption, and conversely capable of innovation--than hierarchies and a range of decentralized alternatives to them. To underscore the practical import of the theoretical ideas Dr. Sabel will focus on the novel governance problems posed by the new organizations, and look at some successful efforts to address them. 14:00 - 18.00 Session II -- Prospects for Complexity Science in Helping Solve Business Prof. W. Brian Arthur, Citibank Professor at the Santa Fe Institute Within the collective of technology, existing elements often become building blocks that build further building blocks--further technologies. Thus Lee De Forest's triode vacuum tube combined with other electronic elements to create the amplifier, the oscillator, the heterodyne mixer, and eventually the logic circuit. These, in their turn, became building blocks in yet further devices: transmission repeaters in telephony, radar, and early computers. Retail performance depends largely upon the retailer's ability to manage the store-customer interface across all stores in a retail network; made more complex by interactions between proximal stores and consumers. Sustained retail success requires simultaneous consideration of the impact of all decisions at all points in the retail network. Simulation models have offered decades of scientific support for managing this complex environment. Optimization methods married to simulation systems now offer a further, significant improvement for managing complexities of the store-customer interface. Evening: Optional Event - Dinner at the Waldorf Hotel, Aldwych at 7.30 p.m. ********* 08.30 - 9:00 Continental Breakfast 09:00 - 17.00 Session III -- Current Research & Future Directions in Complexity Science During the last two decades, the essential role of complex interactions in science, government and business -- and in human interactions generally -- has become very widely appreciated. In parallel, the progress made in developing new analytical and computational tools has made possible some significant progress in many areas. In this talk Dr Eisenstein will outline some potential new directions for the scientific program at the Santa Fe Institute, and its connections to fundamental science, to understanding some aspects of human behavior, and to decisions regarding social policy. Prospects for international collaboration will be discussed as well. Dr. West will explore a quantitative conceptual framework for understanding many of the generic properties of living organisms from molecules and cells to ecosystems. The paradigm presented will be developed as a way of viewing many related phenomena and will include a discussion of the cardiovascular system, trees and plants, growth, aging and mortality, sleep, genome size, cities and corporate structures. While much of economics appropriately focuses on the strategic interactions of agents, there are some situations where this is dominated by other factors, such as market structure. Prof. Farmer will present a model of the continuous double auction based on zero-intelligence noise traders. This can be used to derive simple laws relating order flow to statistical properties of markets, such as volatility and the average bid-ask spread, that agree remarkably well with data from the London Stock Exchange. He will then discuss the effects of adding intelligence, simulating an ecology of arbitrage and exploring its effect on prices. Dr. Luc Steels, Sony Labs, Paris Dr Steels will talk about how the increasing complexity of robots, and the increasing complexity and open-endedness of information technology as experienced in the Web, pushes us to adopt mechanisms whereby communication is no longer pre-programmed but based on emergent conventions and ontologies. These mechanisms are directly inspired by complex systems research and are grounded in complex systems theory. Eric Beinhocker, Senior Advisor, McKinsey & Company, London, SFI Business Network |
