#4 AI Data & Infrastucture
Episode #4 Data Isn’t Boring - It’s Your AI Superpower
Welcome to The Conversation—Model Mind AI’s podcast about how artificial intelligence is reshaping the way we work, lead, and think. I’m Angela Schultz, curriculum developer and trainer here at Model Mind.
Today’s episode isn’t about spreadsheets or SQL—it’s about data literacy and how it can make or break your AI strategy.
To make it approachable, let’s start with a metaphor. Imagine walking into a doctor’s office and saying, “I don’t feel well.” No details. No symptoms. The doctor writes a prescription anyway. Would you trust it?
The same goes for AI. Without clear, structured, and trustworthy data, you’re asking AI to diagnose a problem in the dark.
That’s where data literacy—and good infrastructure—come in. To help unpack this, I’m joined by Model Mind CEO and resident data expert, Loren Horsager.
Key Takeaways:
1. Good Data Beats Big Data (For Most Use Cases)
Generative AI tools don’t need millions of rows—they need relevant examples.
2. Examples are Everything
Providing both “what to do” and “what not to do” helps GPTs perform more consistently—just like training a new employee.
3. Synthetic Data Can Help You Scale
You can use AI to generate additional examples when you're short on data.
4. Think Critically About Your Data
Know where your data came from, how messy it is, and how AI will interpret it.
5. Curiosity, Creativity, and Critical Thinking Win
These traits are more important than advanced math when working with data and AI.
Prompt of the Week: Storytelling with Data
Take a real data point and ask AI to explain it in the style of a movie genre (drama, comedy, horror…). Include: What happened? (Descriptive), Why did it happen? (Diagnostic), What will happen next? (Predictive), and What should we do about it? (Prescriptive)
Listen Here:
The Conversation with Model Mind AI YouTube
the Conversation with Model Mind AI | Podcast on Spotify
the Conversation with Model Mind AI Podcast on Amazon Music
the Conversation with Model Mind AI - Podcast - Apple Podcasts
Episode 4 – Data Literacy & AI: The Hidden Link to Success
In Episode 4 of *The Conversation*, Angela Schultz (Curriculum Developer and Trainer at Model Mind AI) is joined by CEO Loren Horsager to demystify one of the most overlooked and impactful topics in AI readiness: data literacy.
This episode explores why your AI efforts live or die based on the quality of the data you provide—not the size of the dataset, but how it’s structured, what context it carries, and whether it includes clear examples.
Good Data vs. Big Data
Loren explains that for most generative AI applications, volume isn't the priority—clarity is. Instead of feeding AI millions of rows of data, it’s more effective to provide examples, including both positive and negative cases, to help shape reliable and consistent outputs. When training custom models, large datasets matter. But for generative tools, accuracy and context reign supreme.
Real-World Messiness: Why Testing Matters
The team emphasizes that AI systems should be tested with production-quality data—complete with its inconsistencies, noise, and edge cases. Overly 'clean' test data sets up AI for failure when real-world conditions hit. Angela and Loren share stories of how prompt breakdowns often come from unexpected inputs that weren’t accounted for during development.
The Role of Examples in Prompt Engineering
AI responds best when it’s trained like a human: with examples. Loren recounts a legal contract automation project that broke down when a new type of document emerged. The fix? More examples. Clear rules. And a well-defined scope of what’s in and out of bounds. Angela reflects on how this applies to prompting too—especially when embedding parameters, constraints, and stylistic preferences.
Data as a Thinking Partner
Using data isn’t just about charts and summaries. Angela encourages teams to use data to drive storytelling, decision-making, and innovation. This ties into Jordan Morrow’s framework of the 'Three Cs' of data literacy: Curiosity, Creativity, and Critical Thinking—plus a fourth C: Collaboration. Together, these traits help teams ask better questions, explore safely, and use AI as a partner in discovery.
From Descriptive to Prescriptive
Angela shares a framework for evolving how teams use data:
- Descriptive: What happened?
- Diagnostic: Why did it happen?
- Predictive: What might happen next?
- Prescriptive: What should we do about it?
Loren reinforces that AI can be a great thinking partner through each of these layers, as long as it's paired with clean, thoughtful, example-driven inputs.
Prompt of the Week: Storytelling with Data
Angela’s creative prompt invites teams to analyze their data through the lens of storytelling. Ask AI to tell the story of a data point in the style of a movie genre—drama, comedy, horror, or documentary—and explore it across all four levels of analysis. It’s a fun and powerful way to build data confidence and insight.
Final Thoughts
Data is the backbone of any successful AI strategy—not just big data, but good, contextual, tested, and purposeful data. Episode 4 is a must-listen for anyone building workflows, prompts, or custom GPTs—and especially for leaders looking to foster a more data-literate, AI-ready team.