Sitemap

A list of all the posts and pages found on the site. For you robots out there, there is an XML version available for digesting as well.

Posts

Future Blog Post

less than 1 minute read

Published: January 01, 2199

This post will show up by default. To disable scheduling of future posts, edit config.yml and set future: false.

Blog Post number 4

less than 1 minute read

Published: August 14, 2015

This is a sample blog post. Lorem ipsum I can’t remember the rest of lorem ipsum and don’t have an internet connection right now. Testing testing testing this blog post. Blog posts are cool.

Blog Post number 3

less than 1 minute read

Published: August 14, 2014

This is a sample blog post. Lorem ipsum I can’t remember the rest of lorem ipsum and don’t have an internet connection right now. Testing testing testing this blog post. Blog posts are cool.

Blog Post number 2

less than 1 minute read

Published: August 14, 2013

This is a sample blog post. Lorem ipsum I can’t remember the rest of lorem ipsum and don’t have an internet connection right now. Testing testing testing this blog post. Blog posts are cool.

Blog Post number 1

less than 1 minute read

Published: August 14, 2012

This is a sample blog post. Lorem ipsum I can’t remember the rest of lorem ipsum and don’t have an internet connection right now. Testing testing testing this blog post. Blog posts are cool.

portfolio

Digital Dirty Laundry: Conversational AI and the Value of Unguarded Data

Conversational AI tools such as ChatGPT, Gemini, and Claude combine large language models with conversational interfaces. These systems are trained on massive datasets that include formal writing, indexed webpages, and spontaneous personal disclosures, which are later fine-tuned through user interaction. While recent philosophical work has focused extensively on whether chatbot outputs are trustworthy, less attention has been paid to the epistemic significance of the data these systems access. Drawing on standpoint epistemology, I argue that these systems occupy an epistemic position structurally analogous to that of insider-outsiders: because they are treated as socially inconsequential, users often disclose information that would otherwise be filtered out in ordinary conversations. I refer to this class of unguarded disclosures as digital dirty laundry. This analogy does not imply that chatbots possess standpoints or enjoy epistemic agency, but it highlights how social irrelevance can generate privileged access to certain forms of evidence. Ultimately, I argue that, at scale, chatbots’ exposure to unguarded disclosures may generate a novel evidential resource for studying patterns of human behavior, bias, and self-disclosure, but that the epistemic benefits of this position accrue primarily to the organizations that control access to their training and interaction data. This creates an asymmetry in who can access, interpret, and use this novel evidence about human behavior.

Download Paper

On Using Word Embeddings to Study Scientific Metaphor

Metaphors are central to scientific explanation, yet philosophers of science lack scalable methods for studying their use. This paper introduces an approach to the study of scientific metaphor that is informed by methods in natural language processing and the digital humanities. It argues that these methods can be repurposed as a philosophical tool for analyzing scientific language at scale. Using a corpus of reproductive biology articles spanning several decades and thousands of texts, the analysis draws on familiar claims from feminist philosophy of biology about gendered metaphors in descriptions of sperm and egg cells as a case study, and operationalizes these claims to track shifting patterns of linguistic association over time. The results suggest that metaphors previously identified in the literature are recoverable within this framework and that their associated patterns change in measurable ways—often through restructuring rather than disappearance. Overall, this shows how new computational methods can extend qualitative work by making large-scale patterns in metaphor use empirically tractable.

Computing Gendered Pain

Pain research and treatment are unusually dependent on first-person testimony. This puts patients at an epistemic disadvantage—one compounded by gendered stereotypes. Women’s reports of pain are often dismissed, while pain that is acknowledged is sometimes normalized or interpreted as less clinically significant. These problems may be reproduced not only in clinical encounters but also in the scientific knowledge structures from which clinical interpretation draws. History shows that women’s pain has been psychologized or ignored. At the same time, existing philosophical accounts explain several ways in which women’s pain becomes epistemically marginalized, and social-scientific research documents non-trivial gender disparities in how patients’ pain is perceived and the urgency with which it is treated. Still, these bodies of research do not show whether—and how—such disparities are reflected in the semantic organization of the pain-management literature itself. This paper addresses this gap by examining whether the language of the pain-management literature represents patients differently based on gender. I use Word2Vec models to investigate patterns of semantic association involving credibility, emotionality, psychological explanation, pain tolerance, and drug-seeking behavior across a large corpus of scientific publications. I then compare the strength and organization of these associations in language describing female and male pain patients.

Black Box Gaslighting? You’re Imagining Things

Platforms are often accused of ‘shadowbanning’ users, but they also often deny that such practices occur. I argue that these cases can constitute a form of epistemic injustice: users may have evidence that their visibility online is being unfairly suppressed, yet lack both the system-level information needed to determine whether suppression is occurring and the shared interpretive resources needed to make sense of that experience. Because platforms control access to the relevant evidence, users are placed in a position of structurally enforced underdetermination about whether they have been wronged.

Fairness, Algorithmic Monoculture, and Randomization (with James Owen Weatherall)

Increasingly, institutional decisions, from who gets shortlisted for a job to who gets released early from jail, are made by AI-powered decision-makers. We will investigate how algorithmic monoculture, the widespread adoption of a small number of AI products to make these decisions, can generate new dimensions of unfairness; and we will develop ways to identify and reduce the harms from this form of unfairness in real-world applications.

Headlines and Hashtags: Climate News and Online Engagement

In our technologically saturated society, social media platforms increasingly shape public perception of urgent issues, including climate change. However, despite our increasing media exposure and a significant increase in climate-related disasters, public concern about climate change has not increased proportionally. Using data from over 10 million tweets and 140,000 news articles from 2016 to 2018, this study examines the impact of media coverage on online engagement with climate change. This research analyzes the relationships between the volume and sentiment of climate-related news stories and tweets. Although a modest positive correlation was measured between the sentiment of tweets and news articles, a negative correlation was noted across all other metrics. Specifically, there was a negative correlation between the average stance and sentiment of climate-related tweets and the number of climate-related news articles per month and between the average stance of these tweets and the average sentiment of the news articles. These results suggest that increased media coverage may not foster public urgency as expected, raising questions about the media’s role in influencing public perception. This highlights the need for further research into the impact of news coverage on politicized issues.

The Problem With Who I Know: What Contextualism Can Tell Us About Interpersonal Knowledge Claims

‘I know his name.’ ‘I know something about him.’ ‘I know him.’ Consider how these uses of ‘know’ differ. The first two instances of know, seem to point to knowledge about something. Yet in the latter claim, the subject of the assertion is not a singular fact, but another person. I call these knowledge claims interpersonal knowledge. In the following paper, I provide an account for these interpersonal knowledge claims which employs the Conversational Contextualist view of language by synthesizing Allan Gibbard’s Norm-Expressivist account for ‘good’ with an account of knowledge based in social epistemology. Under my theory ‘knowing someone claims’ amount to endorsements of our beliefs; as such, there is no truth-apt interpersonal knowledge. What is occurring is a self-assessment of our relationship to another person, based on our non-cognitive attitude towards the fact that we should know them. Therefore interpersonal knowledge claims are self-affirmations that assert we are doing what we believe we should, in an attempt to embody our perceived relationship with another person.

Download Paper

Defending the Hypothesis of Indifference

The problem of evil is the philosophical question regarding how to reconcile the existence of an omnipotent, omnibenevolent, and omniscient God with the pain and suffering in the world. The Hypothesis of Indifference is Paul Draper’s proposal considering that question. His claim is that the pain and pleasure we experience in our lifetimes has nothing to do with God or some other supernatural force acting as an agent of good or evil. In this paper, I argue that Draper’s Hypothesis of Indifference is a better explanation for why we experience pain and pleasure than theism is and that it survives major contemporary criticisms posed by Peter van Inwagen and William Alston.

Download Paper