Demystifying Generative AI
Dates
October 8, 2026
1-4pm ET
Length of Course
3 hours
Delivery Method
Online
Cost
$425 CAD
Summary
In three fast-moving hours, this masterclass demystifies the technology behind the headlines. Designed for a general professional audience with no technical background, it explains, clearly and without jargon, what generative AI actually is, how large language models work, and why they succeed and fail in the surprising ways they do. We separate hype from reality (is it a “stochastic parrot” or an “all-knowing oracle”?), trace the shift from chatbots to autonomous agents, and confront the issues that matter most to working professionals: reliability, bias, data security, copyright, governance, and the accelerating impact of AI on knowledge work. Delivered live online with interactive polls, chat, and a live Q&A, it can run as a single three-hour block or as two 90-minute sessions. Participants leave with the literacy to think clearly about AI and to use it more critically and effectively in their own work.
Key Themes:
- The “ChatGPT surprise” and what generative AI is
- How LLMs work: training, embeddings, inference, and reasoning (plain language)
- Oracle or parrot? Capabilities, limits, and “jagged” performance
- From chatbots to agents; the knowledge-work transition
- Risk, ethics, data security, copyright, and AI governance
- AI literacy in everyday professional life
Skill Outcomes
By the end of this masterclass you will be able to:
- Explain, in plain language, what generative AI is and how large language models actually work.
- Distinguish genuine capabilities from hype, and recognize where today’s models are strong and where they fail.
- Identify the key risks and ethical issues: accuracy, bias, data security, copyright, and governance.
- Appreciate how AI and “agentic” systems are reshaping knowledge work, and what that means for their own field.
Instructor
Dr. Mark Humphries Mark Humphries is a Professor of History and teaches in the Digital Humanities at Wilfrid Laurier University where he created one of Canada’s first Applied Generative AI programs. He is also a Fellow of the Balsillie School of International Affairs. A programmer since childhood, he has used machine learning and digital methods for nearly two decades. He now applies large language models to real research problems: fine-tuning models, building retrieval-augmented pipelines, and developing accuracy benchmarks for frontier systems, with his ArchivePearl webapp already in use at major archives around the world. He writes the popular Generative History Substack, and his work has been profiled by the New York Times, The Economist, and The Verge.
Format
A single three-hour live online session (can also run as two 90-minute sessions). Interactive polls and chat throughout, with a live Q&A; one short break.
Audience
This masterclass is open to anyone — in the public sector, non-profits, or industry — who wants a clear, credible introduction to generative AI. No technical background required.