It may be easily argued that the biggest change taking place in our world is the shift from the bureaucracy to network organizations; from vertical forms of organization based upon physical limitations and boundaries to horizontal networks based upon computation and connectivity. Algorithmic coordination, connectivity, and automation are the key elements enabling us to turn the bureaucracies of today into the networked organizations of tomorrow.
We are so grounded in our existing forms of organizational structures that it is difficult to think outside of their box. To first clarify our thinking we might ask, what do we need to have an organization? We need things, technologies, resources, people, we need connections between those things and some way of processing information. We need physical technologies because almost all organizations today do what they do through technology. A manufacturing company produces things through a production line, a transport company moves things with vehicles, agriculture happens through farm machinery, etc.
The systems of organization we know today are largely a function of the physical constraints of the Industrial Age. They are centralized because of high transaction costs; everyone had to be grouped together and everything had to be brought into the center to be batched processed. The key characteristic of our existing organizations is a static bureaucratic structure of people. Because of limited information processing capacities, people do most of the more advanced information processing. They are held together by a chain of command that ensures that the decisions of the people at the center are executed upon to create a functioning organization that can coordinate people and resources, sense and respond to changes in its environment and then through the chain of command enact the necessary changes.
So what has changed?
What has changed is the combination of computerized coordination – algorithms – connectivity, but also automation. Let me illustrate how this works to fundamentally restructured and redefine our organizations over time. Greater connectivity means that it is no longer necessary to constrain everything in the center, given high enough levels of connectivity – low enough cost of exchange – it is viable to push resources out to the edges and connect people and things horizontally.
Of course, the main reason we didn’t do that in the past was to do with transaction cost but also due to the complexity of coordinating decentralized systems. We concentrated everything in the center with a few making decisions and then pushed those out to be executed by others – who took a rather passive role – because of our limited capacity for complex coordination.
The second major component that has changed is ever greater computing power capable of running ever more complex systems of coordination – capable of massively parallel processing of information in a way the bureaucracy is incapable. Meaning networked organizations can coordinate quite complex systems. A key characteristic of the bureaucracy is the fact that we have people in fixed positions who do the information processing – the day trader making decisions on which stocks to trade, the shop manager, the air traffic controller the market analyst, the insurance broker, etc – but as the algorithms become more sophisticated these kinds of activities and many others can be done algorithmically.
The other component is automation. If we think about a traditional organization, many of the people will be occupied with a job where they are controlling some form of technology that is doing the physical work; on an assembly line in a manufacturing company, driving a train in a transport company, or baking cookies in an oven. It is important that these people follow procedures and execute on the decisions from above – which is a big part of where we get the chain of command from. Automation does away with this. These emerging networked organizations will increasingly have this production process, technologies, and physical systems connected up directly to the information network and algorithms.
These networks can be open instead of closed because the control and rules are built into the network protocols instead of having to be imposed through a chain of command that requires borders to be maintained. While centralized systems hinge around efficiencies of scale and thus work to concentrate the most efficient components in the center, networks are about network effects, they push outwards being more effective the more people that connect into the network thus there is a drive towards openness. This kind of structural change, from closed to open, makes it possible to harness the kinds of resources and information required to “manage” large complex systems.
In a world of networks, everything is about connectivity and the exchange of value along those connections. We move from a world of centralized, fixed, long-term planning with prespecified outcomes to a much more dynamic world that is defined by the real-time exchange of value within digital markets. Markets come to act as a coordination structure capable of allocating resources in a decentralized fashion without anyone in control or specifying outcomes. This is part of what we give up as we move into a more complex world, the capacity to control outcomes through centralized systems, but we each get the capacity to influence the system with the value that we create and exchange.
So what do we get for all of this? Do we get richer or happier? Not necessarily, but we start to get the capacity to change the structure of these systems so that we are no longer fighting against the complexity but working with it. When everything moves into the world of information networks we get the capacity to design those networks. We start to get the capacity to think about how we might reorganize these networks from creating negative externalities to creating positive externalities and the emergence of new levels of organization.