Introducing generative AI to engineers in 20 minutes forces a choice: showcase its power, or build their mindset. I chose mindset.

In a recent talk, I walked through three real-world use cases that mirror common student workflows: reading dense papers, exploring new topics, and analyzing data. I demonstrated how AI can be an accelerant—if used with intention—and emphasized a simple truth: AI accelerates work, but it doesn’t replace critical thinking.

Why Engineers Must Build AI Fluency

Generative AI is already embedded in the workflows of engineers—through IDEs, productivity tools, and search. Its presence isn’t new. What matters is how intentionally and critically we apply it.

Through Project DAIL (Deeply AI Literacy), I’ve been focusing on helping students and professionals build practical fluency—knowing not just what tools can do, but how to apply them responsibly, validate their outputs, and think critically alongside them.

Reframing AI with Jobs to Be Done

As a product leader, I introduced a Jobs to Be Done (JTBD) framing to help students evaluate where AI tools add value. The question isn’t “What can it do?” but “What job am I hiring it for—and will it help?”

In this session, I focused on three jobs engineering students frequently face:

  • Learning and Research: Absorbing new topics, synthesizing papers, and developing foundational understanding for different types of learners
  • Summarization: Turning complex information into structured, usable insights
  • Data Exploration and Analysis: Sourcing, modeling, and visualizing datasets to accelerate experimentation

Other common jobs students may “hire” AI for include:

  • Debugging and Code Assistance: Identifying errors or generating snippets to explore alternative solutions
  • Communication and Documentation: Writing reports, structuring presentations, or explaining technical decisions

Each job has a clear input, desired outcome, and a way for AI to assist—without replacing the need to think.

What I Demonstrated—and What It Revealed

Demo 1: Summarizing a Research Paper with ChatGPT

I used a recent paper from Microsoft Research and CMU on The Impact of Generative AI on Critical Thinking—a “meta” choice given the context.

Three tailored prompts (for a student, engineer, and business leader) produced three distinct summaries. The student version was bulleted and connected to labs and coursework. The business version focused on reskilling and productivity.

Learning takeaway: Prompt quality drives output quality. Good prompts require clarity and responsibility.

Wow moment: A 23-page paper became a tailored insight in under a minute.

Demo 2: Exploring Topics with NotebookLM

I uploaded three papers on AI in manufacturing into Google’s NotebookLM. It generated a podcast-style audio briefing, synthesized content across sources, and supported contextual Q&A.

Learning takeaway: Multi-document research becomes more interactive—just keep an eye out for hallucinated or overconfident answers.

Wow moment: Custom audio briefings that let students “listen to the literature.”

Demo 3: Rapid Data Analysis with ChatGPT

Time ran short, but here’s what the demo showed: with three short prompts, I asked ChatGPT to simulate energy generation data, generate a table, write MATLAB code to visualize it, and extrapolate trends to 2030.

Learning takeaway: AI scaffolds technical work fast. But assumptions, validation, and accuracy still rest with the engineer.

Wow moment: A full dataset, plot, and forecast created in minutes.

Explore More

A curated set of AI learning resources—courses, podcasts, books, and blogs—is available at DeeplyProduct. The collection will continue to grow with tools, examples, and best practices.

Final Thought

This generation of engineers won’t just use AI—they’ll define how it’s used. That starts with fluency, responsibility, and curiosity. As AI tools grow more capable, our responsibility is to grow more thoughtful in how we apply them. If you’re experimenting, teaching, or rethinking how AI fits into the future of work, I’d love to connect.