As we head into 2026, I’ve been thinking about what business owners actually need to know about AI. Not the hype. Not the fear-mongering. Just the practical fundamentals that help you make better decisions.
I spend a lot of time explaining AI to business leaders across the Midwest. Most of them know a little about what AI can do—they’ve seen the demos, tried ChatGPT, heard the success stories. But when I ask them how AI actually works or why it behaves the way it does, I usually get some version of “I have no idea.”
That knowledge gap matters more than you might think. Without understanding the fundamentals, businesses invest in AI tools but struggle to use them effectively. Teams produce inconsistent results. Leaders can’t figure out what went wrong or how to fix it. The problem isn’t the technology. The problem is that we’ve been treating AI like just another software package when it actually requires a different way of thinking about work.
So this year, I’m starting a series focused on AI fundamentals for business owners. No buzzwords. No promises that AI will revolutionize everything overnight. Just straightforward explanations of concepts that matter when you’re trying to use AI effectively in your business.
This first post covers something that changed how I work with clients. It’s called the 4D Framework for AI Fluency. Developed by Professor Rick Dakan from Ringling College of Art and Design and Professor Joseph Feller from University College Cork, this framework describes four core competency areas that can inform how people work with AI systems. I’ve been using it with my clients over the past year, and it’s given them a much better foundation for AI adoption.
Let me break down what this framework is, why it matters for your business, and how you can apply it across different areas of your operation.
What Is AI Fluency (And Why “Literacy” Isn’t Enough)
Most AI training focuses on tool literacy. How to write prompts. Which buttons to click. What settings to choose. That’s useful, but it’s not enough.
AI fluency requires not just tool-level literacy but a broader and deeper comprehension of AI-human collaboration. Think of it like the difference between knowing how to type and knowing how to write well. One is a mechanical skill. The other is about judgment, strategy, and understanding.
The 4D Framework helps you develop that deeper understanding. It gives you and your team a common language for talking about AI use, a way to evaluate whether you’re using AI effectively, and a structure for improving over time.
The Four Core Competencies: Delegation, Description, Discernment, and Diligence
The framework identifies four core competencies: Delegation (what work should be done by AI vs. humans), Description (can we clearly articulate what we want AI to do), Discernment (are we able to evaluate AI outputs with care), and Diligence (are we acting ethically and responsibly).
These aren’t steps in a process. They’re overlapping skills that you and your team need to develop. Let me explain each one.
Delegation: Deciding What Work Goes to AI
This is about figuring out which tasks AI should handle, which ones you should do yourself, and where collaboration makes sense.
Professor Feller stressed that understanding the nature of the technology is critical to making good decisions, especially in environments where hallucinations or bias can surface. You can’t delegate effectively if you don’t understand what AI is actually good at.
For example, AI excels at pattern recognition, drafting initial content, summarizing information, and handling repetitive tasks. It struggles with nuanced judgment calls, understanding your specific business context without detailed explanation, and tasks requiring up-to-date factual accuracy.
A manufacturing client recently asked me if they should use AI to write their safety procedures. My answer was no for delegation, but yes for augmentation. Don’t hand the whole task to AI. Instead, have your safety manager draft procedures based on their expertise, then use AI to help refine language, check for clarity, and suggest areas that might need more detail. The human expertise stays at the center.
Description: Communicating Clearly With AI
Once you’ve decided what to delegate, you need to describe it effectively. This is where most people think “prompt engineering,” but it’s bigger than that.
Description includes providing context about your business, explaining your goals and constraints, clarifying what success looks like, and giving examples when helpful. The better you describe, the better the AI performs.
I worked with a healthcare practice that was frustrated with AI-generated patient communication. Their prompts were too vague like “write an email reminding patients about appointments.” When we added description like tone (friendly but professional), key information to include (date, time, prep instructions), and constraints (under 150 words, reading level appropriate for diverse patient base), the outputs improved dramatically.
The description skill also involves iteration. You provide context, see what comes back, refine your description based on the results. It’s a conversation, not a one-shot command.
Discernment: Evaluating AI Outputs Critically
This might be the most important competency, and it’s the one people skip most often. Discernment is a real lived experience where workers become editors rather than just creators.
Discernment means checking for accuracy, evaluating whether the output actually solves your problem, watching for bias or inappropriate content, and considering what’s missing or incomplete.
I see businesses get in trouble when they treat AI output as automatically correct. A client in professional services used AI to draft a proposal and sent it without thorough review. The proposal referenced a service they didn’t actually offer and included a case study that sounded good but wasn’t factually accurate. They lost the deal and credibility.
Your team needs permission and training to be critical editors. That means time to review, authority to reject or heavily revise AI outputs, and clear standards for what constitutes acceptable quality.
Diligence: Using AI Responsibly and Ethically
Feller and Dakan suggest starting with Diligence, not Delegation, when embedding AI use responsibly. Before you rush to implement AI, pause and ask what your boundaries are.
Diligence covers questions like: What data should never go into AI systems (customer information, proprietary processes)? How will we ensure transparency when AI is used in customer-facing situations? What are our commitments around bias and fairness? How will we handle it when AI makes mistakes?
For my Midwest manufacturing and healthcare clients, diligence often includes HIPAA considerations, accessibility requirements under ADA and Section 508, and making sure AI use doesn’t create safety risks or compliance problems.
A distribution company I work with created a simple policy: any AI-generated content that goes to customers must be labeled as AI-assisted, and a human must review it for accuracy. That’s diligence in practice.
Three Modes of Working With AI
The framework also recognizes that we use AI in different ways. Understanding these modes helps you apply the 4Ds more effectively.
Automation: AI performs specific tasks based on your instructions. Think email summaries, data entry, basic coding. This is the simplest mode and what most people think of first.
Augmentation: You and AI collaborate as thinking partners. AI helps you brainstorm, refine ideas, explore alternatives. This is where a lot of business value lives but requires stronger skills in all 4Ds.
Agency: You configure AI to work independently on your behalf over time, establishing its knowledge and behavior patterns. This is more advanced and requires serious attention to delegation and diligence.
Most small to medium businesses I work with get the most value from automation and augmentation modes. Agency is powerful but needs careful governance.
Applying the 4D Framework to Your Business
Let me show you how this plays out in different business areas.
Sales: Smarter Prospect Research and Follow-Up
Your sales team probably spends hours researching prospects, drafting personalized emails, and following up on conversations.
Delegation: Let AI handle initial research (company background, recent news, industry trends). Keep human judgment for relationship decisions and customizing approach.
Description: Provide AI with your ideal customer profile, what information matters most for your sales process, and examples of good prospect research. Don’t just say “research this company.”
Discernment: Train your team to verify facts, especially financial information or recent changes. AI sometimes makes stuff up or pulls outdated data.
Diligence: Make sure your team isn’t feeding confidential deal information into public AI tools. Use AI to enhance personalization, not replace genuine relationship building.
A sales consultant I know in Cleveland uses AI to draft initial outreach emails but always rewrites the opening and closing in his own voice. The AI handles the middle section with relevant insights about the prospect’s industry. His response rates went up because emails are more informed, but they still sound like him.
Marketing: Content Creation at Scale With Quality Control
Marketing teams are drowning in content demands. AI can help, but I’ve seen too many businesses publish generic AI content that hurts more than it helps.
Delegation: Use AI for first drafts, headline variations, social media adaptations of longer content. Keep humans in charge of strategy, brand voice, and final messaging decisions.
Description: Give AI your brand guidelines, examples of past content that worked, specific audience insights. “Write a blog post about our service” is lazy description. “Write a 1200-word blog post for manufacturing owners in the Midwest who are skeptical of new technology but interested in practical cost savings” is better.
Discernment: Does this sound like your brand? Is it actually useful to your audience or just keyword stuffing? Does it make claims you can back up?
Diligence: Never publish AI content without human review. Consider disclosing when AI was used in content creation (as I do in my proposals when appropriate). Make sure you’re not accidentally infringing on others’ work.
A retail client creates content calendars with AI help but assigns each piece to a team member who adds specific examples, local references, and their own expertise. The AI speeds up the process, the human makes it valuable.
HR: Streamlining Hiring and Onboarding
HR departments in small to medium businesses are typically lean. AI can help with administrative burden but requires careful handling around bias and compliance.
Delegation: AI can help write job descriptions, screen resumes for basic qualifications, schedule interviews, draft onboarding documentation. Humans must make hiring decisions, handle sensitive employee situations, and customize onboarding to individual needs.
Description: Provide clear criteria for what you’re looking for. “Find good candidates” is useless. “Screen for at least 3 years of relevant manufacturing experience, technical certifications, and evidence of problem-solving in team environments” is specific.
Discernment: Watch for bias in AI-assisted resume screening. Are certain demographic groups being filtered out? Do the AI-generated interview questions actually assess what you need? Is the onboarding material actually helpful or just generic?
Diligence: This is huge in HR. You need to comply with employment law, protect candidate privacy, ensure accessibility in all materials, and document your processes. AI should support compliance, not create new risks.
An HR director I know uses AI to generate initial job description drafts but always reviews for biased language (like “rockstar” or “ninja” which can deter older workers or those who prefer straightforward language). She also has a human review every resume that AI flags for exclusion to catch false negatives.
Manufacturing: Quality Control and Process Optimization
Manufacturing might seem like an unlikely place for AI fluency, but the opportunities are significant.
Delegation: AI can analyze production data to spot quality issues, predict maintenance needs, optimize scheduling. Humans need to make decisions about process changes, handle equipment problems, and manage safety.
Description: Your AI needs to understand your specific processes, quality standards, and constraints. Generic manufacturing AI won’t know that your line has particular temperature sensitivities or that certain materials require extra handling.
Discernment: Is the AI catching real problems or just flagging noise? When AI suggests a process change, does it make practical sense on the shop floor? Are cost projections realistic?
Diligence: Safety is paramount. Any AI recommendations that touch safety procedures need extra scrutiny. Also consider how AI decisions affect workers—are you using AI to support your team or replace them?
A manufacturer in Youngstown uses AI to monitor equipment performance and predict failures. But they trained their maintenance team to question AI recommendations and add their hands-on expertise. The AI catches patterns in the data, the humans know the machines’ quirks. Together they’ve reduced downtime significantly.
Getting Started: Building AI Fluency in Your Team
You don’t need to become an AI expert overnight. Start with these steps.
Talk about it openly. Professor Feller said organizations need to fairly rapidly de-stigmatize the AI conversation, pick a framework, give people shared language, and make actual time for teams to talk about what they’re doing and learning with AI. Make it okay to experiment, ask questions, and admit confusion.
Create simple guidelines. You don’t need a 50-page policy. Start with basics like what data can’t go into AI tools, when human review is required, how to disclose AI use.
Practice the 4Ds on real work. Pick a common task your team does and walk through it together. What should we delegate? How would we describe this well? What would good discernment look like? What diligence considerations matter?
Share examples. When someone on your team uses AI effectively, have them show others. When something goes wrong, treat it as a learning opportunity, not a failure.
Start small and build. Don’t try to transform everything at once. Pick one process or department, apply the framework, learn from it, then expand.
Why This Matters for Midwest Businesses
Companies on the coasts have been faster to adopt AI, partly because they’re surrounded by tech culture. But that’s actually an advantage for us. We can learn from their mistakes.
I’ve watched businesses waste money on AI tools they don’t use well, damage customer relationships with obviously generic AI content, and create compliance headaches by not thinking through diligence.
The 4D Framework gives you a way to adopt AI thoughtfully. You can get the efficiency gains and competitive advantages without the unforced errors.
It also aligns with how Midwest businesses tend to operate. We value practicality over hype. We want clear ROI. We believe in taking responsibility for our work. All of that fits perfectly with the delegation, description, discernment, and diligence approach.
Where to Learn More
The good news is that Professors Dakan and Feller, working with Anthropic, created open access courses built around the 4D Framework that are freely available under a Creative Commons license. You and your team can take these courses at no cost.
The framework also has a dedicated website at aifluencyframework.org with additional resources and documentation.
If you want help applying this framework specifically to your business, that’s exactly the kind of work I do with clients across Northeast Ohio. Sometimes the best first step is talking through your specific situation and figuring out where AI actually makes sense for you.
The Bottom Line
As we start 2026, AI is becoming increasingly common in business operations. The question isn’t whether to use it, but how to use it well.
The 4D Framework helps organizations speak a common language about AI use, especially in times of change. It gives you and your team a structured way to think about AI adoption, avoid common pitfalls, and build real capability over time.
You don’t need to be a tech company to be good at this. You just need to be thoughtful, willing to learn, and committed to using AI as a tool that supports your people and your business goals.
This is the first post in my 2026 series on AI fundamentals for business owners. In the coming weeks, I’ll cover topics like how AI actually learns, why it makes mistakes, what questions to ask vendors, and how to calculate real ROI on AI investments. If there are specific AI topics you want me to cover, let me know.
If you’re thinking about AI for your business and want to talk through what good implementation looks like, I’m happy to help you think it through. The framework gives us a great starting point for that conversation.
About the Author: Jeff Rodgers owns Great Lakes AI Solutions, helping small and medium businesses across the Midwest implement practical AI solutions. Interested in learning more for your business? Let’s chat!
