<Finished SCG Banner2


Upcoming Events

Past Events


Uncertainty in Language and Thought

Prof. Noah Goodman
Stanford University
Tuesday, September 20, 2016 4:00 PM
Building 360

Human cognition is incredibly flexible, partly because common-sense knowledge is uncertain but highly structured. Probabilistic programming languages (PPLs) provide a formal tool encompassing probabilistic uncertainty and compositional structure. I will show that PPLs allow us to model human reasoning and languageunderstanding. I will describe several experimental studies of social cognition, building up to the Rational Speech Act framework for language understanding. This framework captures vague adjectives (“Bob is tall’’), generic language (“Boys are tall’’), hyperbole (“Bob is a hundred feet tall’’), and metaphor (“Bob is a giraffe’’).


The Stanford Complexity Group Presents
Quantitative Trivia and Estimation Night!
Wednesday, May 25th, 6-8pm
Village Center in Escondido Village
140 Comstock Circle
How many shirt buttons were produced last year?
How many insects are there in the world?
How much does the statue of David weigh?
These are the questions that plague no one, but are perfect questions for a night of quantitative trivia, and Fermi-style, order of magnitude estimation! Enrico Fermi had a knack for quickly calculating things using back of the envelope calculations and no regard for a factor of 2. Test your ability to quickly answer strange questions using quick conversions!
Come join the Stanford Complexity Group on Wednesday, May 25th at 6pm at the Village Center, in Escondido Village, 140 Comstock Circle for several rounds of Fermi Estimation.
Come by for fun, refreshments, and to learn more about the Stanford Complexity Group!

Complexity and the Economy

Prof. W. Brian Arthur
Sante Fe Institute
Thursday, December 3, 2015 2:15 PM
Clark Center S361

In the 1980s the Santa Fe Institute began to explore the economy as an evolving complex system. Brian Arthur will talk about the early days of this effort at SFI and explain how this new economics works and why it is needed. Complexity gives a view of the economy not as a perfectly balanced, smoothly functioning machine, but as a system that is organic, evolutionary, ever-changing, and historically-contingent.

Book Club: Probably Approximately Correct

The Complexity Group is hosting a discussion of Probably Approximately Correct: Nature’s Algorithms for Learning and Prospering in a Complex World by Leslie Valiant. This book presents a theory of learning and evolution in a complex world; Marcus Feldman says: “It is erudite, adventurous, difficult, and rewarding. It will be discussed for a long time.” We will meet on June 3rd to discuss the book. Please join us!

A New Path Towards Intelligent Machines

Chris Adami
Michigan State University
Tuesday, May 5 2015 4:00PM
Clark Center Auditorium

For over fifty years, engineers have attempted to achieve machine intelligence that rivals human performance, but with only limited success in some specialized arenas such as chess. I will discuss what I believe is the central reason behind this failure, and how using the biological process of evolution can overcome that problem. I then discuss several applications of our “evolutionary intelligence” approach to understand brains and behavior.

Chris Adami May 5

Can Complexity Improve Public Policy?

Dr. Robert Geyer
Lancaster University
Thursday, November 20

For much of the 20th and early 21st centuries, the focus of public policy has been on creating more ‘order’ in society and public policy. The growth of New Public Management, the audit and targeting cultures in public policy and ‘evidence based’ strategies are the more recent manifestations of this trend. However, as most policy actors recognize, society and policy refuse to be orderly and simple. A different way of looking at and responding to both can be found through the lens of a ‘complexity’ perspective. During his talk, Professor Geyer will briefly outline the influence of the orderly public policy perspective and then introduce some of the main aspects of complexity and how they relate to public policy. He will then apply these insights to a range of UK and European social and health policies and argue that adopting a complexity perspective does not guarantee perfect public policy, but does enables us to better understand its uncertain and emergent behavior and encourages policies that promote open societal growth and development rather than fixed end states and rigid plans.


Dr. Robert Geyer


Book Club

Join the Stanford Complexity Group in reading John Padgett and Woody Powell’s The Emergence of Organizations and Markets, an exposition on a theory of innovation that combines insights from life sciences with new approaches to network analysis.
Preview and order the book at: http://press.princeton.edu/titles/9909.html
We will meet every Wednesday at 12 pm from August 20th to September 10th to discuss key concepts and methods from the book, including:
  • Innovation Theory
  • Autocatalysis and Biological Innovation
  • Multiplex Networks
  • Innovation in Academia

Woody Powell will be joining the discussion in our final meeting, on September 10th.


Percolation and Cascades in Interdependent Networks

Dr. Raissa D’Souza
Professor, Computer Science, Mechanical and Aerospace Engineering, UC Davis
Monday, August 4th

Dr. D’Souza is Professor of Computer Science and Mechanical and Aerospace Engineering at UC Davis. She is an External Professor at the Santa Fe Institute, Managing Editor of Internet Mathematics, Associate Editor of the Journal of Complex Networks, and member of the Editorial Board of Scientific Reports and IEEE Transactions on Network Science and Engineering.

Dr. D’Souza’s research includes networks of networks, controlling cascades in self-organizing systems (see attached DSouza 2013 and DSouza Viewpoint 2013), and microtransitions predicting percolation thresholds (see attached DSouza 2014).


Alfred Hubler, University of Illinois at Urbana-Champagne

Emergence of Functionality in Evolving Physical Networks
April 8, 2014


Self-assembling nonlinear wire networks in complex virtual environments have some amazing properties: They learn to play computer
games such as Tetris, if the network inputs represent the state of the virtual environment and the network outputs control the environment with actuators and if the virtual environment offers energy as a reward for successful controls. The network dynamics appears complicated and irregular, but observables, which classify the parts according to their function in channeling the energy flow through the system, are quite reproducible. Experimental studies suggest that the tendency of networks to evolve towards self-organized critical states, states with large entropy production, and states with minimum resistance drives them to predict and control complex environments. We contrast the computational capacity of these networks with conventional digital computers.

Sandra Mitchell, University of Pittsburgh

Representation, Partiality and Pragmatism or, Why after 50 years are X-ray Crystallographers Still in Business
January 15, 2014

<<No Video Available>>

Abstract: Sir John Henry Kendrew’s 1962 Nobel Prize lecture predicted there would be no need for experimental models of protein folding, once the ab initio algorithms for predicting structure were refined. So far, this has not happened. I will suggest how two features of representation in general, and representing protein folding in particular, lead to the necessity of both pluralism and integration of alternative representations guided by pragmatic goals.


Geoffrey West, Santa Fe Institute

A Unifying Framework for the Dynamics and Structure of Organisms, Ecosystems, Cities and Companies:from Metabolism, Growth and Mortality to Cancer and Sleep
June 20, 2013


This talk explores ramifications of a coarse-grained quantitative framework for understanding many structural and dynamical phenomena associated with biological and socio-economic systems from organisms to cities. It rests on the observation that, despite the extraordinary diversity and complexity of life, almost all life-history and physiological variables such as metabolic rates, lifespans, genome lengths, DNA nucleotide substitution rates, wages, profits, patents, crime, disease, and road lengths scale with size in a simple and approximately “universal” fashion across a huge range of spatio-temporal scales. This suggests that these systems are constrained by underlying principles that transcend evolved design. The scaling laws are derivable from generic dynamic and geometric properties of optimized space-filling, fractal-like, networks that sustain life, leading to a framework that can potentially capture many essential features of these systems. This will be used for understanding growth, aging, mortality, cancer and sleep based on the role of metabolism as fueling growth and sustaining life, yet also as the major source of damage, repair and entropy production. This will be extended to discuss why all companies, like organisms, die yet cities remain viable and the pace of life increases leading to speculations regarding innovation and global sustainability.


Jessica Flack, University of Wisconsin at Madison and the Santa Fe Institute

Collective Computation
April 19, 2013


We will discuss a set of computational techniques, called Inductive Game Theory, for extracting strategic decision-making rules from time-series data and constructing social circuits. These circuits, which capture the collective implementation of the strategies, serve as hypotheses about how microscopic dynamics generate functional macroscopic properties of social organization.

Iain Couzin,

Princeton University
From democratic consensus to cannibalistic hordes: the principles of collective behavior
April 3, 2013

Jean Pierre-Dupuy,

Stanford University
Complexity and the metaphysics of time
Februrary 13, 2013

Lecture Slides



Contrary to what politicians claim, we do not “create our future.” It is no less true that the future is not “already written”. I will present a metaphysics of time that is neither constructivist nor fatalistic. It is born when a community of people is pulled forward by an image of the future that it has projected in front of itself, an image that, once overtaken by events, becomes a part of reality. This is a form of bootstrapping or, as I say, of self-transcendence – a key figure in complexity theory that has been fleshed out by thinkers as diverse as social philosopher Friedrich Hayek or neuroscientist Francisco Varela.

If time permits, I will apply this view of the future to the current economic crisis, to the issue of climate change and the precautionary principle, and to the mad logic of nuclear deterrence.

Bernardo Huberman, Hewlett Packard Laboratories

Social media and the economics of attention
October 18, 2012

Lecture Slides


The past decade has witnessed a momentous transformation in the way people interact and exchange information with each other. Content is now produced, shared, classified, and rated on the Web by millions of people, while attention has become the ephemeral and valuable resource that everyone seeks to acquire.  This talk will describe the role that social attention plays in the production and consumption of content within social media, how it can be used to predict future trends and the how it determines the public agenda.


Michael Savageau, University of California at Davis

Genotype to Phenotype: Deconstructing Complex Systems
October 5, 2012

Lecture slides


Throughout the pre-genomic era there was sustained interest in the relationship between genotype and phenotype. However, the announcement of the draft sequence of the human genome revealed the true magnitude of the challenge. Although we now have a generic concept of ‘genotype’ provided by the detailed DNA sequence, there is no corresponding generic concept of ‘phenotype’. We have only some intuitive notions of what is meant by phenotype at the level of the organism: hair color of cats, shape and size of flowers, height and weight of livestock, not to mention disease states in humans. However, without a generic concept of phenotype there can be no rigorous framework for a deep understanding of the complex nonlinear systems that link genotype to phenotype. Achieving predictive understanding of complex nonlinear systems, such as those manifested at various levels of biological organization, represents an enormous challenge. The task could be greatly facilitated if such systems could be generically decomposed into a series of tractable subsystems and the results of their analysis reassembled to provide insight into the original system. My colleagues and I have developed such an approach in which the subsystems are integrated into a system design space that allows qualitatively distinct phenotypes of the complex system to be rigorously defined and counted, their relative fitness to be analyzed and compared, their global tolerance to be measured, and their biological design principles to be identified. This approach will be described in the context of the genotype-phenotype question for a couple of well-studied systems.

Melanie Mitchell, Portland State University and the Santa Fe Institute

Using analogy to discover the meaning of images
September 13, 2012

Enabling computers to understand images remains one of the hardest open problems in artificial intelligence. No machine vision system comes close to matching human ability at identifying the contents of images or visual scenes or at recognizing similarity between different scenes, even though such abilities pervade human cognition. In this talk I will describe research—currently in early stages—on bridging the gap between low-level perception and higher-level image understanding by integrating a cognitive model of pattern recognition and analogy-making with a neural model of the visual cortex.

Walter W. Powell, Stanford University

The problem of emergence
May 17, 2012

Dr. Powell spoke from his forthcoming book:

John F. Padgett and Walter W. Powell, The Emergence of Organizations and Markets, Princeton University Press, 2012

“The social sciences have sophisticated models of choice and equilibrium but little understanding of the emergence of novelty. Where do new alternatives, new organizational forms, and even new types of people come from? Combining biochemistry insights about the origin of life with historically oriented social network analyses, Padgett and Powell develop a theory about the emergence of organizational, market, and biographical novelty from the co-evolution of multiple social networks. Their mantra is that in the short run actors make relations, but in the long run relations make actors. Organizations, markets, and people are conceptualized as intersecting flows of reproducing products, skills, and language. Novelty is spillover and tipping in the mutually catalytic constructions of underlying multiple social networks and reproductive flows. This theory of the emergence of novelty as a byproduct of life is developed both through formal deductive modeling and through an exceptionally wide range of careful and original historical case studies. Early modeling chapters in the book extract and develop the biochemical concept of autocatalysis, which is the chemical definition of life. Those chapters extend and apply constructivist thinking from chemistry to social processes of production and communication. A second set of chapters presents historical case studies of the emergence of organizational novelty in early capitalism and state formation: namely, medieval banking, Renaissance partnership systems, the stock market in early modern Netherlands, and the construction of nineteenth century Germany. Spillover and feedback among multiple dynamic social networks are stressed. A third cluster of chapters offers historical case studies of the co-evolution of political mobilization and economic reform in pre- and post-communist Russia, China, and Eastern Europe. A final set of case studies focus on contemporary high-tech capitalism. Those chapters analyze the intertwining networks of science, finance, and commerce to explain the emergence of the biotechnology industry, regional high-tech clusters, and the open-source community.”

Terrence Deacon, University of California at Berkeley

Incomplete Nature: radically reformulating the concept of emergence.
May 3, 2012<

Lecture slides

By recasting emergence theory in dynamical terms and focusing on the role of constraint in explaining ascending levels of causal relationships, this talk will argue that such basic scientific challenges as explaining the origins of life, the nature of information, and the dynamics of mental function need to be rethought. Though relying heavily on complex systems approaches, I will argue that various approaches to biological and neurological processes that depend on models of self-organizing dynamics are fundamentally incomplete, and that a higher order emergent dynamical approach is necessary, which I call teleodynamics.

Deacon, TW (2012). Incomplete Nature: How Mind Emerged from Matter . W.W. Norton


Walter J. Freeman, University of California at Berkeley

How brains create knowledge and meaning from fragments of information.
Octobery 27, 2011


Commonly we sniff, glance, palpate, or hark some event or object and experience a sudden flash of meaning: we know what it is. This experience highlights the difference between sensation, which is the acquisition of fragments of information by intentional search, and perception, which is the contextualization of the acquired information: every fragment is related interactively to every other, creating a spatial pattern. Neural networks implemented on digital platforms are fully competent to simulate the dynamics by which sensory systems extract and refine fragments of information and store them for future use. Examples are provided by ‘concept cells’, once known as ‘grandmother cells’ and now as ‘Jennifer Aniston cells’, hippocampal ‘place cells’, ‘mirror neurons’, etc. They are observed as spike trains that carry meanings for observers, whether or not they do so for the subjects in whom they are observed. Attempts to simulate the processes of perception with neural networks have foundered in combinatorial explosions. An alternative mode of cortical dynamics has been brought to light by measuring and modeling the spatial patterns of the epiphenomenal EEG potentials that are caused by the dendritic currents that control myriad spike trains. The textures in these activity patterns reflect maintenance by neocortex of a global superhighway through which every neuron can interact with every other neuron simultaneously in transitory frequency bands. I postulate that a set of vector fields formed by synaptic interactions provides the multi-tiered matrix that is needed for context construction in the emergence and expression of knowledge. In my talk I will show some examples of the scale-free, self-organized patterns ranging in size from the olfactory bulb of rats to the entire human scalp, and explain how they emerge and what role they play in the action-perception cycle.

Freeman WJ (2001) How brains make up their minds. New York: Columbia UP

Deborah M. Gordon, Stanford University

The regulation of foraging activity in harvester ant colonies.
May 31, 2011
Ant colonies operate without central control and resemble large distributed systems. An ant’s behavior depends on its recent experience of brief interactions with other ants. In the course of a brief antennal contact, one ant can assess the task of the other using odor cues. A long-term study of the behavior and ecology of harvester ants in the Arizona desert shows how colonies regulate foraging to balance the tradeoff imposed by spending water, while foraging in the desert sun, to obtain water, which is metabolized from seeds. The goal is not to send out more ants than are justified by the current food supply. The ants collect seeds that are widely scattered, each retrieved by a single ant without the use of pheromone trails. The duration of a foraging trip depends mostly on how long the forager had to search to find a seed. A forager leaves the nest on its next trip in response to the rate at which it meets foragers returning to the nest with food. Thus foraging activity is adjusted to food availability without any information about the location of food. I will discuss a model of the algorithm colonies use to regulate foraging, the ecological and evolutionary consequences of variation among colonies, and analogies to other distributed networks.