Printing was a necessary but not sufficient input to the scientific revolution. The Invisible College, the group of natural philosophers who drove the original revolution in chemistry in the mid-1600s, were strongly critical of the alchemists, their intellectual forebears, who for centuries had made only fitful progress. By contrast, the Invisible College put chemistry on a sound scientific footing in a matter of a couple of decades, one of the most important intellectual transitions in the history of science. In the 1600s, though, a chemist and an alchemist used the same tools and had access to the same background. What did the Invisible College have that the alchemists didn't?From Danny Hillis.
They had a culture of sharing. The problem with the alchemists had wasn't that they failed to turn lead into gold; the problem was that they failed uninformatively. Alchemists were obscurantists, recording their work by hand and rarely showing it to anyone but disciples. In contrast, members of the Invisible College shared their work, describing and disputing their methods and conclusions so that they all might benefit from both successes and failures, and build on each other's work.
The chemists were, to use Richard Foreman's phrase, 'pancake people'. They abandoned the spiritual depths of alchemy for a continual and continually incomplete grappling with what was real, a task so daunting that no one person could take it on alone. Though as schoolchildren, the history of science we learn is often marked by the trope of the lone genius, science has always been a networked operation.
In this we can see a precursor to what's possible for us today. Just as the Invisible College didn't just use the printing press as raw capability, but created a culture that used the press to support the transparency and argumentation science relies on, we have the same opportunity.
Consider as a simple example, a program that needs to know the time of day. In the unconnected world, computers often asked the operator to type in the time when they were powered on. They then kept track of passing time by counting ticks of an internal clock. Programmers often had to write their own program to do this, but in any case, they understood exactly how it worked. Once computers became connected through the Internet, it made more sense for computers to find out the time by asking one another, so something called Network Time Protocol was invented. Most programmers are aware that it exists but few understand it in detail. Instead, they call a library routine, which asks the operating system, which automatically invokes the Network Time Protocol when it is required.Both answers obviously focus on the networking potential of the Internet. But both make it clear that networking itself won't create the magic. It's what we do with the networking that matters.
It would take a long time to explain Network Time Protocol, how it corrects for variable network delays and how it takes advantage of a partially-layered hierarchy of network-connected clocks to find the time. Suffice it to say that it is complicated. Besides, I would be describing version 3 of the protocol, and your operating system is probably already using version 4. It really does not make sense for you, even if you are a programmer, to bother to understand how it works.
Now consider a program that is directing delivery trucks to restock stores. It needs to know not just the time of day, but also the locations of the trucks in the fleet, the maps of the streets, the coordinates of its warehouses, the current traffic patterns, and the inventories of its stores. Fortunately it can keep track of all of this changing information by connecting to other computers through the Internet. It can also offer services to other systems that need to track the location of packages, pay drivers, and schedule maintenance of the trucks. All of these systems will depend upon one another to provide information, without depending on exactly how the information is computed. All of these communicating systems are being constantly improved and extended, evolving in time.
Now multiply this picture by a million fold, to include not just the one fleet of trucks, but all the airplanes, gas pipelines, hospitals, factories, oil refineries, mines and power plants not to mention the salesmen, advertisers, media distributors, insurance companies, regulators, financiers and stock traders. You will begin to perceive the entangled system that makes so many of our day-to-day decisions. Although we created it, we did not exactly design it. It evolved. Our relationship to it is similar to our relationship to our biological ecosystem. We are co-dependent, and not entirely in control.
Shirky points out that people have to learn to use our new tool to collaborate—which we are already quite advanced at doing.
Hillis talks about something that's more interesting to me. I've become increasingly interested in thinking about complex systems as a sea of interacting energy flows. Here are some slides I used to introduce a panel on Energy and Information in Complex Systems at one of the AAAI symposia last November. The fundamental point is that we live in a world of energy flows. Most living things survive only because they are able to exploit some of the energy flows around them. Hillis points out that we now have the ability to track many of those energy flows. With that ability we are likely to have a much greater ability to make use of the energy potentials in the world—leading to enormously greater and (at least for a while) continually increasing productivity and gains in the quality of most people's lives.