Excerpted from Paul Mason‘s latest book on Post-Capitalism (selection by Nathan Cravens):
“Given that we are decades into the info-tech era, it is startling that – as Oxford maths professor J. Doyne Farmer points out – there are no models that capture economic complexity in the way computers are used to simulate weather, population, epidemics or traffic flows.
In addition, capitalist planning and modelling are typically unaccountable: by the time a major infrastructure project starts delivering results, ten or twenty years after its impact was first predicted, there is no person or organization still around to draw conclusions. Thus, most economic modelling under market capitalism is actually close to speculation.
So one of the most radical – and necessary – measures we could take is to create a global institute or network for simulating the long-term transition beyond capitalism.
It would start by attempting to construct an accurate simulation of economies as they exist today. Its work would be Open Source: anybody could use it, anybody could suggest improvements and the outputs would be available to all.
It would most likely have to use a method called ‘agent-based modelling’ – that is, using computers to create millions of virtual workers, households and firms, and letting them interact spontaneously, within realistic boundaries. Even today such a model would be able to draw on realtime data. Weather sensors, city transport monitors, energy grids, postcode demographic data and the supply chain management tools of global supermarket groups are all giving off relevant macro-economic data in realtime. But the prize – once every object on earth is addressable, smart and feeding back information – is an economic model that does not just simulate reality but actually represents it. The agents modelled virtually are eventually substituted by granular data from reality, just as happens with weather computers.
Once we are able to capture economic reality in this manner, then planning major changes in an accountable way becomes possible. Just as aircraft engineers model millions of different stress loads on the tail-fin of a jet, it would be possible to model millions of variations of what happens if you reduce the price of Nike trainers to a point between their present $190 and their production price, which is likely to be lower than $20.
One specific problem is how to record the experience of failure into persistent data that allows us to retrace our steps, amend them and roll out the lessons across the whole economy. Networks are bad at memory; they are designed so that memory and activity sit in two different parts of the machine. Hierarchies were good at remembering – so working out how to retain and process lessons will be critical. The solution may be as simple as adding a recording and storing function to all activities, from the coffee shop to the state. Neoliberalism, with its love of creative destruction, was happy to dispense with the memory function – from Tony Blair’s ‘sofa’ decision-making to the tearing up of old corporate structures, nobody wanted to leave a paper trail.”