AI and Design

SENG 365 — Software Engineering

Neil Ernst

nernst@uvic.ca

University of Victoria

2026-07-24

Learning Objectives

Learning Objectives

Placeholder — to be developed.

• Use AI tools to support design work while remaining accountable for quality
• Critically evaluate AI-generated designs, code, and documentation
• Document AI-assisted work transparently
• Understand how AI shifts the engineer’s role toward direction and evaluation

Topics to Develop

Directing an Agent

TODO: prompting as specification; giving an agent design context; iteration.

Evaluating AI Output

TODO: what “good” looks like; spotting popular-but-wrong choices; security and maintainability review.

Cognitive Debt and Legibility

TODO: maintaining a mental model; diagnostic legibility. Link to philosophy.md.

Transparency and Accountability

TODO: AI usage logs; disclosure; where the human stays responsible. Link to ethics.qmd.

In-Class Activity

Critique an Agent’s Design

TODO: present an AI-generated design/PR; students identify the tradeoff made and what it missed.