A world run by software

A few days ago I reshared this talk from Balaji Srinivasan, along with my initial comments defending the position against what I took to be a superficial rejection from +David Brin and others. It was my first watching of the lecture, and my comments were borne of the passion that comes from having considered and argued for similar conclusions over the last few years, against those I felt were resisting the alternative framework BSS was suggesting without due consideration.

But there is always room for critical reflection, and now that I've had a few days to digest the talk I'd like to write a more considered response. I am utterly convinced that a world run by software can be more fair, inclusive, and sustainable than any mode of organization the industrial age had to offer. Nevertheless, BSS says precious little in the talk of what such a world would look like, or what reasons we have for believing the conclusion to be true. BSS's argument is largely critical about the problems and constraints of the existing system, with the goal of motivating interest in an alternative. I agree with much of his critique, especially his observation that people are already eagerly fleeing industrial age "paper" technologies in favor of digital alternatives. But the Silicon Valley audience to which the talk is directed might give the impression that a world run by software would benefit primarily those privileged few who are already benefiting from our nascent digital age, as yet another way to widen the gap between the wealthy and the rest. I think this is a misleading impression. A positive story that constructively described how a world run by software would operate would go a long way towards helping people imagine it as a real and plausible alternative, with distinct advantages over the existing order of things. I hear that BSS is planning to publish a more detailed treatment of his views, but while we wait I'd like to add my own thoughts to the discussion.

Because after all, there is lots to say and lots of discussion still to come from many different quarters. We are talking about about a fundamental change of the social order-- a revolution-- that will bring about many changes at many different scales. These changes might manifest in surprising and unpredictable ways, so feedback from everyone is important. The task of integrating diverse perspectives and feedback was not possible (and barely even conceivable) before the advent of digital technologies. We're just beginning to use these tools to organize digital populations at scales that rival nations, and already these digital populations overlap and engage with industrial age political structures in complex and twisted ways. But we are getting ahead of ourselves. There are discussions we must have about how to use the tools and organizations around us to bring about the changes we want, and then there are discussions about what changes we want and what the world will look like when our revolution is complete. There are important things to say about the former, but I want to use this space to speculate on the latter.

Here's what a world run by software looks like:

A world run by software is run on on open-source principles. It is participatory, and encourages experimentation in an open, collaborative development. There's already been a lot written about open source and peer networks. I don't want to just talk about the values of a digital society, I want to talk about the nuts and bolts of how it works. So I'm taking these principles as given. If any Silicon Valley types think a world run by software is a world of tightly controlled walled gardens, I urge you to consider another line of work immediately. (I'll say more about dealing with the libertopian "walled gardens" later in the essay.)

More abstractly, a world run by software is task-oriented. A program performs a function; a world run by software is designed to execute specific functions reliably and on demand. This involves three basic functional parts:
  1. Executors, who carry out some specified task,
  2. Programmers, who design and optimize the tasks to be carried out
  3. Selectors, who select which tasks get carried out
In democracies like the US, the government is typically charged with all three roles, separated somewhat in the three branches, with democratic feedback taking the form of voting for the representatives who ultimately select, program, and execute our political world. Democratic feedback also takes the form of economic and other sociopolitical activity which the representatives consider, but the feedback in both cases is coarse-grained and unreliable. In a world run by software, these roles can all be more widely distributed and participatory to allow individuals a more direct engagement with the the political order.

Consider first, and perhaps most easily, the role of selection. What activities deserve our efforts and attention, and what should should our attitudes towards these topics be? Representative bodies tend to give such questions a very narrow treatment that is susceptible to corruption and divergence from the general will. Digital communities, on the other hand, self-organize into affinity groups and coalitions that regularly broadcast their positions on any number of social issues and policies. This activity can be regularly and automatically harvested for feedback from the community about any topic you'd like, and can be used to represent and anticipate the dynamics of the general consensus of the people. We don't need to set aside time and rituals for voting and participating in a democracy, because every action you take is a vote in a world run by software. 

If our policy directions and collective planning is handled automatically by self-organized digital communities, the role of traditional governance looks radically different In a world run by software, governance is system administration. "Politicians" are no longer elected representatives empowered to make political judgments, they are administrators tasked with maintaining the tools for supporting and maintaining the people's political judgments. This takes the form, on the front end, of maintaining a public and powerful interface for engaging with the machinery of the political system, and on the back end maintaining the protocols and infrastructure that make it all work smoothly together. Politics becomes a form of engineering, of network management, of organizational facilitation. In other words, a digital government consists entirely of open source programmers who contribute to the code bank of governmental procedure.

The last functional role is played by the executors, who carry out the work requested by the people according to the standard procedures instructed by programmers. Thus far the entire system I've described has been horizontal and self-directed, and the executors are no different. They are free to engage with the general will as they see fit, and can use that information to decide what jobs need doing and where their time and skills are best spent. Executors are the "working class" of the traditional caste system, but in a world run by software every worker is guaranteed recognition for the work they do. This will take the form of gamification techniques where people earn "badges" that represent the many dimensions of their skills and mastery. This system of rewards not only motivates work and encourages a competitive atmosphere, but also serves to more efficiently connect requested work with the people in the best position to handle the job. It also frees the executor from any institutional obligations, so they might follow the vocation of their choosing.

A properly implemented system of this sort, where requests are initiated by users and handled by executors and mediated by political facilitators, is sufficient for managing all political and economic activity. Even under improved automation, there will always be human work to do, including physical human labor. But there are more than enough of us to do all that work many times over, including many who want the challenge and recognition of success at even the difficult or "dirty" jobs. This completely obviates the need for paid employment as a means of compensation and distribution of wealth. So in a world run by software we are each free to follow our idle desires. We can pursue our curiosities confident that our needs and requests (including for basic food, shelter, medicine and education) will be provided on request, and that the political infrastructure exists to support any grievances that might arise. There's an opportunity cost to mere idle fantasy in potential badges unearned,  but the possibility that our endeavors can bring genuine recognition and success will continue to stimulate innovation and progress.

In a digital world, influence replaces material wealth as the engine of power. The loud voices in strong communities may command disproportionate influence, with no proper authorities to check that power from above. Instead, their power is checked by the wax and wane of digital populations, as they freely settle into the communities of their choosing. This is a power peasants rarely had against their feudal lords, and requires the protections of the people's right to speech and movement. In other words, it means there can be no perfectly walled gardens; all systems must be porous, allowing information to flow in and out in standardized ways. This is the key to understanding the distinction between the libertopian castle doctrine nightmare many people legitimately worry about, and a collectively self-organized digital society. In a world run by software, privacy is a setting not a right. If you aren't providing feedback and aren't transparently documenting your activity and successes, you will not receive recognition from a digital society, and will have difficulty engaging with or accessing any of its benefits. In other words, you are opting out. People should be free to opt out, but at a certain point it means moving off the grid entirely. And given the freedoms of a digitally self-organized society, that's not something people are likely to do.

I've laid out, as briefly as I can manage, some very high-level organizational structures for managing a world run by software, and pointed towards the potential advantages and disadvantaged it might bring. Many people see such a world as a utopian potential in our distant technological future. I don't see anything in the proposal I've offered here that isn't already possible with existing technologies; in fact, many of the ideas suggested here are already in various stages of implementation and discussion in many active online communities already. I'd argue that the theoretical and philosophical dimensions of this proposal are also tractable within existing political and ethical models.

We do not lack for the tools and talent to implement ideas as ambitious and revolutionary as these. We do not lack the political will to seek out and advocate for real alternatives to our political situation. I feel these changes are inevitable and already well underway, whether I advocate for them or not. I hope that the vision of the future that BSS has provoked me to lay out here helps us resolve that future more clearly so we can better prepare for its arrival.