Weekly Blog Post 1: The Convergence of Emergence
In parsing through the readings of the past two weeks, both Lawrence Lessig and Steve Johnson appear to travel from opposite ends of the spectrum regarding the evolution of the internet in their respective books Code and Emergence. Lessig focuses on the architecture of the web as the key to providing it more usefulness to society, in particular concentrating on the ability of that architecture shape what human beings use it for and how exactly that happens. Lessig advocates for employing smarter ways to regulate the internet’s architecture through approaches based on law, commerce, and cultural preferences, that will, in turn, enhance its value. To some extent, international commerce has already driven some of the innovations in determining the net’s architecture, but not to the point where it constitutes Lessig’s ideal. By contrast, Johnson, in Emergence, studies the internet and its web-based delivery of information from a bottom-up perspective in accordance with the theme of his book. Johnson believes that emergent macro behaviors from small entities determine a particular system’s adaptive qualities; he highlights ant colonies and cities as prime examples of emergent activities. Ants and urbanites might be quite simple on an individual basis, but together simple behaviors aggregate to create intelligent, learning systems, for better or worse. Despite the internet’s scale, however, Johnson argues that it, as historically and presently configured, does not constitute an adaptive system. The web, the primary software protocol employed on the internet, is an example of a directed network; the links only work in one direction, and the other pages on the web have no way of knowing what its neighbor is doing. The feedback loops present in complex adaptive systems with emergent behavior provide them with their adaptive capabilities. At best, argues Johnson, the net as presently structured is akin to “a snowflake”: a unique shape that can scale to massive size under the right conditions. At worst, in sticking with neurological comparisons to the “global brain” of de Chardin, Johnson argues that the internet has more in common with a brain tumor; uncontrolled growth with maladaptive qualities. While futurists like Kurzweil and others optimistically believe that the sheer expansion of information created through networked systems will result in an inevitable global intelligence, Johnson takes a more cautious approach and focuses on the devil in the details as to what programs might infuse the current network with more adaptive qualities. Johnson highlights programs that find relationships between web content and archival storage of the internet as an embryonic method to imbue the web with learning characteristics that he finds lacking at present.
This is where I believe that Lessig and Johnson begin to converge: in order to make the web and the internet a more productive too / place / thing, it requires structural improvement. Lessig argues from a more control based standpoint that includes big government, big commerce, and society at large hashing out differences and creating rules that determine network behavior and use. In this way, Lessig hopes to transform from the Wild West into some sort of digital heartland based off shared values. Johnson, however, thinks that structural improvement can be achieved through the use of smarter programming that learns from what people actually choose to do with the internet. In this way, we might be able to approach the ideal of some sort of “intelligence” in the internet, albeit not necessarily the classical self-aware archetype from science fiction. I am pretty sure that Johnson does not believe in Kurzweil’s Age of Spiritual Machines. His approach is much more bottom up, at least from my cursory examination. I find Johnson’s approach much more compelling, given my natural predilections towards the open-field internet and suspicion of the nexus of big business and big government. Johnson’s emphasis on search engine protocols that map the network links, relationships, and archival information as a way to help the internet demonstrate emergent, learning behavior strikes me as something much more democratic, even if some might consider it invidious in terms of mapping out millions of individual preferences and then logging it away. Given the alternative, I feel as though this looks like a much less heavy-handed approach that could be employed for nefarious purposes. Lessig admits as much that shaping and coding the internet architecture could be misused by governments not inclined to permitting civil liberties. Provided, of course, that we actually want our networks to display learning capabilities, which is an entirely different consideration.
