Large language models reduce agency costs


With Jason Potts, Darcy W E Allen, and Nataliya Ilyushina. Available on SSRN.

Large Language Models (LLMs) or generative AI have emerged as a new general-purpose technology in applied machine learning. These models are increasingly employed within firms to support a range of economic tasks. This paper investigates the economic value generated by the adoption and use of LLMs, which often occurs on an experimental basis, through two main channels. The first channel, already explored in the literature (e.g. Eloundou et al. 2023, Noy and Wang 2023), involves LLMs providing productive support akin to other capital investments or tools. The second, less examined channel concerns the reduction or elimination of agency costs in economic organisation due to the enhanced ability of economic actors to insource more tasks. This is particularly relevant for tasks that previously required contracting within or outside a firm. With LLMs enabling workers to perform tasks in which they had less specialisation, the costs associated with managing relationships and contracts decrease. This paper focuses on this second path of value creation through adoption of this innovative new general purpose technology. Furthermore, we examine the wider implications of the lower agency costs pathway on innovation, entrepreneurship and competition.

The problem of ubiquitous computing for regulatory costs

Working paper on SSRN

The benefits of regulation should exceed the cost of regulating. This paper investigates the impact of widespread general-purpose computing on the cost of enforcing of regulations on generative artificial intelligence (AI) and decentralized finance (DeFi). We present a simple model illustrating regulators’ preferences for minimising enforcement costs and discuss the implications of regulatory preferences for the number and size of regulated firms. Regulators would rather regulate a small number of large firms rather than a large number of small firms. General-purpose computing radically expands the number of potentially regulated entities. For Defi, the decentralized nature of blockchain technology, global scale of transactions, and decentralised hosting increase the number of potentially regulated entities by an order of magnitude. Likewise, locally deployed open-source generative AI models make regulating AI safety extremely difficult. This creates a regulatory dilemma that forces regulators to reassess the social harm of targeted economic activity. The paper draws a historical comparison with the attempts to reduce copyright infringement through file sharing in the early 2000s in order to present strategic options for regulators in addressing the challenges of AI safety and DeFi compliance.

The Case for Generative AI in Scholarly Practice

Available at SSRN

Abstract: This paper defends the use of generative artificial intelligence (AI) in scholarship and argues for its legitimacy as a valuable tool for contemporary research practice. It uses a emergent property rights model of writing to shed light on the evolution of scholarly norms and practices in academic practice. The paper argues that generative AI extends the capital-intensive nature of modern academic writing. The paper discussing three potential uses for AI models in research practice: AI as a mentor, AI as an analytic tool, and AI as a writing tool. The paper considers how the use of generative AI interacts with two critical norms in scholarship: norms around authorship attribution and credits for contributions, and the norm against plagiarism. It concludes that the effective use of generative AI is a legitimate research practice for scholars seeking to experiment with new technologies that might enhance their productivity.

Voting with time commitment for decentralized governance: Bond voting as a Sybil-resistant mechanism

With Vijay Mohan and Peyman Khezr. Available at SSRN

Abstract: Blockchain applications are increasingly experimenting with novel governance mechanisms that address issues that are important for their community: resistance to voter fraud in the form a Sybil attack; resistance to the formation of a plutocracy within the community; and, the ability to express preference intensity. In this paper, we take a closer look at these issues confronting decentralized governance. Our contribution is three-fold: first, we lay some analytical foundations for the formal modelling of the necessary and sufficient conditions for a voting system to be resistant to a Sybil attack; second, we show that a voting mechanism with a single instrument for expressing preference intensity, such as the quantity of tokens, cannot simultaneously achieve resistance to both Sybil attacks and plutocracy formation; and third, we design a voting mechanism, bond voting, that is Sybil resistant and offers a second instrument of voting influence (time commitment) for plutocracy resistance.

Buyback and Burn Mechanisms: Price Manipulation or Value Signalling?

With Darcy WE Allen and Sinclair Davidson. Available at SSRN

Abstract: A core finding in traditional corporate finance is that manipulating funding instruments does not increase the value of a firm. Several Web3 projects have mechanisms to buy their tokens on the market and burn those tokens. If the finding from corporate finance holds in the Web3 environment then this manipulation of the value of tokens should not increase the value of those projects. This paper asks if these mechanisms serve more of a purpose than price manipulation. We provide an efficiency explanation for buyback and burn mechanisms: value signalling. A buyback and burn enables projects to signal that their business model has genuine network effects, and that it is not a Ponzi scheme. This finding has implications for the motivation, justification and design of buyback and burn mechanisms across Web3.

The exchange theory of web3 governance (or ‘blockchains without romance’)

Working paper with Jason Potts, Darcy W E Allen, Aaron M. Lane and Trent MacDonald. Available on SSRN

Abstract: Blockchains have enabled innovation in distributed economic institutions, such as money (e.g. cryptocurrencies) and markets (e.g. DEXs), but also innovations in distributed governance, such as DAOs, and new forms of collective choice. Yet we still lack a general theory of blockchain governance. James Buchanan once described public choice theory as ‘politics without romance’ and argued instead for an exchange theory of politics. Following Buchanan, we argue here for an exchange view of blockchain governance. The ‘romantic’ view of blockchain governance is collective choice and consensus through community voting. The exchange view, instead, is focused on entrepreneurial discovery of opportunities for value creation in governance space through innovation in protocols (e.g. Curve, Convex, Lido, Metagov, etc) that facilitate exchange of coordination and voting rights, that are newly made possible through tools that enable pseudonymous, composable and permissionless governance actions. The exchange lens on web3 governance also helps illuminate how this emergent polycentric process can generate robustness in decentralised systems.

Interoperability as a critical design choice for central bank digital currencies


Working paper available at SSRN

Abstract: Interoperability is a key economic and technical consideration for payment systems. This paper explores the implications of interoperability for central bank digital currencies (CBDCs). CBDCs are digital representations of central bank money. A critical question is how those digital representations can interoperate with other CBDCs, private blockchains, and permissioned blockchains. By comparing prevailing CBDC interoperability models with interoperability in blockchain ecosystems, the paper finds that CBDC architectural choices are deeply intertwined with policy choices in a way not yet understood by the scholarly and policy literature. Widely discussed CBDC policy questions (such as whether a CBDC should be retail or wholesale, whether interest should be paid on CBDC holdings, and how privacy should be protected) are better understood as choices around interoperability. The paper concludes by connecting the CBDC policy debate to a parallel debate about fiat-backed stablecoin architecture and governance.

Trust and Governance in Collective Blockchain Treasuries

With Darcy WE Allen and Aaron M Lane. Available at SSRN

Abstract: Blockchain treasuries are pools of digital assets earmarked for funding goods and services within a blockchain ecosystem that have some public purpose, such as protocol upgrades. Ecosystem participants face a trust problem in ensuring that the treasury is robust to opportunism, such as theft or misappropriation. Treasury governance tools, such as expert committees or stakeholder voting, can bolster trust in treasury functions. In this paper we use new comparative economics to examine how treasury governance mechanisms minimise different types of costs, thereby bolstering trust. We interpret case studies of innovative treasury governance within this framework, finding that the costs shift throughout the lifecycle of an ecosystem, and those subjective costs are revealed through crisis. These changes lead ecosystem participants to choose and innovate on treasury governance.

Setting the reserve price for the Tracer DAO Gnosis auction

With Peyman Khezr

Introduction: Selling multiple units of a homogeneous good in an auction is one way of determining the market price. Uniform-price auctions have been used in many real-world markets because of their price discovery property: All winning bidders pay the same price (either highest losing bid or lowest winning bid). The question is how a seller could compute an optimal reserve price in a uniform price auction. First we should note that literature suggests a positive reserve price is usually better than no reserve price as it reduces the chance of underbidding by bidders. However, to compute the reserve price for a uniform price auction there are no clear criteria. In this note we follow the criteria given for the second-price auction as the best approximate of the uniform-price auction.

PDF available here

An economic theory of blockchain foundations

With Jason Potts, Darcy WE Allen, Sinclair Davidson and Trent MacDonald

Abstract: Blockchain (or crypto) foundations are nonprofit organizations that supply public goods to a crypto-economy. The standard theory of crypto foundations is that they are like governments with respect to a national or regional economy, i.e. raising a public treasury and allocating resources to blockchain specific capital works, education, R&D, etc., to benefit the community and develop the ecosystem. We propose an alternative theory of what foundations do, namely that the treasury they manage is a moat to raise the cost of exit or forking because the benefit of the fund is only available to those who stay with the chain. Furthermore, building and maintaining a large treasury is a costly signal that only a high quality chain could afford to do (Spence 1973). We review these two models of the economic function of a blockchain foundation – (1) as a private government supplying local public goods, and (2) as a moat to raise the opportunity costs of exit. We outline the empirical predictions each theory makes, and examine the implications for optimal foundation design. We conclude that foundations should be funded by a pre-mine of tokens, and work best when large, visible, transparent, rigorously managed, and with a low burn rate.

Available at SSRN.