Skip to main content

Leveraging Large Language Models for Enhanced Blockchain Social Governance

Student Final Reports from Prof. Luyao Zhang’s project entitled “From ‘Code is Law’ to ‘Code and Law’: A Comparative Study on Blockchain Governance,” Summer Research Scholar Program, Duke Kunshan University

Published onAug 19, 2023
Leveraging Large Language Models for Enhanced Blockchain Social Governance
·

Disclaimer: This article is a final deliverable from Prof. Luyao Zhang’s project entitled “From ‘Code is Law’ to ‘Code and Law’: A Comparative Study on Blockchain Governance,” supported by the 2023 Summer Research Scholarship (SRS) program at Duke Kunshan University. Our Summer 2023 scholars also receive additional support from Wang-Cai Biochemistry Lab Donation Fund.

Part I. Research Summary

1.1 Research Title

Leveraging Large Language Models for Enhanced Blockchain Social Governance

1.2 Abstract

Blockchain technology has revolutionized various industries, and effective governance is crucial for ensuring the success and sustainability of blockchain projects. Community forums, as essential platforms for community engagement, play a significant role in shaping blockchain governance decisions. Concurrently, large language models (LLMs) have emerged as powerful tools for revolutionizing data analysis in various industries. This study investigates the feasibility and methodologies for integrating LLMs into blockchain social governance.

The Research article highlights LLMs' transformative capacity in augmenting decision-making within blockchain networks. By excelling in text generation, summarization, and sentiment analysis, LLMs foster transparency and inclusivity in decentralized communities. Analyzing textual discussions and justifying decisions, LLMs enhance diversity and fairness in the decision-making process. The Innovate article outlines practical methodologies to implement LLMs in blockchain governance. Utilizing data from the Uniswap community's Discord discussions, we introduce a comprehensive dataset for LLM training and evaluation. Employing tools like h2o.ai, we detail a systematic process to fine-tune LLMs for analyzing blockchain discourse. This approach empowers LLMs to provide insights and answers pertinent to ongoing debates, enabling informed governance choices.

The study reveals the potential of LLMs to guide decision-makers by extracting data-driven insights from community discussions. It advances our comprehension of the impact of LLMs on blockchain governance, emphasizing implications for decision-making processes.

1.3 Relevant Deliverables

1.3.1. Exploratory Research Article 1

1.3.2. Exploratory Research Article 2

Part II. Scholar Reflections

2.1 Intellectual Growth

My intellectual growth from this SRS project is substantial, particularly in comprehending AI and large language models (LLMs). Previously, I limited my understanding of LLMs to basic AI chatbot interactions, like ChatGPT, but this project has expanded my grasp significantly. I now appreciate the diverse applications of LLMs and their technical nuances in fine-tuning and generalization. Moreover, I've gained a comprehensive insight into blockchain governance, identifying areas for reform essential for my future endeavors in blockchain research.

As I delved into the research, I realized how LLMs can address pivotal economic and social issues within the blockchain ecosystem. This experience broadened my perspective on LLMs' potential to reshape decision-making, transparency, and inclusivity in blockchain communities. My efforts to comprehend complex concepts like fine-tuning LLMs and using tools like h2o.ai yielded valuable insights. This journey not only enhanced my technical knowledge but also refined my analytical skills, making me better equipped for complex problem-solving.

In short, this research transformed my view of LLMs, from basic understanding to recognizing their potential. It deepened my grasp of AI models, including technical aspects, and expanded my insights into blockchain governance reform. By merging different cutting-edge technologies, I'm better prepared to navigate complex challenges and contribute effectively.

2.2 Professional Development

In terms of tangible deliverables, I delved into the feasibility and practical methods of integrating large language models (LLMs) into blockchain social governance. This process involved designing and implementing fine-tuning methods using tools like h2o.ai. It demanded innovation and technical acumen, particularly due to the rapidly evolving and transformative nature of both AI-based LLMs and blockchain. References and previous research in these novel fields were limited, posing unprecedented challenges to my resource integration and innovation capabilities. Nonetheless, I embraced these challenges as opportunities for innovation and creative problem-solving.

Working within an interdisciplinary team enriched the research process by bringing diverse viewpoints to the forefront. I learned to effectively communicate and collaborate with peers from varying backgrounds, enabling holistic problem-solving approaches and exploration of different research avenues related to blockchain.

Throughout this journey, I encountered numerous technical and conceptual hurdles. The cutting-edge technologies like LLMs and blockchain, both evolving rapidly, posed unique difficulties. Additionally, conveying the intent of the research to a broader audience was a challenge. Nonetheless, I transformed these adversities into stepping stones toward better results, harnessing the power of innovation to find creative solutions.

In summary, my SRS journey has yielded substantial achievements. By navigating challenges, innovating in uncharted territories, and communicating research with empathy, I have evolved both professionally and personally. This growth equips me to drive forward pioneering research, leverage innovative technologies, collaborate seamlessly, turn adversities into opportunities, and communicate effectively in pursuit of impactful solutions to critical issues.

2.3 Leadership Practice

In the SRS project, I've come to realize the profound impact of independent leadership on research, which goes beyond merely putting research into practice, but also involves shaping the research direction. Throughout this journey, I've practiced leadership skills that guide research direction, identify gaps in existing knowledge, and design research questions and processes. This type of leadership practice isn't just about guiding a project; it's about shaping the path of exploration in research. I've come to understand that leadership not only drives a deeper understanding of subjects but also inspires impactful contributions.

In particular, the fusion of cutting-edge technologies like Large Language Models (LLMs) with the realm of blockchain underscores the necessity of innovative leadership. Leading the integration of LLMs into blockchain governance requires conceiving entirely new solutions and daring to implement them. This involves embracing risks, navigating uncertainties, and challenging conventional thinking. Through my own experience, I've gradually grasped the significance of these aspects.

In conclusion, my SRS experience has revealed that successful research isn't confined to the technical realm alone; it necessitates leadership to drive innovation, foster collaboration, and navigate through uncharted territories. My journey resonates with this insight that true leadership spans from conceiving research to execution, transforming challenges into opportunities for growth and influence.

Part III. Future Plans

3.1 Joint academic research papers

  • Paper 1: I am completing a research paper entitled “Empower Blockchain Governance Decisions: Exploring the Application of Large Language Models in Community Forums,” with Prof. Luyao Zhang as the corresponding author and Xintong Wu as student co-authors, built upon the SRS project. We plan to submit the paper for international conference presentations and conference proceedings or/and journal article presentations in the upcoming year.

  • Paper 2: I am also co-authoring a paper that delves into the intriguing realm of cryptocurrency exchanges. While Xintong Wu will be leading as the first author, our joint efforts aim to analyze the future direction of the cryptocurrency exchanges, thus proposing a centralized and decentralized possible collaborative systems between finance. By leveraging our collective expertise, we are set to help users better understand the cryptocurrency market and support informed decision-making by examining trends between CEX and DEX platforms and their distribution. Prof. Luyao Zhang will be the corresponding author.

3.2 DKU Signature Work

My signature work primarily delves into the realm of blockchain research, and the SRS project has provided invaluable experience that directly aligns with my signature work's focus and methodologies. The SRS project has not only enhanced my understanding of the blockchain field but has also equipped me with valuable insights into research methodologies that can be seamlessly integrated into my signature work.

From the exploration of blockchain governance dynamics to the practical implementation of innovative technologies like Large Language Models (LLMs), the SRS project has exposed me to a wide spectrum of interdisciplinary approaches. This exposure has allowed me to refine my research focus and gain a deeper appreciation for the multifaceted nature of blockchain-related challenges. The methodologies employed in the SRS project, such as fine-tuning LLMs and utilizing advanced tools like h2o.ai, can be directly transferable to my signature work. The technical skills and innovative thinking cultivated during the SRS project will enable me to approach my research with a more comprehensive and informed perspective.

The SRS project has been a key in my journey towards creating a meaningful and impactful signature work in the blockchain field. The experiences gained in terms of research focus, methodologies, and innovation will undoubtedly shape the quality of my signature work, contributing to the advancement of knowledge and solutions within the blockchain field.

3.3 Industry Internship or Job

The SRS project has greatly influenced my aspirations to pursue a career in the finance industry while contributing to the convergence of finance and technology.

The project's exploration of integrating Large Language Models (LLMs) into blockchain governance has illuminated the potential of advanced technologies in transforming traditional financial practices. This experience has shown me how technology can be harnessed to enhance transparency, security, and efficiency within financial systems. When I enter the finance industry in the future, I will try to leverage such insights to innovate existing financial processes. Additionally, the project's engagement with emerging technologies aligns well with the rapid technological advancements shaping the financial field. The insights gained from integrating AI into blockchain governance highlight the importance of staying updated with cutting-edge developments.

In summary, the SRS project aligns with my ambitions to contribute to the intersection of finance and technology. As I embark on my journey in the finance industry, I am eager to leverage my interdisciplinary background to drive positive change and innovation in the financial sector.

Part IV. About the Author

Yutong Quan is a student in the Class of 2024 at Duke Kunshan University, majoring in Political Economy with an Economics track. Interested research areas are blockchain governance, decentralized finance, and interdisciplinary fields of Artificial Intelligence and economics. 

Email: [email protected] / [email protected]

LinkedIn: https://www.linkedin.com/in/yutongquan/ 

Twitter/X: @yutong_quan



Comments
0
comment
No comments here
Why not start the discussion?