Recent Work
My work sits at the intersection of complex systems, business strategy, and technology. These are some of the problems I’ve been thinking about.
Academic Publications
Generative AI, Bullshit, and Democracy
This chapter argues that widespread deployment of large language models and other generative AI systems threatens the epistemic foundations of democracy by enabling scalable production of misinformation, disinformation, hallucinations, and what the authors call ‘bullshit’—content generated without regard for truth. It shows how these capabilities, alongside the social and environmental impacts of AI, erode a shared factual background and citizens’ ability to evaluate political claims, and it analyses the mechanisms by which actors can automate misleading political messaging. The chapter concludes with institutional and technical recommendations to defend democratic epistemic environments.
Kinkead, D. & Douglas, D. (Forthcoming) Generative AI, Bullshit, and Democracy. In K. Macnish & J. Galliott (Eds.), Artificial Intelligence and Democracy. Edinburgh University Press.
Identifying flawed reasoning in contrarian claims about climate change
This forthcoming article systematically identifies flawed reasoning patterns in contrarian climate claims using corpus analysis and human coding to categorise argumentative weaknesses, and it discusses implications for computational detection tools and pedagogical interventions.
Flack, R.; Cook, J.; Ellerton, P.; Kinkead, D.; Coan, T.; Boussalis, C.; Nanko, M.O.; Gallant, A.; Dargaville, R. (2026) Identifying flawed reasoning in contrarian claims about climate change. Environmental Communication.
This paper develops a comprehensive taxonomy of contrarian climate claims and maps common reasoning fallacies across those categories, producing a coded dataset intended to support automated detection, empirical analysis, and educational countermeasures for climate misinformation.
Flack, R.; Cook, J.; Ellerton, P.; Kinkead, D.; Coan, T.; Boussalis, C.; Dargaville, R. (2024) Identifying reasoning fallacies in a comprehensive taxonomy of contrarian claims about climate change. Environ. Commun.
This study evaluates the “Cranky Uncle” game, a humor-driven, gamified intervention that inoculates players against climate misinformation by exposing them to common misleading techniques in a playful format; experimental and classroom evaluations demonstrate improved recognition of misinformation and increased resilience to misleading tactics.
Cook, J.; Ecker, U.K.H.; Trecek-King, M.; Schade, G.; Jeffers-Tracy, K.; Fessmann, J.; Kim, S.C.; Kinkead, D.; Orr, M.; Vraga, E.; others (2023) The cranky uncle game—combining humor and gamification to build student resilience against climate misinformation. Environmental Education Research, 29(4), 607–623.
The Network and the Demos: Big Data & the Epistemic Justifications of Democracy
A stable democracy requires a shared demos and a common communication network. This chapter argues that the internet and big-data analytics have decoupled network reach from message discrimination, enabling individually tailored political content to be delivered at scale; that decoupling fosters echo chambers and targeted persuasion which undermine key assumptions of epistemic justifications of democracy (citizen independence, competence, and a shared public identity); and it offers practical suggestions for mitigating these harms.
Kinkead, D. & Douglas, D. (2020) The Network and the Demos: Big Data & the Epistemic Justifications of Democracy. In K. Macnish & J. Galliott (Eds.), Big Data and Democracy. Edinburgh University Press.
Reichenbachian Common Cause Clusters
This paper develops a probabilistic framework for “common-cause clusters,” showing how structured sets of shared causal factors can generate observed correlations; it provides existence results, characterises cluster properties, and discusses implications for causal modelling and philosophical accounts of explanation.
Mazzola, C., Kinkead, D., Ellerton, P., Brown, D. (2020) Reichenbachian Common Cause Clusters, Erkenntnis.
Reasons: A digital argument mapping library for modern browsers
This article presents Reasons, an open-source, browser-native argument-mapping library and platform; it describes the data model, interactive visualisations, and authoring tools, documents pedagogical use-cases and implementation choices, and argues that structured argument maps aid critical-thinking instruction and assessment.
Kinkead, D., Brown, D., Ellerton, P. & Mazzola, C. (2019) Reasons: A digital argument mapping library for modern browsers, Journal of Open Source Software, 4(37), 1044.
Deconstructing climate misinformation to identify reasoning errors
This paper combines corpus analysis and conceptual coding to identify recurring reasoning errors and rhetorical techniques in climate misinformation (e.g. cherry-picking, impossible expectations, false causation), evaluates how these errors propagate, and recommends educational interventions and deconstruction methods to help learners resist misleading argumentative strategies.
Cook, J., Ellerton, P., & Kinkead, D. (2018). Deconstructing climate misinformation to identify reasoning errors. Environmental Research Letters, 13(2), 024018.
What can student results tell us about school performance? (paper)
This simulation-driven paper shows how selectivity and school choice can bias inferences from student achievement measures, demonstrating via runnable simulations that commonly used ranking metrics can mislead about school effectiveness and arguing for caution in interpreting raw test-score comparisons.
Kinkead, D. (2015) What can student results tell us about school performance?. In: PESA Conference 2015: Re-engaging with politics: Re-imagining the university. Philosophy of Education Society of Australasia 2015, Melbourne, Australia, (1-23). 5-8 December 2015.
The Impact of IMPACT: How instruction in critical thinking has a beneficial impact on standardised test performance
This study reports analyses showing that targeted instruction in critical thinking (the IMPACT program) produces measurable gains on standardised assessments (reading, numeracy and writing), reporting effect sizes and discussing dataset, methods, and implications for scaling critical-thinking interventions in schools.
Kinkead, D. & Ellerton, P. The Impact of IMPACT: How instruction in critical thinking has a beneficial impact on standardised test performance. OSF
Open Source
Dependency analysis and visualisation tooling for Ruby projects. Klarity helps surface dependency structure and potential issues by analysing codebases and producing visual, navigable graphs to aid comprehension and refactoring.
A web-based argument-mapping library and application for modern browsers. “reasons” provides interactive tools for building and visualising argument maps, intended for researchers and educators who want to represent structured reasoning on the web.
Small, zero-dependency libraries for generating short, hard-to-guess, reversible IDs from auto-incrementing integers. Useful when you need compact public identifiers that are deterministic but not trivially guessable.
A command-line gem for interacting with Linear from the terminal. linear-rb provides a compact, scriptable interface to manage Linear issues and workflows without leaving your shell.
School Performance • Live Paper
An interactive paper and simulation exploring what student test results tell us about school performance. The project includes runnable simulations, visualisations, and a written paper demonstrating how selectivity and school choice affect inferences about school effectiveness.