AI & Governance

Scoping Cybernomos is concerned with the intersection of artificial intelligence-AI and governance. With the need to create a foundational framework toward the creation of a substantive institution towards its governance. One capable of framing and addressing concerns relating to the alignment issue: normatively understood as the requirement for the machine intelligence of AI to be safely contained and congruent with human values. Of how with the advent of AI we have an opportunity to adapt the way we govern ourselves, to strive to evolve toward a certain moral pragmatism. To model a pathway from our human to our post-human future will require us to create suitable institutions, bodies that reflect the transition of nation states into de facto technological republics. Just as the ‘Framers’ who assiduously debated-drafted the United States Constitution in 1787, we must now turn to the creation of trans-national global governance organizations. Cybernomos is then a first pitch, a scoping exercise at formulating what such an institution should embody; of how we are now obliged to return to nurturing the fundamentals of the human condition in the light of emergent omnipresent computational intelligence.

An Agnostic Theology

As significant as the US constitution has proven to be it belongs to an evolving governance process, of which government is but a subset of the concept of governance itself. The last time humanity embraced such monumental change in the governance sphere it produced the seismic contributions of monotheism and democracy, over several millennia ago. It’s the former of these that functions as the point of departure for the theoretical propositions put forward here. Hence the importance of characterizing this discursive agenda as a form of agnostic theology. It was never my intention to adopt the monotheistic prism it just unfolded in that direction till we were on the same page. AI and monotheism, at first glance, a somewhat incongruous yoking together of disparate worlds granted, but nevertheless an intriguingly productive frame of reference.

What’s implicit from this unexpected discursive conjunction is that the complementary issues of AI governance, governance with AI, and governance within the context of AI, are not being thought about from a wide enough perspective. That not only might AI be dangerous, as in catastrophically disruptive, but the potential promise of such technology, if not developed and deployed wisely, might be lost along the way as well. A paradoxical irony indicating a situation where we become so obsessed with the notion of a loss of control, with narrow conceptions of containment, that we effectively ignore the intrinsic promise of AI’s potential to completely reset the culture of governance we currently have.

The Visionary Aspect of Governance

Governance is concerned with cultivating a steering capacity that facilitates, authorizes power, to control certain outcomes. Its fundamental normative concern has as much to do with fostering the beneficial as it does with containing the harmful. We’re familiar with the regulatory aspect of governance, the necessary concern regarding AI governance, of how to regulate the technology, how to safely contain it. This text observes this fundamental requirement, but it’s not its primary concern. Governance is also good at perspective, the visionary aspect of governance. Its virtue is its requisite ability to step outside of itself, to pose the awkward questions regarding what’s changed, how do we adapt? How do we align AI with human values, if we technically could, or should? Or even, given the wonderfully diverse human family that we are, can the collective ‘we’ come together to forge an agreement on such values? Perhaps; yet the responsibility of the governance perspective is perfectly happy inverting this dilemma. Instead of asking—How can we align AI with human values? The focus of this communication, an exploration of the visionary aspect of governance, embraces the prescient opportunity that we might want to repose this question to read—How can AI be utilized to help further establish what those values could be; can it be the impetus to expand the moral sphere; will it help us moralize ourselves? In this scoping exercise alignment is reset to encompass the opportunity that emergent AI affords to advance a certain moral pragmatism. For human maturation to evolve to such an extent, that in the accidental or otherwise formation of a formidable intellect to share our lives with, such an enigmatic presence might choose to align itself with us.

The Nomos of Cybernomos

The proposed Cybernomos Institute is a prescriptive attempt to model what such a structure and policy driven body should be aiming to achieve at such an important interregnum in the evolution of the human condition. The neologism Cybernomos is understood here as a form-and-content compound: a kind of portmanteau where cyber names the technological, cybernetic, and systems-based form through which governance will be increasingly administered, while nomos names the law, order, or governing principle by which that form must be directed. Where the nomos of cybernomos is supplied by the governance principle of altruistic self-interest–ASI. Retracing the journey of discovery regarding the virtue of this governance principle, unpacking what this formulation offers for grounding the agenda of the putative Cybernomos Institute is the principal purpose of this communication.