Emergence: The Connected Lives of Ants, Brains, Cities, and Software

Rating:
Summary: Shedding Light to Distributed Phenomena in the World
Review: "Emergence: The Connected Lives of Ants, Brains, Cities, and Software" explores a very important, yet poorly recognized, phenomenon in the world. Drawing lessons from different distributed systems, the author explains how intelligent behaviors and learning can 'happen' from very simplec omponents. An ant in a colony does not know much about its global environment--it just follows certain simple rules. Yet the whole colony can manage its own structure, allocating a suitable area for each purpose, as if there is a central architect who designed it. Moreover, an individual ant lives for only a year or so, while the colony changes its behavior over its lifetime of 15 years. The ant queen, from all we know, is not in any direct way involved in the colony's evolution.

Lessons from the study of ants carry over to other well-known emergent systems. What an ant is to a colony is what a neuron is to a brain, a human to a city, and an individual to an on-line community. The author missed an extremely important example, however: our market economy where resources are organized to produce what individuals need without too much centralized control.

Emergent behaviors and learning cannot just occur in any complex system. Certain principles must be satisfied before it can happen. The book explores these in each of the four chapters in its second section: neighbor interaction, pattern recognition, feedback, and indirect control.

In the last section, the author explains the 'recommender software', just like what's in operation at Amazon, which combines decisions from many individuals, each with only local, limited knowledge, withp attern-recognition software to recommend items to millions of surfers in an intelligent way. Moreover, the author alludes to the possibility that 'emergence' would be an innovative and powerful way to organize many of existing human systems, political, economic, and others.

Overall, this is an insightful and well-researched book. To those already familiar with the concept of emergent behaviors, however, this book might be too basic. Its literary writing style makes the book enjoyable to read, but may obscure some core concepts that can be delivered more succinctly and clearly.

Bottom-line: recommended for those not already well versed in the field of emergence.

Rating:
Summary: Reads like a magazine article
Review: This book attempts to explain artificial intelligence in terms of how ant colonies, cities, and modern software operate. If it seems to have the feel of a magazine article, it's because it's not written by a professional in the field but by a professional writer who is a frequent contributor to trendy, popular publications such as Feed and Wired. Although it did not give me the understanding I was looking for about emergence theory, I would not dismiss it completely because it does have a lot of interesting information, as any good magazine article would. It has an overview of Jane Jacobs new urbanism that is both complete and illustrating, it explains how an intelligent kind of feedback makes some web sites successful as virtual communities, and what I found most interesting, how video games are evolving in ways that seem to give them a life of their own. If you are looking for an insightful, deep look at artificial intelligence for the layman, Douglas Hofstadter's "Godel Escher Bach" is still unchallenged. On the other hand if you are looking for a more relaxed, amusing and down to earth approach, filled with cool stuff you can impress your friends with, this book is for you.

Rating:
Summary: Read All About It!
Review: Emergence is the next big thing. Read everything you can about it. This book does a good job of talking about how complexity theory applies across a wide range of situations.

Rating:
Summary: Bottom-up ideas in a top-down world
Review: Emergence is a solid introduction to how low level decision making can lead to high level, emergent intelligence. It starts off with the life of ants, describing how their sporadic interactions with each other, and the limited decision making capabilities that they have based on those interactions, can lead them to develop complex systems when you zoom out far enough. Johnson then looks at human cities from the same viewpoint, extrapolating recurring behaviors that have evolved from the great cities of the world.

It was an interesting association to attribute ant behaviors to humans, and sparked many ideas in my head for my own correlations.

Last, he described some instances of how the same, low-level decision making entities can be created in software, thus producing intelligent systems in an indirect route. As a developer, this was by far the most interesting part of the book. I would, however, have liked to see him expand on this aspect of emergent behavior a bit more than he did, as it was the most relevant section that I could apply to today's challenges.

Rating:
Summary: light reading... very light
Review: I was disappointed with this book. A little too narrative.
Wanders a bit and is unfocused.
There are absolutely no diagrams, pictures, graphs, graphics
nothing. The topic of emergence is a lot more interesting
than you would get from this book.

Rating:
Summary: Worth the read for the content.
Review: As a person reading this out of curiosity I was slightly turned off by the semi-pompous pace (if that makes sense) and the touting of the author about the subject. But the subject is quite fascinating and the points made are brilliant. I found myself putting this book down to read other similar books in which I connect with the author a little more. This is just a personal issue and I would highly recommend this book otherwise. If the subject matter tickles your curiosity (which it should) then I would go ahead and buy this book

Rating:
Summary: Mediocre At Best
Review: Johnson has a riveting introduction and opening but the rest of the book falls flat with a superficial treatment of emergence. The author would also have the reader think that he knows alot about cities and their development, but his actual understanding of the subject is very, very thin.

Try "Signs of Life" by Richard Sole and Brian Goodwin for a much better elucidation of complexity science and the role of emergence. Another book just out is "Self Organization in Biological Systems" published by Princten University Press as part of its series on complexity science.

Rating:
Summary: Very pleasant, but not too deep
Review: I'm a big fan of Johnson's writing, so I wanted to like this one more than I actually did. I can find something of real interest to me on almost every page : mentions of John Maeda, Danny Hillis, Mitch Resnik and StarLogo, interface design, the Sims; as well as plenty of stuff I have a passing interest in, like open-source software. But it looks like there's an awful lot of recycling of Johnson's work here, like the section on interaction design that mentions Maeda, Will Wright, and Jodi.org which is lifted almost verbatim from Johnson's essay in ID Magazine's 1999 Interactive Design awards issue. Several of his stuff for Salon seems to be cut-and-pasted here as well.

In other places he just hasn't gotten much deeper into topics than I have just by reading around on the web. There's not much more about the emergent properties of online communities (like Slashdot) than I've picked up in the last few years without even really reading that much on the topic, for example. Having read Mitch Resnik's book on StarLogo, there's essentially zero insight to be gained by Johnson's section on the project, other than to see it in the broader context of emergence.

Occasional references to more scholarly topics, like Walter Benjamin, give the book the overall feel of a really fine Master's Thesis: impeccable bibliography and footnoting, but is it really that insightful after all? It's certainly true that emergence is everywhere once you start to look for it, but a more in depth discussion might have been just as fun to read.

Rating:
Summary: A good introduction - and more
Review: I wasn't sure, when I bought this book, how much I would get out of it. The reviews I had read painted it as an introductory work and, since I already know a bit about emergence in the context of ecosystems, economies, social insects and human brains, I wondered if it might be too basic.

What can I say? Having read it, I agree that it is an excellent introduction to the subject: clear, wide-ranging and readable. But it is also far more. Even if you know much more than the author about, lets say, ant nests, the quality of the writing and the constant excursions into other fields to draw illuminating comparisons will keep you reading sections you might otherwise want to skip.

Even the book's style says something about the new sciences of complexity: instead of a linear trail of argument from axiom to conclusion, Johnson's thesis grows by picking out repeated patterns from seemingly unrelated fields, adding resolution like a Mandelbrot set slowly emerging from what at first looks like a random scatter of dots. In one chapter an unpromising section on the pitfalls of discussion groups suddenly backlinks to a previous discussion about city growth, gives a quick blast of Adam Smith, segues into media feeding-frenzies and reprises the theme of feedback mechanisms. By the end I was avidly reading about how some bunch called slashdot.org had dealt with the exponential growth of their Star Wars, programming and related geek stuff discussion group, not a topic that would normally grab me.

Unfortunately, the book does flag in a big way in the last few chapters, unless you're seriously interested in video gaming and the future of passive entertainment. In the author's defence, it must be very hard to write about the future of emergence, since its essence is that you never know what will pop up til your system plays out.

To sum up, Johnson is an engaging, insightful writer. He is particularly strong on the interaction between emergence and selection, realising that emergence in itself is not necessarily adaptive or good. He is sometimes a little weak on the difference between bottom-up organisation and true emergence. Finally, look out for the comparison between scientific revolutions and slime moulds: easily the cutest piece of science writing I have seen lately.

Rating:
Summary: Good introduction, with serious flaws
Review: This book provides a good introduction to the new field of emergence, the study of how complex, apparently "organized" global behaviors arise from the interaction of many autonomous parts operating locally without central control. Steven Johnson explains the principles and brings together many examples from biology (ants, slime molds, neurons) and other areas (games, software, the growth of cities).

Unfortunately, Johnson has not made the effort to study his field thoroughly. He is very familiar with game software (e.g., SimCity), but I was shocked to find no mention of the first analysis of emergent behavior. In his classic "The Wealth of Nations" (1776), Adam Smith coined the term "the invisible hand" to describe the seemingly orchestrated order that emerges from the actions of individuals looking for things they need in a free marketplace. Smith's analysis, by the way, is both detailed and profound--a must for anyone interested in the topic of emergence.

Also, Johnson seems to wander from his central topic at times, for example in the chapter on mind reading.

Despite its gaps and occasional lapses, the book is definitely worth reading. The field is important both socially (do we need a centrally-run society or will the invisible hand work?) and technically. Johnson has done a good job of introducing it.