ABSTRACT. The authors demonstrate how a population of agents considered by and large as a possible market is always able to find a series of local optimum points through the coordination of decisions and choices designed using swarm intelligence. In this regard, agents’ solution space depends on whether it comes from a well-structured or ill-structured knowledge concerning the operating mechanisms of the environment in which they operate. In the case of well-structured knowledge, general domain-independence abstract methods emerge. While, in the more common ill-structured knowledge, agents need domain-specific methods. In both cases, there remains an underlying illusion of reaching optimum points both because problem-solving activities are domain- and context-dependent, and because there may be different levels of knowledge on which to develop the strategies of optimum seeking. In any case, the solution space is strongly influenced by decisions and choices formulated on the basis of a priori beliefs and knowledge. Methodologically, therefore, the authors develop both a space of ‘zero-knowledge’ solutions based on random simulations and an alternative space of satisficing solutions built by the N random seeds of the agents simulated by a set of N neural networks. The authors represent the dynamics of evolutionary trajectories towards local optimum points and comparative results. Ultimately, the aim of the paper is to demonstrate that an ecologically rational population tends to find gradually better local optimum points even if in contexts of extended variability at the root of the basic rationale of N random seeds. For the above, zero-intelligence may become a reductio ad absurdum of the fact that information and knowledge underlying ecological understanding of behavior are still necessary especially to improve how decisions are made in determining the price of assets and in the face of decision-making pattern that applies to uncertain outcomes. Not surprisingly, indeed, the authors demonstrate how such a population operates in conditions of sub-rationality on a regular basis.

ABSTRACT. Emergent behavior in repeated collective decisions of minimally intelligent agents -- who at each step in time invoke majority rule to choose between a status quo and a random challenge -- can manifest through the long-term stationary probability distributions of a Markov Chain. We use this known technique to compare two kinds of voting agendas: a zero-intelligence agenda that chooses the challenger uniformly at random, and a minimally-intelligent agenda that chooses the challenger from the union of the status quo and the set of winning challengers. We use Google Co-Lab's GPU accelerated computing environment, with code we have hosted on Github, to compute stationary distributions for some simple examples from spatial-voting and budget-allocation scenarios. We find that the voting model using the zero-intelligence agenda converges more slowly, but in some cases to better outcomes.

ABSTRACT. We follow up on our talk last year with some novel experiments on trade in the CDA when each trade causes a negative/positive externality effect on a third-party bystander. Our methodology always varies the information and feedback content, thus offering a behavioral perspective on how simpler rules and more complex rules might differ in terms of resulting market outcomes and individual behaviors.

We compare the effects on individual decision-making between markets with externalities and without externalities, in both high- and low-information environments. In particular, we ask how much trade reduction occurs the more negative the externality of a trade, and how does individual behavior respond to information and feedback.

We also compare the resulting convergence patterns in terms of price and efficiency.

In short, we find significant effects at the individual level that suggest some externality sensitivity (in line with pso-social motives), but no significant traces at market outcome level. The CDA appears so robust that it implements CE reliably through time independent of information, feedback and social consequences.

We compare these findings with related ones from the literature, where there has been mixed evidence regarding whether markets are effective or not at supporting/killing moral behaviors.

The connection with a ZI/MI approach is important, as this kind of approach has traditionally not worried too much about trying to integrate social preferences, etc. Our results indicate that this simplification does not come at a huge cost in terms of descriptive accuracy, at least over time, in the context of CDAs and related trading institutions where these theories have been proposed most prominently.

ABSTRACT. Double auctions with human traders and “zero-intelligence” programmed traders converge to Pareto optimal allocations in partial equilibrium settings. We show that these results remain robust in two-good general equilibrium. We explain how market structure, not optimizing behavior, guides efficient resource allocation.

ABSTRACT. We study behavior in a Continuous Double Auction. We use a large data set from in-class experiments to test various models. We find that neither the Individual Evolutionary model (IEL) of Arifovic and Ledyard nor the Zero Intelligence (ZI) model of Gode and Sunder explain the data. IEL explains the efficiency data better than ZI and ZI explains the price data better than IEL. We introduce a no-intelligence (NI) agent who generates bids that violate individual rationality. We show that a subject pool of 70% IEL and 30% NI does well at explaining the human data. That mix explains the data better than either IEL alone or ZI alone.

ABSTRACT. This paper investigates how the structure of an inter-dealer market may be dynamically determined by the trade-off between the benefits of inventory risk-sharing and the costs of maintaining relationships. We employ a Zero-Intelligence (ZI) approach, in which agents behave without strategic intent, to understand the key interactions in this system. Each ZI agent randomly determines a quantity and price for the traded assets constrained solely by their capacities of holding inventories. Trades occur between ZI agents when their price-quantity pairs intersect with others such that neither agent makes a loss. The results indicate that the trade-off of dealers determines the shape of the inter-dealer network, and we observe the core-periphery feature of inter-dealer network: dealers with large capacities comprise the core of the network, while dealers with small capacities are at the periphery.

ABSTRACT. We build a microeconomic systems simulation of agents may move from a decentralized market for a homogeneous good to a centralized market using a limit book. In our markets we assume that a major source of transaction cost are the computational costs of making and completing a transaction, and the computational costs of maintaining a network of trading partners. We explore the conditions which lead agents to minimize transaction costs by converging to focal point meeting locations and adapt their bargaining strategy to both look for, and compete for, the best offer at their location. We then explore the conditions under which a centralized limit book market will emerge, and we show how agents might co-adapt a budget constrained zero intelligence strategy to minimize computational costs. We finally explore conditions which lead back to decentralized markets using blockchain technologies, and we study the adaptive fitness of zero intelligence strategies in markets using one of the emerging types of blockchain markets using `smart` contracts.

ABSTRACT. Human society is an open system that evolves by coupling various known and unknown fluxes. How does this complex system dynamics precisely unfold? A science of human society may provide further insights. The evolution of our civilization has depended largely on cooperation among human beings. However, our science is yet to figure out ‘How did cooperative behavior evolve?’ This question [1] from the 'What don’t we know’ series from the Science journal has been awaiting physical reasoning for a long time. We generalize the classical field theories for an ensemble of human beings and proposed the social field theory [2]. The social field theory formalizes the social force and the Hamiltonian of an individual in the social field. We underpin this new Hamiltonian as a physical basis for cooperation among human beings and the evolution of human society. With the Hamiltonian defined [3], we use the Navier-Stokes’approach to study the dynamics in the social field that evolves with time. The equations for the evolution of an individual and human society are sketched based on the time-dependent Hamiltonian that includes the power dynamics. Lotka-Voltera type equations can be derived from the Hamiltonian equation in the social field. A fundamental understanding of why cooperation evolved may have a resounding effect on our understanding of social, political, and economic rationale. Indeed, Darwin uncovered some ideas of cooperation in his theory of evolution. However, the science of evolution doesn't provide physical reasoning, and also not adequate for 21st-century human knowledge. Biologists are refining Darwin's ideas a bit by bit. Here, we aspire to uncover the science of cooperation and many social dynamics in terms of the Hamilton of an individual in the social field. Human cooperation is not much different literally from cooperation that takes place between an electron and nucleons in a model of the Hydrogen atom. It is our consciousness that makes human cooperation special among social beings. We speculate on some possible directions in which the science of human society may develop over the next few decades, especially by connecting the natural and social sciences -- the two eyes of our human knowledge. Money is a concept of paramount significance to economic science, social science in general. This new framework conceives of money following the original insights of Howard Odum [4]: money flows in circles, but energy flows through a system and ultimately comes out in a degraded form. These concepts may bring forth a useful connection between the sciences at the fundamental level. Even if the field-based approach based on homology may supply important insights to the core economic concept such as capital, development, business cycles, etc., there are still many important questions at the intersection of the two cultures [5] to which we currently have no satisfactory answers. We believe that a two-eyed approach may inform some important questions at the crossroads of the natural and social sciences in the 21st century.

REFERENCES [1] E. Pennisi, “How Did Cooperative Behavior Evolve?,” Science, vol. 309, no. 5731, p. 93, 2005. [2] R. C. Poudel and J. G. McGowan, “The Dynamics of Human Society Evolution,” Newsletter, July 2019, Forum on Physics and Society, APS. [3] A. J. Leggett, “Reflections on the past, present and future of condensed matter physics,” Science Bulletin, vol. 63, p. 1019–1022, 2018. [4] H. T. Odum and E. C. Odum, Energy Basis for Man and Nature, McGraw-Hill Book Company, 1976. [5] C. P. Snow, The Two Cultures and the Scientific Revolution, New York: Cambridge University Press, 1961.