Practically Speaking: A Look at How AI Can Practically Be Implemented in Our Everyday Work

Resources

At the first “Generation AI | Conversations and Perspectives on Artificial Intelligence” webinar (February 26, 2025), James Moore began with a provocative statement from OpenAI’s Sam Altman: “95% of what marketers use agencies, strategists and creative professionals for today will be easily, nearly instantly and almost no cost handled by AI.” For many, this represented both an enormous opportunity and a daunting challenge.

Understanding AI: From Magic to Methodology

Moore’s approach centers on what he calls being “AI literate” — understanding the provenance, pairing, and price of different AI systems (he uses the analogy of a sommelier for AI). He referenced John McCarthy, who coined the term “artificial intelligence,” noting that “as soon as it works, no one calls it AI anymore.” This insight captures a crucial point: once we understand how AI works, it stops seeming magical and becomes a tool we can strategically deploy.

For Chicago businesses grappling with where to start, Moore demonstrated this principle through a practical example. When asked to present to the Broadway League, an organization he knew little about at the time, he used Adobe Acrobat to download their entire public website into a 371-page PDF, then uploaded it to Google’s NotebookLM. The result? A chatbot fully versed in the all aspects of the Broadway League and a 30-minute AI-generated podcast that gave him a comprehensive understanding of the organization’s history, needs, and operations.

This example illustrates a fundamental insight from Moore’s presentation: the power of Retrieval Augmented Generation (RAG), a technique that dramatically reduces AI “hallucinations” by grounding the outputs of the Large Language Model (LLM). NotebookLM, which Moore recommends as a free starting point, combines large working memory with RAG capabilities, making it particularly valuable for business applications.

The Framework: Moving from Knowledge to Implementation

Perhaps the most valuable aspect of Moore’s presentation was his four-step framework for organizational AI adoption. This isn’t just about technology; it’s about change management and strategic thinking.

  • Step One: Develop an AI Policy. Moore emphasized keeping policies simple, short, and grounded in organizational mission and strategy. Most importantly, he stressed that perfection isn’t the goal initially — these policies will evolve as the organization learns.
  • Step Two: Educate Everyone. Here, Moore made a bold recommendation: every employee should complete Finland’s free AI courses, launched in 2018 when that country recognized AI as transformational for their entire workforce. These courses, completable in weekends, provide foundational understanding across three crucial areas: AI introduction, enterprise integration, and ethics.
  • Step Three: Create Learning Communities. This step reveals Moore’s deep understanding of organizational dynamics. He recommends cross-functional teams mixing different management levels and departments. The goal isn’t just to learn about AI, but to discover how it can assist with actual organizational tasks while identifying gaps and opportunities.
  • Step Four: Share and Revise. The framework becomes cyclical here, with communities sharing discoveries, documenting learnings, and continuously updating policies based on real-world experience.

The Security Imperative: Your “House AI”

One of Moore’s most practical recommendations addressed a concern raised by multiple audience members: security and privacy. His solution is elegantly simple — every organization needs to provide “house AI” to all employees. At DePaul, this currently means Microsoft Copilot Chat, available at no cost to students, staff, and faculty.

The reasoning is compelling: if organizations don’t provide secure AI tools, employees will use whatever’s available, potentially compromising sensitive information. Moore outlined a progression from basic secure access through APIs for sensitive data, cloud-contained instances, and ultimately local AI models running on organizational hardware.

This security-first approach becomes even more critical given Moore’s warnings about emerging risks. He highlighted the danger of “prompt injection,” where malicious websites can hijack AI agents, and the proliferation of always-listening devices from AI-enabled sunglasses to smartwatches that could inadvertently capture privileged conversations.

Environmental Awareness: Choosing the Right AI Tool

Moore’s concept of “environmental awareness” provides crucial guidance for organizations overwhelmed by AI options. He outlined how to evaluate the growing landscape of AI models, from the six major cloud-based frontier models in the US to emerging options from China and Europe.

His recommended evaluation framework considers multiple factors: basic performance through platforms like LMArena, safety considerations through holistic evaluation tools, capability testing through comprehensive assessments, and transparency about training data sources. This multifaceted approach helps organizations choose AI tools that align with their values and requirements, not just raw performance metrics.

The Human Element: Change Management and Adoption

Perhaps the most sophisticated aspect of Moore’s presentation was his discussion of change management, drawing from Kurt Lewin’s force field analysis from the 1940s. His insight is particularly relevant for AI adoption: resistance to change typically stems from perceived costs outweighing benefits. The solution isn’t to push harder, but to understand and address the restraining forces.

Moore’s prescription for successful AI policy implementation centers on clear communication about goals, acknowledging that change involves discomfort, listening to concerns, and finding advocates within the organization. This approach recognizes that AI adoption isn’t just a technical challenge — it’s fundamentally about people and organizational culture.

Looking Forward: Opportunities and Challenges

Moore concluded by examining near-term developments that will impact Chicago businesses. The emergence of reasoning models that spend more time analyzing complex queries is changing prompt engineering strategies. Agentic AI that can act independently on behalf of users offers tremendous potential but introduces new risks, particularly around prompt injection attacks.

The proliferation of AI-enabled devices — from Meta Ray-Ban’s AI sunglasses to always-recording wearables — presents both opportunities for enhanced productivity and significant privacy challenges for organizations. Moore’s message is clear: these technologies are arriving whether we’re ready or not, making proactive policy development and employee education crucial.

The DePaul Advantage: Partnership in Practice

Throughout his presentation, Moore emphasized DePaul’s role as a partner to the Chicago business community. With more alumni working in the Chicagoland area than perhaps any other university, DePaul’s AI Institute represents a unique resource for organizations navigating AI implementation.

This partnership approach reflects the university’s commitment to practical, inclusive innovation. Rather than treating AI as an abstract academic topic, the AI Institute focuses on real-world application, offering frameworks and resources that businesses can immediately implement.

Moore’s presentation ultimately delivered on its promise to move beyond theory to practical implementation. By providing specific tools, frameworks, and considerations, he offered the Chicago business community a pathway from AI’s apparent magic to methodical, strategic deployment.

For organizations ready to move beyond AI experimentation to systematic integration, Moore’s framework provides a thoughtful, security-conscious approach that acknowledges both the tremendous potential and inherent challenges of this transformational technology. In a landscape where change is the only constant, this kind of practical guidance becomes invaluable for businesses committed to staying ahead of the curve while maintaining their values and security standards.