A few weeks ago, I hosted a webinar called “How to AI: Treating AI Like a Teammate, Not a Magic Wand.”
And that title is intentional: AI is not a magic wand. It’s a teammate.
That framing has been incredibly helpful for me, because it moves us out of hype and into something much more practical. It helps us think in terms of how we actually work.
In the webinar, we walked through what this means operationally. Not at the 30,000-foot level. Not at the ultra-technical level. But somewhere in the middle—where most of us live.
If you’re trying to figure out how to use generative AI well (and responsibly) in your education program, here are four keys that have made the biggest difference for me.
Key #1: Shift Your Mindset Before You Touch the Tool
Most AI frustration starts with the wrong expectation.
Too often, we open ChatGPT (or Gemini or Claude) and type something broad like “Create a course on _____.” What we get isn’t bad… but it’s not good, either.
But if we think about onboarding AI like a new member of our team, we have a different experience. When we hire someone new, we don’t give them a directive to start (i.e., “create a course on ____.”) Instead, we give them a variety of inputs to understand our company, our business, our products, and how we create content.
We need to treat AI the same way. Give it onboarding content the way you would a new hire. Provide examples of what “great” looks like and examples of past work. Offer feedback on first drafts and why we’d recommend those changes.
When we treat Generative AI like a tool, we get mediocre results. When we train it like a teammate, we get much better results.
(Check out our Ultimate Guide to GenAI for L&D for more tips!)
Key #2: Context Is the Multiplier
We hire learning experience designers or instructional designers because they’re good at what they do. But when we onboard them, we provide them with something they don’t have: context.
GenAI is programmed with lots of frameworks, methodologies, and best practices—but all of that is useless if it doesn’t have the right context in which to apply those tools.
GenAI doesn’t know:
- Your audience maturity
- Your business priorities
- Your executive pressures
- Your brand voice
- Your internal politics
- Your past failures
You have to provide that context to get the best result possible.
So what does “good context” look like?
Here are some things that would be beneficial to provide when training GenAI to support your education programs:
- Clear audience definitions (Who is this for? What do they already know?)
- Specific learning objectives (What should change after this?)
- Constraints (Time, format, modality, budget)
- Brand voice guidance
- Examples of previous strong work
- Your unique point of view
The difference in output is dramatic.
This isn’t about prompt engineering tricks. It’s about clarity of thinking. AI surfaces whatever structure you give it. If your inputs are vague, your outputs will be too.
Key #3: Always Assume It’s a First Draft
This is one of the most important guardrails.
AI can hallucinate. It can confidently state something that isn’t true. It can create assessments that appear aligned, but subtly miss your learning objectives. Even with specific instructions, it can adopt a tone that sounds more collegiate than perhaps your company’s voice is.
When I review AI-generated content, I look at five things:
- Accuracy – Are the claims correct? Are citations real?
- Alignment – Does the content actually support the learning objectives?
- Tone and level – Is this written for the right audience at the right reading level?
- Inclusivity and bias – Are examples representative? Is the language appropriate?
- Clarity and usability – Can someone actually act on this?
We are still accountable for what we publish. AI accelerates production, but it does not remove responsibility.
One of the best practices we discussed in the webinar is building a simple human-in-the-loop workflow: define the goal, generate the draft, review intentionally, refine, and only then move forward.
That rhythm keeps speed and quality in balance.
Key #4: Use AI as an Accelerator, Not a Decision-Maker
Generative AI is exceptional at certain things.
It’s fantastic at:
- Drafting content quickly
- Summarizing long material
- Rewriting for tone or difficulty level
- Generating variations
- Helping you think through angles
It is not exceptional at:
- Strategic prioritization
- Organizational nuance
- Ethical tradeoffs
- Navigating cross-functional politics
It doesn’t know your stakeholders. It doesn’t sit in your leadership meetings. It doesn’t understand the emotional landscape of your organization.
That’s your role.
And here’s the encouraging part: AI doesn’t make our jobs obsolete. It makes them more strategic.
If drafting a module takes 30 minutes instead of three days, what do we do with that reclaimed time? We spend it on the things we need to do that we don’t always have time for: cross-functional meetings, reporting, strategy, etc.
From Tool to Teammate
The teams seeing real value from generative AI aren’t just experimenting. They’re operationalizing.
They document what good prompts look like. They define review standards. They maintain human accountability. They clarify where AI is appropriate, and where it isn’t (let’s face it, friends, AI is not the right solution for every problem!).
In other words, they’ve stopped treating AI like a shortcut and started treating it like a teammate.
If you’d like to go deeper into this framework, we unpacked it in detail during our webinar, “How to AI: Treating AI Like a Teammate, Not a Magic Wand.” You can watch the full recording and hear how we think about predictive AI, data foundations, and operational readiness as well.
.png)
