The traditional "calculator analogy"—suggesting AI is merely a faster way to compute—is now dangerously insufficient. Generative AI and advanced machine learning models demonstrate the capacity to write code, optimize complex topologies, and synthesize design constraints with a speed and accuracy that rivals human capability. Consequently, engineering education faces an existential pivot: we must transition from teaching students how to solve problems to teaching them how to orchestrate the solution of problems through intelligent agents.

This presentation addresses the core question: What shall we do with AI in Engineering Education? It argues that the response must go beyond simply adding "AI Ethics" or "Prompt Engineering" as elective modules. Instead, it proposes a fundamental restructuring of the engineering pedagogy based on three pillars: Cognitive Offloading, Verification Literacy, and Human-Centric Systems Thinking.

First, we examine Cognitive Offloading. If AI can handle syntax, calculation, and routine derivation, engineering education must abandon the "boot camp" mentality of rote memorization. The curriculum must shift focus toward "Problem Formulation." In an era where answers are cheap, the value of an engineer lies in asking the right questions. This presentation outlines a pedagogical framework where students are assessed not on their ability to manually derive a solution, but on their ability to decompose complex, ambiguous real-world problems into architectures that AI can process.

Second, we introduce the concept of Verification Literacy. As reliance on black-box algorithms increases, the primary technical skill of the 2026 engineer must be scepticism. We discuss methodologies for teaching "adversarial engineering"—training students to audit, stress-test, and debug AI-generated outputs. The engineer of the future is not just a creator, but a sophisticated editor and validator. We will present case studies of "flipped assessment" models, where students are graded on their ability to identify flaws in AI-generated engineering blueprints.

Third, we explore Human-Centric Systems Thinking. As technical barriers lower, the engineer's role expands into social, ethical, and environmental domains. We argue that AI allows us to re-humanize engineering education. With the technical drudgery automated, the curriculum can finally prioritize empathy, interdisciplinary communication, and ethical foresight. We propose a "Human-in-the-Loop" educational model where AI serves as a personalized Socratic tutor, freeing faculty to mentor students in high-level critical thinking and professional judgement.

Finally, the presentation confronts the digital divide. We will discuss the infrastructure required to ensure that AI-augmented education is accessible globally, preventing a tiered system where only resource-rich institutions produce "super-engineers."

We stand at a crossroads. We can continue to train students for a world that no longer exists, competing futilely against algorithms, or we can embrace a symbiotic future. This presentation offers a roadmap for the latter, defining the "AI-Augmented Engineer" not as a user of tools, but as an architect of intelligence.