Humanity's Last Exam
Co-author on a multi-modal LLM benchmark designed to be the final closed-ended academic benchmark of its kind with broad subject coverage. Project of The Center for AI Safety (CAIS).
A comprehensive portfolio spanning academic research, AI ethics, technology education, and community engagement.
Formal qualifications spanning philosophy, technology, and education
Publications, research contributions, and academic work
Co-author on a multi-modal LLM benchmark designed to be the final closed-ended academic benchmark of its kind with broad subject coverage. Project of The Center for AI Safety (CAIS).
A comprehensive linguistic and ethical analysis of key terms and arguments relevant to the ethical concepts of eudaimonia and makaria across several centuries of ancient Greek thought. Published by Peter Lang Press.
Chapter in Seelow (ed.) Lessons Drawn: Essays on the Pedagogy of Comics and Graphic Novels. An approach to teaching philosophy inclusively, through Gaiman's Sandman graphic novels. Published by McFarland Press.
"Aristotle, Aquinas, and the Convertibility of Truth and Being", Philosophical Writings. Aquinas's metaphysics of adequation prevents his thought from being considered a philosophical extension of Aristotelian theory.
Talks, presentations, and contributions to AI ethics discourse
Invited talk at Institute of Management Accountants Northeast Regional Council annual meeting, on financial risks of generative AI in business contexts.
Public talk sponsored by Williams College Philosophy Department, on using classical rhetorical techniques to bypass safety guardrails in generative AI models at a 2024 red-teaming event.
Member of the IEEE working group for Ethically Aligned Design: A Vision for Prioritizing Human Well-being with Autonomous and Intelligent Systems, a standards document to prioritize ethics in the creation of AI systems.
Invited participant in CAMLIS AI red-teaming exercise sponsored by NIST and Humane Intelligence. Collaborated with industry and government partners to test generative AI security using NIST AI 600-1 framework. Co-authored evaluation report.
Presentation at NERD Summit, UMass-Amherst: "Conversational Computing; or, how genAI reveals (and subverts) our HCI assumptions". An exploration of how tech folkways can mismatch our intentions with genAI.
Invited speaker at Osher Lifelong Learning Institute series on technology and its social impacts. Presented in conjunction with Berkshire Innovation Center, 1Berkshire, Williams College, MCLA, and Bard College at Simon's Rock.
Interview on the "John Krol Show," a local television program in Pittsfield, MA. Discussed AI ethics research at MCLA in the show's "Rapid Research" segment.
Signatory on the "Autonomous Weapons: an Open Letter from AI & Robotics Researchers," published by the Future of Life Institute. Early engagement with AI ethics and autonomous systems governance.
Educational initiatives and course development
Williams College course (PHIL-098) interrogating original texts from antiquity on ethical concerns relevant to artificial entities: agency, control, sentience, language, creativity, soul, and more.
Summer course for Osher Lifelong Learning Institute. Surveying ethical issues confronting us with the growth of generative AI, from commerce to the arts to education to the military to health care.
A 5-week training course designed to certify faculty to teach online courses. Topics: pedagogy, UX, design, accessibility, online presence, media, troubleshooting. CC-BY-SA, designed natively in Instructure Canvas LMS.
An introductory philosophy curriculum redesigned as a hybrid gamified RPG. Case study in Bell, K, Game On! Gamification, Gameful Design, and the Rise of the Gamer Educator, published by Johns Hopkins University Press.
Open-source tools, technical projects, and community resources
A complete, minimalist attendance solution for college instructors. Generate QR codes for students directly from Canvas, then use a scanner app to take attendance in large classrooms. Everything runs locally with full privacy protection.
Kit for local genAI experimentation, on loan through Williams College libraries. OLMo large language model on Raspberry Pi 5 via the experimental Mozilla Llamafile project.
Blogpost for Williams Makerspace, recipe for spinning up a hyper-local Apache webserver on a Raspberry Pi for community access: festivals, teach-ins, emergencies (e.g. ISPs and cell towers non-functional).
Presentation at Connecticut Academic Library Conference: "Multilingual Media Editing with Local, Open AI Tools". Demonstration of local AI to produce film captions. Co-presenter with Trevor Murphy, Williams College.
Code project developed as a participant in the Critical Code Studies Working Group at the University of Southern California, to generate accessible content from StoryMapJS (a project of Knight Lab at Northwestern University).
Writing and creative projects
A collection of microfiction, poetry, and images written throughout April 2020, conveying imaginings built on the uncertainties of worlds, and lives, disrupted in a time of global tragedies. Published by Shires Press.
I'm available for consulting on AI ethics, speaking engagements, educational workshops, and collaborative projects.
Get In Touch