51茶馆

• Nathaniel McLeroy,听Graduate Student, School of Social Work, and Graduate Production Assistant, Media Technology Services

[Two Presentations in 1 hour] [Pair A]

As AI reshapes labor, education, health, housing, and public safety, its impacts fall earliest and hardest on communities already facing structural vulnerability. Yet higher education鈥檚 AI discourse often centers on pedagogy and productivity, overlooking the social systems into which these technologies are deployed.

This session argues that AI governance is fundamentally a social challenge. Social work offers tools that AI efforts lack: systems thinking, equity analysis, community engagement, and an understanding of unintended consequences. But the field itself remains underprepared, with limited training on how emerging technologies shape policy, practice, and social outcomes.

The session examines this dual gap and makes the case for integrating social-work-informed frameworks into AI education across disciplines. Using examples from workforce development, community practice, and AI governance, it outlines practical steps universities can take to build more ethical AI infrastructure and prepare a workforce capable of navigating technological change.

This is a 30-minute presentation paired with "Linguistics and AI: an inquiry-based course from a disciplinary perspective."

• Christopher Geissler,听Visiting Assistant Professor of Linguistics,听Eastern, Slavic, and German Studies (MCAS)
• Emily Hay '26, Linguistics major
• Jasmine Maas '26, Linguistics major
  听

[Two Presentations in 1 hour] [Pair A]

Linguistics and Artificial Intelligence, a new course taught in Fall 2025, turned the tools of linguistics to study text-to-speech (TTS) systems and large language models (LLMs). Each student conducted five small-scale research projects, writing two-page abstracts formatted like submissions to computational linguistics conferences. Students defined their own research questions, but were constrained by topics appropriate for a particular methodology: vocal resonance, pronunciation, syntactic variation, and discourse analysis.. This format required students to critically examine AI systems, while leveraging their developing knowledge of linguistics to understand this new topic. Students report developing a greater understanding of what AI systems are and reflecting on AI systems in new ways. Overall, the course provides a case study in how disciplinary study and learning about AI can benefit each other.

This is a 30-minute presentation paired with "AI as Social Infrastructure: Why Tech Literacy Needs A Social Work Lens."听

• Vincent Cho,听Associate Professor,听Educational Leadership and Higher Education, LSEHD

[Two Presentations in 1 hour] [Pair B]

This hands-on workshop invites faculty to confront a practical question: where in your teaching do students get stuck, and could a well-designed chatbot help? Drawing on the facilitator's experience developing custom chatbots for graduate students in professional programs, the session moves participants from identifying a specific friction point in their teaching to drafting a deployable chatbot prompt. Along the way, it surfaces the design decisions that matter: how rubrics and assignment expectations can be made visible through prompt design, how guardrails shape what a chatbot will and will not do, and how involving students in that process deepens their understanding of what is being asked of them. The workshop assumes no prior experience with custom chatbot design and is intended to leave participants with both a working tool and a framework for thinking about when one belongs in their teaching.

This is a 30-minute presentation paired with "Socratic AI: Adaptive Oral Assessments, AI-Driven Conversations and Pedagogy in the Classroom."

• Can Erbil,听Professor of the Practice, Economics

[Two Presentations in 1 hour] [Pair B]

In first-year courses, generative AI has quietly broken traditional assessment. Written homework and short answers no longer reveal what students actually understand. This talk presents a set of applied, classroom-tested strategies for redesigning introductory courses in an AI-rich environment. Drawing on large-enrollment teaching at Boston College, it shows how AI-supported oral assessments, AI-augmented lectures, and adaptive learning tools can be used to require students to explain concepts aloud, apply ideas in real time, and demonstrate understanding that cannot be outsourced to a chatbot. Rather than banning AI, these approaches integrate it directly into course design while restoring clarity, rigor, and engagement in foundational courses. The session focuses on concrete implementation choices, what worked, what failed, and how instructors can immediately adapt these methods to first-year classrooms.

This is a 30-minute presentation paired with "Making the Invisible Visible: Designing Custom Chatbots to Support Student Work."

• Chris Strauber, Senior Liaison Librarian,听University Libraries
• Steve Runge, Senior Liaison Librarian, O'Neill Library

[Two Presentations in 1 hour] [Pair C]

Conversations about Generative AI at the 51茶馆 libraries tend to ask some ethical and epistemological questions the general discourse does not. This presentation will discuss how LLM鈥檚 include only a fraction of human knowledge, and how LLMs obscure scholarly communication.

The Internet is at best a convenience sample of human knowledge. Most archives exist primarily on paper, and archives are only one possible store of knowledge. Much knowledge is lost entirely, much is in minority languages with limited web presence, and much is yet to be found or archived, let alone digitized.
LLM chatbots hide their sources. Most companies have been evasive about what data LLMs have ingested, and that data is, in the words of one researcher, a 鈥渟lurry鈥� of information. Few people know how LLMs develop parameters in their training and fewer yet the programmatic sources of unpredictable or inaccurate output.

This is a 30-minute presentation paired with "The Coming AI Crisis: Why Most Companies Are Already Out of Bounds."

• Lindsay Timcke, Managing Partner Timcke Risk Management LLC, Accounting

[Two Presentations in 1 hour] [Pair C]

AI adoption is accelerating far faster than the governance structures required to manage its ethical, societal, and operational risks. More than 70% of organizations now deploy generative AI, yet fewer than 20% maintain formal governance frameworks, and only a small minority document model lineage or training鈥慸ata provenance. This gap has become a defining ethical risk: opaque systems are making consequential decisions without explainability, auditability, or accountability. Regulators are responding with escalating expectations around transparency, safety testing, and verifiable control, signaling that AI will be treated as a regulated system rather than a productivity tool. Meanwhile, structural weaknesses鈥攗nstructured data, fragile pipelines, shadow AI use, and undisclosed vendor models鈥攁re amplifying systemic exposure. This presentation examines why AI opacity is emerging as a societal risk, how liability shifts to deploying organizations, and why verifiable, governed, and independently validated AI is now the ethical baseline for responsible enterprise adoption.

This is a 30-minute presentation paired with "LLM鈥檚 Are Not What They Say They Are: A View from the Library"

• Kyle Fidalgo,听Academic Technologist,听Law School
• Maureen Van Neste, Associate Professor of the Practice, Law School
• Jake Samuelson, Legal Information Librarian & Lecturer in Law, Law School听
• Raul Carrillo, Assistant Professor, Law School
• Ross Martin, Adjunct Professor, Law School

[Panel Discussion]

This session brings together faculty, librarians, and academic support staff from 51茶馆 Law to share practical approaches to AI integration across teaching, learning, and administrative functions. Through lightning-style presentations, panelists will demonstrate how they've implemented AI tools in their daily work鈥攆rom developing educational workshops using custom AI assistants and templates, to integrating chatbots into classroom pedagogy, to designing immersive simulation exercises for 1L students. Rather than theoretical speculation, these short talks focus on real workflows, lessons learned, and tangible outcomes from our ongoing AI initiatives. Attendees will gain insight into the 51茶馆 Law AI Fluency program, see examples of AI-enhanced course design in legal education, and learn how different roles across a school can collaborate to build institutional AI literacy. The session concludes with Q&A, offering an opportunity to discuss challenges, considerations for legal education contexts, and strategies for cross-functional AI adoption.

• Adrian Aziza,听Boston College student,听Applied Data Science
• Julia DeVoy, PhD, MTS, MBA, MLS '26, Associate Dean of Undergraduate Programs and Students, LSEHD & Co-Founder of Inter-institutional Design for Impact Initiative

[Workshop/Demo]

This session presents an epistemic, student-facing AI study tool that turns course materials into do-now actionable steps while supporting knowledge-making rather than rote memory. The AI tool鈥檚 workflow scaffolds core epistemic moves: question formation, claim鈥揺vidence mapping, uncertainty calibration, counterargument testing, and next-action experiments (what to verify, read, ask, or measure) and tags each step to course learning objectives so individual progress becomes visible. Instead of producing final submissions, the AI Study Tool uses a voice-preserving refactor loop (student inputs draft; AI offers cited revision strategies and reasoning; student then chooses and rewrites) plus explicit constraint capture (rubric, audience, citation rules). A provenance layer: AI Dialogue Log, change history, and verification ledger; records prompts, short output excerpts, what the student adopted/overruled, and what was fact-checked, producing a concise 鈥榩rogress brief鈥� that enables more individualized, strategic feedback. The result is a learning-centered pattern for ethical, transparent AI use that strengthens reasoning and authorship.

• Mimi Tam,听Computer Science/Cybersecurity/AI-ML Professor,听Woods Computer Science Department

[Presentation]

Join Dr. Mimi Tam for a strategic roadmap into the Boston College AI Ecosystem. Moving beyond simple automation, this session explores how 'Institutional Intelligence' integrates Agentic AI across every pillar鈥攆rom faculty research and student success to campus operations. Discover how 51茶馆 can scale Cura Personalis through a governance-first framework that balances cognitive rejuvenation with ethical, institutional-grade security.

• Philip Aldrich,听CTO - Verterim, Adjunct Professor - 51茶馆
• Alexia Prichard, Senior Instructional Media Producer, 51茶馆; Winner 2025 MIT AI Film Hackathon; SXSW FuturePIXEL House 2026-Official Screening

[Presentation]

Complex learning concepts are difficult to learn, much less understand and retain. What if students could "experience" the concept through AI and VR? This presentation will outline real world problems Chief Information Security Officers and Chief Risk Officers face at many organizations. The traditional teaching method would be to show students a busy visual meant to convey intricate scenarios and problems without experience or context. This presentation will show how AI and VR can bring these concepts to life for students to watch, interact, and engage within the reality of the concept itself.

• Melanie Hubbard, Head of Digital Scholarship & Data Services
• Micah Lott, Associate Professor of Philosophy
• Paula Mathieu, Associate Professor of English
• Dave Thomas, Digital Scholarship Specialist (Libraries)
• Andrea Vicini, SJ, Michael P. Walsh Professor of Bioethics (Theology)

[Panel Discussion]

The 鈥淲hat Is Wrong with Generative AI鈥� panel will feature contributors from the Departments of English, Theology, and Philosophy, as well as the Libraries. Drawing on their disciplinary perspectives and technical expertise, they will articulate the issues they believe students and faculty should understand to engage with generative AI critically and ethically within and beyond the classroom. Generative AI鈥檚 embedded biases, frequently misunderstood capabilities, depersonalizing tendencies, effects on critical thinking, and material impacts, among other topics, will be discussed.听
Despite the deliberately provocative title, this session is not intended to be a takedown of generative AI. Instead, it aims to give attendees a more holistic and realistic understanding of what it is, its capabilities, and its impacts, recognizing that it is here to stay. There will be ample time for audience responses and questions.