Posted by cadsmith on June 4, 2010
Responses to complexity include modeling and innovative technology. Networks provide computational and social leverage. Tools are adaptive to realtime, combinatorial, fractal, and quantum considerations. There are various opinions about where all this may be leading, with respect to order or limits for example, and what degrees of freedom can be exercised.
Recent Links: (of about 23): 3D: car wrap; semantic web: RDFa checker; robotics: transportation; mobile: local ads; security: civilian net lockdown; surveillance: text stream, electrical network frequency analysis; tracking: eye movement, sleep monitor; quantum: simulation; complexity: matrix decomposition; business: internet of things, travel guide; finance: investing dashboard.
The author covers the field in a readable narrative, rather than mathematical, fashion. There is no common measure of complexity since theory and science are still undefined. Research involves interdisciplinary collaboration. It is compared to cybernetics which had more extent than content, though this is more mainstream. There are both adaptive and nonadaptive complex systems. A proposed definition is “a system in which large networks of components with no central control and simple rules of operation give rise to complex collective behavior, sophisticated information processing, and adaptation via learning or evolution”.or briefly “a system that exhibits nontrivial emergent and self-organizing behaviors”. Measures include entropy, algorithmic information content, logical depth, thermodynamic depth, computational capacity, statistical or effective measure complexity, fractal dimension, degree of hierarchy and near-decomposability. Some new areas of research are listed, e.g. self-organized criticality and computational mechanics. These fall into two groups, either more specific applications, or higher level mathematical theories. Historically, emergence arose as a reply to reductionism. Computers mimic evolution and nature mimics computation. Network thinking is more concerned about relationships than entities. The web is a scale-free network. This also occurs in physiology since cells do not scale with body size; space is filled using a fourth dimension of fractal circulatory networks. Ecology extends the food chain to food web. The book has five parts for nineteen chapters, a bibliography of several hundred authors, and extensive notes and index. It is dedicated to Douglas Hofstadter, her doctoral adviser for analogy making programs, and John Holland for genetic algorithms.
The author is a pioneer of network science as a modeling activity that combines complex adaptive systems, chaos, and mean-field theory. This text is dense mathematically and includes Java code. There are thirteen chapters with exercises, a bibliography, and index. The history of significant events is outlined from Euler’s Bridges of Konigsberg in 1736 to Gabbay in 2007. Topics include structure, emergence, dynamism, autonomy, bottom-up evolution, topology, power, and stability. Graph theory describes properties, matrix representation, classes, modeling and simulation. Regular networks are constructed by a generative procedure. Network-centric organizations reduce links and path lengths to lower costs and latency. A new metric, link efficiency, compares network types. Entropy initially increases as nodes are added, flattens, then diminishes to zero as structure predominates. Networks have topological phase transitions as rewiring probability increases. Network emergence describes macroscale properties resulting from microscale rules. Hub emergence is not scale-free. Cluster emergence is not small world. Feedback-loop, adaptive or environmental ermergence connects the next state to input microrules on goal-oriented networks. A network epidemic, characterized by spectral radius, propagates state or condition via links, as do antigen countermeasures which use superspreaders to decrease time and peak incidences. The classic is the Kermack-McKendrick model from 1927. Networks which follow Kirchhoff’s first law are shown where commodity flow in and out is equal. Influence networks are great models of social networks where nodes are actors. Network vulnerability is the probability that an attempted attack will succeed. Strategies such as linear are good defender and exponential for attacker. Risk is reduction of vulnerability or consequence. Resilience is defined for links, where small-world has highest followed by random then scale-free, as well as for stability, and flows where expected flow is availability times actual flow. Percolation adds links, depercolation removes them. Game theory assumes independent success probabilities. The attacker-defender problem is asymmetric. Netgain is a property where nodes compete for value proposition such as preferential attachment. Multiproduct emergence shows how shakeouts and monopolies can occur. Other market emergence types include nascent, creative destructive, or merger and acquisition. Network science can be used to model metabolism. Biology includes protein expression using Boolean networks. Chemistry uses bounded mass kinetic networks. Readers interested in quantum mechanics would seek additional sources.
A theme of this book is that scientific common-sense shows that a long-term perspective is essential and that particular quantitative, technical formulations reveal behaviors which apply to many areas. Rather than verbosity, illustrations from network theory are used, such as graph statistics and probability generating functions. There are seven chapters, each having exercises and further readings. Ideas are linked across fields and demonstrated by examples, e.g. oscillators, neural nets or epidemics. The brain is the most complex adaptive system and life is an adaptive network. Complexity theory is a tool for modeling and scenarios, and is used for futurology, e.g. universal prediction tasks. The origin of life involved molecular cooperation in autocatalytic networks. Fitness landscapes are a function of chances of survival for species. Coevolution involves effects across multiple species within time and space scales, resulting in a red queen phenomenon of running in place. Kauffman gene regulation is an example of a random Boolean network. Critical variables dominate the dynamic at point of phase transition, e.g. temperature. This results in punctuated equilibrium and synchronization phenomena exhibited by evolving realworld networks. An order parameter measures degree of asymmetry. There is a small world effect where distance between nodes is a small fraction of the number of nodes. In social networks, diffusion has a role in transport, reported in the 1960s by Milgram. Game theory looks at survival strategies. Adaptive systems alternate between absorbing energy and dissipation.
This book’s thesis is that the universe is becoming alive according to a Selfish Biocosm hypothesis which is consilient, falsifiable and retrodictable. Emergence is the controller and intelligence the copier. Some of it is related to the Singularity as outlined in the forward by Kurzweil, or to a “deep DNA” universal genetic code. It looks at many types of artificial life research, and various alternatives to singularity theory such as Virtual Cambrian or Omega point. Humanity becomes the missing link, but its track record of maintaining lesser species is hopefully not repeated up the chain. The future of religion may include cosmotheology, becoming a subject of scientific study, or a “biocosm aborning”. There is a chart of NASA’s mission telescopes to observe black holes, dark matter and the big bang. Rather than “the long hello” of direct communications, SETI would look for computational meaning of seemingly natural noise, prediction markets, artificial exo-society modeling, or artifacts of cosmic macroengineering. An Intelligence Principle results in a post-biological universe.
The style is narrative. There are three parts, nine chapters, an afterword, three appendices, notes, bibliography, and index. It quotes from other scientists’ publications at length. Shaded boxes highlight key concept definitions and explanations, e.g. a notion of quantum evolution.
Robb on US censorship Commented: A general risk is revealed. May be reminded of notion of ecotechnology in Steward Brand’s “Whole Earth Discipline”. If they are going to treat the planet like a gas station, then a spill side-effect suppression system would be nice, e.g. by a genetic engineering solution capable of changing crude to something friendlier. Of course, this becomes a controlled substance to avoid a Vonnegut Ice 9 scenario where green refinery is used to vanish all the reserves.
The Secret Life of Chaos (Part 1/6), Jim Al-Khalili, 2010 on Turing morphogenesis, Belousov nonlinear chemical oscillator, Lorenz chaos theory, Mandelbrot fractal geometry.
Irreducible Complexity 01/04, The Cassiopeia Project, 2010 adds quantum mechanics to organic chemistry and biology, and amino acids from interstellar gas clouds.
Authors@Google: Christos Papadimitriou, 2010 wrote Logicomix and researches algorithms and complexity.
An Evening with Dr. Atul Gawande, 2010 wrote The Checklist Manifesto to handle complexity.
The Most Exhilarating Ode to the Future You’ll See All Day (Batteries Not Included) | Motherboard on Singularity