The Launch of Something New
We launched the GW Engineering AI Academy with our inaugural session in November 2025. Fifteen faculty members from across the School of Engineering and Applied Science gathered (virtually) to begin a six-month journey exploring how artificial intelligence can transform our teaching, research, and daily work.
The goal of our first session was ambitious but clear: to shift our thinking from individual AI tools to integrated workflows that amplify both productivity and creativity.
Beyond the Shiny Object Syndrome
In the current landscape of AI tools, it's easy to fall into what we might call "shiny object syndrome"—constantly chasing the newest application, comparing features, trying to identify the single "best" tool for working with AI. This approach misses something fundamental about how AI can genuinely enhance our work.
The real power emerges not from any single tool, but from understanding how to orchestrate multiple AI capabilities into coherent workflows. This requires a mindset shift: from being an AI "operator" to becoming an AI "orchestrator"—someone who manages a team of digital specialists, each with distinct strengths.
The Research & Synthesis Workflow
To demonstrate this principle in action, we explored a concrete example: a research and synthesis workflow that any faculty member could implement immediately.
The Workflow Components
Step 1: Curate with Perplexity (The Researcher) We started with a simple prompt in Perplexity, an AI-first search engine: "Find out everything you can about the George Washington University School of Engineering and Applied Science with a target audience of potential students and parents."
Unlike traditional chatbots, Perplexity automatically searches the web, compiles sources, and creates a comprehensive report—complete with citations to every claim. In less than 10 minutes, it autonomously:
- Identified and reviewed 20+ relevant web pages
- Explored financial aid, rankings, and facilities
- Generated a detailed report with live links to all sources
The key advantage: every statement in the report links directly to its source, enabling verification—what we call "grounding" the AI output in verifiable truth.
Step 2: Synthesize with NotebookLM (The Analyst) Next, we took the Perplexity report and uploaded it to NotebookLM, along with the GW logo and brand color palette. With a single prompt—"Create a slide deck for admissions professionals visiting high schools and promoting GW SEAS to potential students and parents"—NotebookLM generated a complete presentation.
The results were remarkable:
- Professional slides with GW branding consistently applied
- Automatically sourced and incorporated relevant images
- Data visualizations and statistics drawn from the research
- Verified short URLs and contact information
Step 3: Refine with Gemini (The Designer) While NotebookLM produces high-quality PDFs, they aren't immediately editable (yet). So we took one more step: uploading the PDF to Google Gemini with the prompt "Create a Google Slides presentation using the PDF attached."
Gemini analyzed the PDF and recreated it as an editable Google Slides deck, which could then be:
- Fed back into NotebookLM as a new source
- Further customized for specific audiences
- Shared and collaboratively edited
The Power of Grounding
Throughout this workflow, we emphasized a critical concept: grounding. When AI outputs are grounded in specific, verified sources (rather than pulling from general training data), the risk of hallucinations drops dramatically.
In NotebookLM specifically, any question asked in the chat interface, any slide deck generated, any infographic created—all are grounded exclusively in the sources you provide. This transforms the tool from a creative but unreliable assistant into a reliable analyst working only with your curated materials.
The verification gap—the hard work of checking sources—still exists in the initial research phase. But once you've curated trusted sources into NotebookLM, everything that follows maintains that foundation of truth.
A Second Example: Understanding Entrepreneurial Mindset
To further demonstrate the versatility of these workflows, we explored a second use case: developing our own understanding of the Entrepreneurial Mindset framework that underlies the AI Academy itself.
Starting with:
- A document summarizing materials from the KEEN Network's Engineering Unleashed platform
- Three YouTube videos explaining the entrepreneurial mindset framework
- Custom instructions for visual style
NotebookLM generated:
- Infographics visualizing the three pillars (Curiosity, Connections, Creating Value)
- A complete slide deck in a "professor's whiteboard" style
- Study guides and briefing documents
This demonstrated how the same workflow approach applies whether you're creating admissions materials, developing teaching resources, or simply trying to understand a new concept yourself.
The Invitation to Experiment
The session concluded with an invitation to all participants: try this workflow with your own topics and sources. See what happens when you:
- Use Perplexity to research a topic in your field
- Add your own verified documents to NotebookLM
- Experiment with custom instructions for different visual styles
- Chain the outputs through different tools
The goalposts are constantly moving with these technologies. Features change, new capabilities emerge, and what works best today may be different tomorrow. This is why the focus on workflows and mindset matters more than mastery of any particular tool.
What's Next
Over the coming months, our AI Academy cohort will continue exploring:
- Advanced prompt engineering techniques
- Building custom GPTs and AI assistants for specific courses
- Redesigning assessments for the AI age
- Integrating AI into research workflows
- Developing course-specific AI tools
But the foundation we established in Session 1 remains constant: AI literacy does not mean knowing which buttons to click in a given web app. It means understanding how to orchestrate different AI capabilities into workflows that serve your specific goals while maintaining academic rigor and integrity.
As we tell our students when teaching them to code: the syntax changes, the languages evolve, but the fundamental thinking skills—decomposition, pattern recognition, algorithmic thinking—those endure. The same principle applies here.
The GW Engineering AI Academy is a strategic initiative to position SEAS as an AI-forward institution through systematic faculty development, anchored in the Entrepreneurial Mindset framework of our KEEN partnership.