The future influences the present just as much as the past – F. Nietzsche
The current system of global economic governance is increasingly encountering limitations in dealing with the challenges of trade protectionisms, environmental challenges, pandemics and global economic downturns. This is a system that evolved largely as a result of the power dynamics on the international stage rather than a pre-meditated and truly comprehensive effort of the global community aimed at building an inclusive, efficient and adaptive global economic system. The current global economic architecture may be broadly categorized as a two-layered system consisting of a layer of national economies and the layer of global international organizations such as the WTO, IMF and the World Bank. There remain glaring gaps within this global construct, most notably with respect to regional organizations and regional integration arrangements that still do not have platforms of cooperation on the international stage.
As a result, the global economic system does not employ all of its available resources – it is not able to bring together all parts of the Global Financial Safety Net (GFSN), most notably the parts that pertain to the regional layer of financial institutions (regional financing arrangements (RFAs) and regional development banks (RDBs)) into a coordinated and duly comprehensive anti-crisis package during periods of economic downturns. Indeed, the limitations of the current system in bringing about a coordinated response came to the fore during periods of economic crises in 2008-2009 and 2020, when G20 resorted to coordinated anti-crisis stimuli. A related problem is that the global financial system is segmented, with silos and pockets of economies that are isolated with limited connectivity to the rest of the global economic system. Most importantly, the system is not adaptive and has limited capacity to learn from its past failures – as has been amply demonstrated by the crises episodes of the past several decades.
In overcoming some of the above deficiencies, one possibility would be to structure global governance as a neural network (GGNN), whereby the global system would be predicated on several layers – the input layer being the layer of national economies, with each national economy serving as a node; a regional hidden layer composed of the interconnected nodes of regional integration blocs such as ASEAN, Mercosur and others; there may also be additional regional layers whose nodes are composed of regional development institutions such as regional development banks; an output layer of global international organizations such as the WTO, IMF and the World Bank. Such a setting would be based on all parts of the system being inter-connected, with all three main layers serving as platforms at the national, regional and global levels respectively. It would also enable the global system to operate as a “learning organization” by refining each successive iteration and policy round via feedback loops and a reevaluation of the weights of the components of the global economic system.
The creation of the GGNN would enable the global economy to react more efficiently and in a more inclusive way to periods of downturns during coordinated stimuli. It may also enable the system to multilateralize the trade liberalization at the various levels of the system (whether national, regional or the level of the WTO trade rounds) with feedback loops allowing for improving the precision of the scale of such liberalization with each round. It may also be instrumental in testing the global financial system for potential shocks and ex-ante preventing disruptions in global financial markets. Perhaps most importantly, the GGNN could serve as a model/reference point/benchmark in terms of how the global economic system could potentially evolve towards greater efficiency and inclusivity – currently there does not appear to be a credible model that addresses the gaps and the shortcomings of the current set-up.
The operation of the GGNN would be very demanding, however, in terms of the commonality of standards and the universal coverage of the different parts of the global costruct. It would need a common digital infrastructure that would allow for the processing and sharing of real time data; it would also need a high degree of data standardization and the availability of APIs at the national, regional and global level to ensure connectivity with the entire platform. Apart from the technological and data requirements another indispensable element is a critical degree of international cooperation – such a global governance construct would need to be supported by all participating national authorities, regional blocs and global organizations.
While the truly global infatuation with all AI-like and neural networks as models of “learning organizations” may be traced to the past several years, in reality organizational theory (OT) has borrowed extensively from AI development over the course of the past several decades as argued in a survey undertaken by Tom Steinberger and Felipe Csaszar[1]. As stated in their paper, “a rarely acknowledged fact about organization theory (OT) is that many of its ideas stem from the field of artificial intelligence (AI). For example, key OT concepts such as problemistic search, heuristics, exploration, requisite variety, and organizational scripts all have their roots in AI”[2].
In the sphere of business and management theory, neural networks are increasingly discussed as prototypes of learning organizations that use feedback loops to improve performance: “teaching a NN involves a process of iteration and optimization. The system learns by adjusting the weightings of the network connections based on the difference between the actual and desired outputs. This is akin to how organizations shape their culture… Intertwining NNs and organizational culture provides a unique lens for understanding and influencing the behaviors within a company. By embracing these behaviors, executives can help shape their culture”[3].
In the end, a neural network for global economic governance may address some of the core problems experienced by the world economy today. The critical element within such a paradigm is that of continuous learning and improvements in the global economic system. There may be plenty of questions and doubts about such a framework – the current realities of top-down, still unipolar global governance suggest that the GGNN system may be a rather distant, if at all realistic goal. But while the GGNN framework may strike many as wildly futuristic, the pace of technical progress and the scale of perturbations on the international arena greatly widen the range of possibilities. With AI growing exponentially in importance for the global economy, it may now be the AI corporate leaders of the world, rather than economists such as Keynes and White who will redefine the contours of “new global governance”. At the same time, the more the global economic system progresses towards a multipolar setting, with a more equal distribution of weights across countries, regions and within global organizations, the more feasible could be the creation of such a system. Rather than the global governance system being a function of the tug-of-war among the few heavyweights, an inclusive system that pursues economic efficiency, learning and the capability to continuously adjust to evolving challenges should become the guiding paradigm in terms of how we think about global economic governance.
[1] FELIPE A. CSASZAR University of Michigan, TOM STEINBERGER Korea Advanced Institute of Science & Technology (KAIST). ORGANIZATIONS AS ARTIFICIAL INTELLIGENCES: THE USE OF ARTIFICIAL INTELLIGENCE ANALOGIES IN ORGANIZATION THEORY. Academy of Management Annals 2022, Vol. 16, No. 1, 1–37. https://doi.org/10.5465/annals.2020.0192. Retrieved from: https://d30i16bbj53pdg.cloudfront.net/wp-content/uploads/2024/08/Organizations-as-artificial-intelligences-The-use-of-artificial-intelligence-analogies-in-organization-theory.pdf
[2] Op. cit.
[3] https://evolution.team/blog/unraveling-organizational-culture-through-the-lens-of-neural-networks
Yaroslav Lissovolik, Founder, BRICS+ Analytics
Image by ChatGPT