About

D. David Williams is an Assistant Professor of Classics at the University of Virginia, specializing in the intellectual history of archaic and classical Greece. His research particularly focuses on Athenian drama, with a special emphasis on Aristophanes, as well as on Socrates, the sophists, and Plato. Williams completed a joint PhD from the Department of Classics and the Committee on Social Thought at the University of Chicago in 2022. Prior to his current position at UVA, he was a Solmsen Fellow at the Institute for Research in the Humanities at the University of Wisconsin–Madison during 2023-24 and a visiting scholar in the Department of Classical Studies at the University of Pennsylvania in 2022-23. Williams is engaged in three long-term research projects. The first investigates Aristophanes' engagement with contemporary intellectual culture. The second is a collaborative volume on the sophist Hippias of Elis. The third project is a comprehensive study of the use of gnomic statements on the tragic stage. His work contributes to a deeper understanding of classical intellectual traditions and their representation in ancient Greek literature and drama.

Research topics

• Political Science
• Business
• Computer Science
• Finance
• Public administration
• Public economics
• Accounting
• Medicine
• Law
• Economics
• Macroeconomics

Selected publications

• Classport: Designing Runtime Dependency Introspection for Java

  SSRN Electronic Journal · 2026-01-01

  preprintOpen access
  PublisherDOI

• Using Large Language Models to Forecast Local Government Revenue

  Public Finance Journal · 2025-06-08 · 4 citations

  articleOpen accessSenior author

  We examine the use of a public access large language model (LLM) to make local government revenue forecasts. ChatGPT is an LLM that is not specifically designed to perform quantitative analysis. However, it is capable of completing a wide range of tasks. The goals of this article are to determine the accuracy that can be obtained and to examine its potential bias. This study is based on a government revenue dataset from the Government Finance Officers Association (GFOA). The benefits of determining the accuracy and bias of LLM forecasts include providing a low-cost forecast method for small- and medium-sized governments and enabling external observers to validate forecasts made by official sources. Discovering the limitations of ChatGPT and similar LLMs, as well as the specific conditions required to use them wisely, may help localities avoid adverse outcomes. We find that a combination of LLM and human input provides a viable alternative forecasting method for small- and medium-sized governments, and it enables external observers to validate forecasts made by official sources. Errors in forecasting with the human-in-the-loop can be as low as 9.9 percent at the aggregated annual level. Using ChatGPT results alone can lead to high-error forecasts that may not be reliable.

  PublisherDOI

• AI-driven cloud-edge synergy in telecom: An approach for real-time data processing and latency optimization

  World Journal of Advanced Engineering Technology and Sciences · 2025-03-30 · 1 citations

  articleOpen accessSenior author

  In recent years, the telecommunication industry has seen significant advancements with the integration of AI, cloud computing, and edge computing. These technologies, when combined, enable telecom providers to process data more effectively, minimize latency, and enhance service delivery. This paper explores the synergy between AI, cloud, and edge computing in the telecom sector, highlighting innovative approaches to real-time data processing and latency optimization. Through a deep dive into emerging trends, this article identifies novel methodologies and applications in AI-driven cloud-edge integration, with a focus on telecom infrastructure, 5G networks, and IoT ecosystems.

  PublisherDOI

• Establishing an Agenda for Public Budgeting and Finance Research

  Deleted Journal · 2024 · 20 citations

  • Political Science
  • Political Science
  • Public administration

  Public budgeting and finance is a discipline that encompasses communities of research and practice. Too often, however, these communities fail to engage each other, instead choosing to operate independently. The result is that the research being conducted fails to address the questions of the day and our governments’ challenges. In this article, we come together as a community of academics and practitioners to establish an agenda for where future research should be conducted. This agenda aims to align the research being undertaken within the academic community with the needs of those working in the community of practice. After establishing ten areas where research is needed, we followed a ranked-choice voting process to establish a prioritization for them. Based on the outcome of this process, the two primary areas where research is currently needed most are the fiscal health of our governments and the implementation of social equity budgeting.

  PublisherOA PDFDOI

• For Better or Worse? Revenue Forecasting with Machine Learning Approaches

  Public Performance & Management Review · 2022-05-22 · 18 citations

  article

  The recent rapid development of artificial intelligence (AI) is expected to transform how governments work by enhancing the quality of decision-making. Despite rising expectations and the growing use of AI by governments, scholarly research on AI applications in public administration has lagged. In this study, we fill gaps in the current literature on the application of machine learning (ML) algorithms with a focus on revenue forecasting by local governments. Specifically, we explore how different ML models perform on predicting revenue for local governments and compare the relative performance of revenue forecasting by traditional forecasters and several ML algorithms. Our findings reveal that traditional statistical forecasting methods outperform ML algorithms overall, while one of ML algorithms, KNN, is more effective in predicting property tax revenue. This result is particularly salient for public managers in local governments to handle foreseeable fiscal challenges through more accurate predictions of revenue.

  PublisherDOI

• Evaluating the Comparative Accuracy of COVID-19 Mortality Forecasts: An Analysis of the First-Wave Mortality Forecasts in the United States

  Forecasting · 2022-09-29 · 3 citations

  articleOpen accessSenior author

  The sudden onset of the COVID-19 pandemic posed significant challenges for forecasting professionals worldwide. This article examines the early forecasts of COVID-19 transmission, using the context of the United States, one of the early epicenters of the crisis. The article compares the relative accuracy of selected models from two forecasters who informed government policy in the first three months of the pandemic, the Institute of Health Metrics and Evaluation (IHME) and Columbia University. Furthermore, we examine whether the forecasts improved as more data became available in the subsequent months of the pandemic, using the forecasts from Los Alamos National Laboratory and the University of Texas, Austin. The analysis focuses on mortality estimates and compares forecasts using epidemiological and curve-fitting models during the first wave of the pandemic from March 2020 to October 2020. As health agencies worldwide struggled with uncertainty in models and projections of COVID-19 caseload and mortality, this article provides important insights that can be useful for crafting policy responses to the ongoing pandemic and future outbreaks.

  PublisherOA PDFDOI

• Early [18]FDG PET/CT scan predicts tumor response in head and neck squamous cell cancer patients treated with erlotinib adjusted per smoking status

  Frontiers in Oncology · 2022-08-30 · 1 citations

  articleOpen access

  Translational Relevance: Evaluation of targeted therapies is urgently needed for the majority of patients with metastatic/recurrent head and neck squamous cell carcinoma (HNSCC) who progress after immunochemotherapy. Erlotinib, a targeted inhibitor of epidermal growth factor receptor pathway, lacks FDA approval in HNSCC due to inadequate tumor response. This study identifies two potential avenues to improve tumor response to erlotinib among patients with HNSCC. For the first time, this study shows that an increased erlotinib dose of 300 mg in smokers is well-tolerated and produces similar plasma drug concentration as the regular dose of 150 mg in non-smokers, with increased study-specific defined tumor response. The study also highlights the opportunity for improved patient selection for erlotinib treatment by demonstrating that early in-treatment [18]FDG PET/CT is a potential predictor of tumor response, with robust statistical correlations between metabolic changes on early in-treatment PET (4-7 days through treatment) and anatomic response measured by end-of-treatment CT. Purpose: Patients with advanced HNSCC failing immunochemotherapy have no standard treatment options. Accelerating the investigation of targeted drug therapies is imperative. Treatment with erlotinib produced low response rates in HNSCC. This study investigates the possibility of improved treatment response through patient smoking status-based erlotinib dose optimization, and through early in-treatment [18]FDG PET evaluation to differentiate responders from non-responders. Experimental design: In this window-of-opportunity study, patients with operable HNSCC received neoadjuvant erlotinib with dose determined by smoking status: 150 mg (E150) for non-smokers and 300 mg (E300) for active smokers. Plasma erlotinib levels were measured using mass spectrometry. Patients underwent PET/CT before treatment, between days 4-7 of treatment, and before surgery (post-treatment). Response was measured by diagnostic CT and was defined as decrease in maximum tumor diameter by ≥ 20% (responders), 10-19% (minimum-responders), and < 10% (non-responders). Results: Nineteen patients completed treatment, ten of whom were smokers. There were eleven responders, five minimum-responders, and three non-responders. Tumor response and plasma erlotinib levels were similar between the E150 and E300 patient groups. The percentage change on early PET/CT and post-treatment PET/CT compared to pre-treatment PET/CT were significantly correlated with the radiologic response on post-treatment CTs: R=0.63, p=0.0041 and R=0.71, p=0.00094, respectively. Conclusion: This pilot study suggests that early in-treatment PET/CT can predict response to erlotinib, and treatment with erlotinib dose adjusted according to smoking status is well-tolerated and may improve treatment response in HNSCC. These findings could help optimize erlotinib treatment in HNSCC and should be further investigated. Clinical Trial Registration: https://clinicaltrials.gov/ct2/show/NCT00601913, identifier NCT00601913.

  PublisherOA PDFDOI

• The Challenges of Project Expos and Pop-Up Voting

  2020-01-01

  book-chapter1st authorCorresponding
  PublisherDOI

• Introduction

  2020-01-01

  book-chapter1st authorCorresponding
  PublisherDOI

• Between Policy Promises and Program Implementation

  2020-01-01

  book-chapter1st authorCorresponding
  PublisherDOI

Frequent coauthors

• Thad D. Calabrese

  13 shared

• Michael D. Cook

  National Cancer Institute

  10 shared

• Luis Alfredo Rodríguez

  9 shared

• Don Waisanen

  8 shared

• Mario Werner

  8 shared

• Jack W. Davidson

  6 shared

• Hans Krueger

  6 shared

• Sonia Lamont

  BC Cancer Agency

  6 shared

Labs

• University of Virginia ClassicsPI

Education

• Ph.D., Public Administration/Policy Analytics

  Virginia Commonwealth University

  1995

• M.Ed., Social Foundations of Education

  University of Virginia

  1979

• BA, Philosophy

  University of Virginia

  1975

Awards & honors

• Solmsen Fellow at the Institute for Research in the Humaniti…
• visiting scholar in the Department of Classical Studies at t…