Talks
Freshman Seminar, Department of Artificial Intelligence May 13, 2026 · 80 slides

AI Systems 2026 Preview

Guest lecture for first-year AI students previewing AI Systems 2026, covering model-to-system thinking, governance, MCP, harnesses, multi-agent SDLC, and open-weight MLOps.

AI Systems AI Education MCP Multi-Agent Systems MLOps Cheju Halla University

Event Details

  • Date: May 13, 2026
  • Location: Cheju Halla University
  • Audience: First-year students, Department of Artificial Intelligence
  • Speaker: Prof. Young Joon Lee

Overview

This freshman seminar previews the senior-level AI Systems 2026 course for first-year AI students. The central message is that AI is completed as a system, not as a model.

The lecture introduces the shift from using AI tools to designing reliable AI systems that connect models, tools, permissions, validation, telemetry, and human judgment.

Talk Outline

  1. Prologue — Why first-year students should see the system view early
  2. Model to System — From a single answer to a full workflow
  3. Governance — Designing control, approval, and accountability
  4. MCP — Connecting AI systems to tools through a common protocol
  5. Harnesses — Using loops, tests, context, instructions, and logs to absorb failure
  6. Multi-Agent SDLC — Coordinating planner, worker, reviewer, and QA roles
  7. Open-weight MLOps — Understanding local models, serving, telemetry, and evaluation
  8. Career Takeaways — Learning to build systems, not only prompts

Slides

1 / 80
Slide 1