by Steven Johnson
Ants — and slime mold — are more interesting, and more intelligent, than I ever would have guessed. Their intelligence isn’t the same as human intelligence, as I wouldn’t exactly say slime mold or ants are sentient. What they do possess is a kind of “swarm intelligence” or emergence.
Emergence is a term for describing a self-organized, ‘whole is greater than the sum of the parts’ type situation. That is the case with slime mold, which functions both as tiny individual specks and a coagulated mess of well, slime. The individual bits, without a leader or any recognizable capacity for intelligence, nevertheless communicate with each other, and coalescing and breaking apart again as environmental conditions warrant.
Johnson spends a fair amount of time disabusing his readers of the notion of the necessity for individually-intelligent components. There doesn’t need to be a slime leader telling all the other tiny bits of slime what to do. The queen ant doesn’t actually order the other ants around; they follow pheromone trails and act in the colonies’ best interest. Ants can follow these pheromone trails to food sources, but they are also used by individual ants to figure out if it is a good idea to switch from foraging duty to garbage disposal. Ant colonies can be a dozen years old, and older ant colonies are “smarter” than younger colonies — even though the only ant who lives longer than a year is the queen. (In some species, the males live for so short a time, they are born without jaws to chew food with.) Humans have a hard time seeing this kind of intelligence, and of embracing emergence, because emergence is about lack of control, of letting systems regulate themselves, and evolving from the consequences — not how people in groups usually approach problem-solving.
Cities exhibit the same kind of group intelligence. For evidence, talks about how neighborhoods work — the silk district of Florence being in the same location for hundreds of years for instance, or the importance of the sidewalk as communications routes. Businesses clustering together benefit from the economics of agglomeration, yet they tend to evolve naturally, and not by urban planning from on high. Johnson relies on Jane Jacobs’s The Death and Life of Great American Cities for much of his argument here. Human beings can easily miss the emergent intelligence of cities, because we think on too-small a scale. Decades may not reveal the patterns in the chaos, but a hundred years or 500 years will.
Software that has been produced by generations of evolution instead of hours of caffeine-induced coding is an interesting idea. It has been done, though it remains to be seen if growing “code gardens” will catch on. Some software seeks to model the real-world behavior of slime mold — and then there are the Sims. The Sims is based on emergent principles though emergence needs to be hobbled for a good game, as otherwise the player wouldn’t have much to do.
Johnson raises some great questions, and his willingness to write about the web — which mutates much faster than a book can go from green light to paperback — is important, yet is it also where the ground is less firm. He refers to the web as “feedback-intolerant” and links working as one-way pointers. While this is true for much of the web, it isn’t true for the whole web. While I am loathe to appear in the “blogs will change world” camp, I do think they have potential for changing the web. I would not describe successful blogs as feedback-intolerant, or blind to incoming links: successful blogs can generate communication. This book was first published in 2001 — before the “bloggers as journalists” meme heated up. It would be interesting to see what Johnson would say about blogs, and what questions he would think to ask about, and of them.
He ends his book with an acknowledgement of the best questions are nowhere near being answered:
Are there new scales to conquer, new revolutions that will make the top-down revolutions of the industrial age look minor by comparison? On the hundred-year scale, or the scale of millennia, there may be no question more interesting, and no question harder to answer.
Emergence is a good book that generates questions, opens a new perspective on intelligence, and will make you think twice (at least) about unintended consequences. Definitely recommended.