We've heard this story before. When the Internet made technical knowledge easier to access, people said computer science expertise would matter less. When Stack Overflow made solutions searchable and shareable, the same claim returned. Today, generative AI has made it faster and easier to produce code, and once again people are asking whether formal computer science education is becoming obsolete.
This talk argues the opposite. The easier it becomes to produce code, the more important it becomes to understand computation in depth. Computer science has never been just about writing programs; it spans correctness, abstraction, algorithms, complexity, systems, design, security, maintainability, and consequences. Some of that work is visible in the public-facing software people use every day, but much of it happens in subtler mathematical, theoretical, and infrastructural spaces, where progress depends on identifying and solving problems at the edge of what we currently understand. Generative AI can produce plausible output, but plausibility isn't progress, output isn't accountability, and neither replaces understanding.

Many Atlantic Canadian companies have started exploring AI in small ways. They have tried new features in Microsoft 365, asked ChatGPT a few questions, or experimented with meeting transcription tools. These early steps show curiosity and openness to change, but most businesses have not yet turned that exploration into lasting impact.
In this short talk, Tony Sheehan, Co-Founder and CTO of NorthBound Advisory, will share what helps small and mid-sized organizations move from trying AI to truly adopting it. Through practical examples from across Atlantic Canada, Tony will highlight three simple lessons on how any business can start building real productivity gains with AI.

This talk presents a research perspective on Responsible AI centered on the progression from understanding machine learning systems to controlling their behavior. It begins by examining how AI systems can be understood through Explainable AI (XAI), and why explainability is essential for transparency, accountability, and trust. The presentation will review major directions in explainability for machine learning, including interpretable models, global and local explanation methods, and data-centric approaches, and will discuss our work on post-hoc, model-agnostic, and data-centric methods for explaining complex black-box models. It will then argue that explainability, while fundamental, is only one component of trustworthy AI, which also requires reliability, faithfulness, safety, privacy, and adaptability. From this foundation, the talk will move to the problem of control, considering how AI systems may be guided, edited, or made to unlearn specific information in privacy-preserving and minimally invasive ways. The final part of the talk will address the new challenges introduced by large language models and agentic systems, including questions of verification, controllability, information routing, autonomy, and responsible deployment, and will highlight these as important open problems for the next generation of Responsible AI research.

AI is reshaping public health, from epidemic surveillance and disease prediction to identifying social determinants that drive health inequities. This session showcases real-world deployments and critical ethical questions around surveillance, consent, and equitable access to AI-enabled health tools.

Rural and remote communities face chronic shortages of healthcare providers and diagnostic resources. This presentation explores how AI-powered diagnostic tools, telemedicine, and remote monitoring are bridging geographic gaps — and what infrastructure and policy changes are needed to make these solutions equitable and sustainable.

Canada has positioned itself as a global leader in AI research and policy — but translating academic leadership into commercial and industrial competitiveness requires deliberate investment and coordination. This talk examines Canada's national AI strategy, the role of Atlantic Canada's emerging tech ecosystem, and what the next decade holds.

Small and medium-sized enterprises (SMEs) are increasingly adopting AI tools, yet face unique barriers — limited data, smaller budgets, and fewer technical staff. This session provides practical guidance on accessible AI strategies that SMEs can implement today to improve operations, customer experience, and competitiveness.

Scientific discovery is being accelerated by AI in unprecedented ways — from protein structure prediction in biomedical labs to autonomous underwater vehicles mapping uncharted ocean ecosystems. This presentation highlights breakthrough research applications and what they mean for Atlantic Canada's ocean economy and research institutions.

Fredericton has established itself as one of Canada’s most promising small-city innovation ecosystems, with recognized strengths in cybersecurity, research, and talent development. But the next phase of growth will require more than momentum, it will demand greater coordination, sharper positioning, and a more deliberate approach to scaling what works.
In this talk, Josh shares insights from recent strategic work and stakeholder engagement across industry, academia, and government, outlining both the opportunities and structural challenges shaping the region’s future. He will explore how Knowledge Park is evolving to play a more central role in this next stage, acting as a platform for alignment, investment attraction, and ecosystem development.
The session will offer a candid look at what it takes to move from a strong regional ecosystem to a globally competitive innovation hub, and what Fredericton must do next to get there.