You have seventeen browser tabs open right now. (I just counted mine, 29 for me 🤦♀️).
One of them is an article about AI tools you bookmarked three weeks ago. Maybe four. You meant to read it during your lunch break, but lunch was at your desk and you spent it answering a stakeholder Slack message that turned into a thread.
That tab is still there. You haven’t closed it because closing it would feel like giving up on something. But you also haven’t opened it, because when would you? Between the needs assessment that got dropped on you without warning, the compliance module due by Friday, and the stakeholder who needs a ‘quick update’ that will take 40 minutes to prepare, there is no uninterrupted hour. There probably won’t be next week either.
This is where a lot of in-house L&D professionals are with AI right now. Not resistant. Not afraid. Just… behind the moment they were waiting for.
Here’s what I want you to know: the moment doesn’t come on its own. And the thing in the way isn’t time. Not really.
The real barrier isn’t your schedule
Those who haven’t started yet aren’t lacking motivation. They’re not technophobic. They’re not even particularly behind.
They’re uncertain. And uncertainty is a surprisingly effective blocker, especially for people who are already operating at capacity.
Because here’s what starting feels like when you’re stretched thin: a risk. What if you spend an hour trying something and it doesn’t work? What if the output is bad? What if you build a dependency on a tool and then can’t afford it? What if you start using AI and your work gets worse before it gets better?
Those aren’t irrational fears. They’re the fears of someone who doesn’t have bandwidth to absorb a learning curve right now. And the instinct to wait for a better moment, more time, more certainty, is completely understandable.
It’s also exactly what keeps you stuck.
Let me be honest with you: The better moment is not coming. The free hour is not coming. The conditions you’re waiting for are not how this starts. You pick a low-stakes task, you try it, and you learn from what happens. That’s the whole on-ramp.
What ‘low stakes’ actually means
I’m not asking you to redesign your entire workflow. I’m not asking you to overhaul your process or commit to AI as a philosophy. I’m asking you to find one task in your current week that fits two criteria:
- It takes more time than it probably should.
- It doesn’t require your best thinking. It’s mostly just output.
A few things that fit that description for most in-house designers:
- Drafting a first round of SME interview questions before a kickoff.
- Writing a course overview paragraph (you know, the one you rewrite three times because you hate all the versions).
- Turning a wall of compliance text into something resembling a learning objective.
- Building out a list of ways your learners might misunderstand the key concept in your current module.
- Writing the facilitator notes for a section you’ve already designed.
None of these requires you to trust AI with the parts of your work that actually require your expertise. They just let you spend less time on the parts that don’t.
The thing nobody tells you about the first try
It probably won’t be great. Let me say that louder: the first output you get will likely need editing. Possibly a lot of editing. It may miss your org culture entirely. It may give you something technically correct that sounds nothing like you.
That’s normal. That’s actually useful.
Because the gap between what AI gives you and what you actually need, that gap is information. It’s the place where your expertise lives. Every time you edit an AI output, you’re clarifying what ‘good’ looks like in your specific context. You’re building a working relationship with a tool that gets better the more specific you are.
The first try is not supposed to be the final output. It’s supposed to be the first draft of a process you’ll keep refining.
Pro tip: The most useful thing you can do after your first AI output lands flat? Ask yourself why it missed. Was the prompt too vague? Did you give it context it couldn’t have? Did it default to generic when you needed specific? Those answers are your prompt for next time.
Where to start if you’re starting this week
- Pick the task on your list right now that you’ve been putting off because it’s tedious. Not the strategic, high-stakes work. The tedious stuff. The stuff you’d describe to someone else as ‘it’s not hard, I just hate it.’
- Open your AI tool of choice.
- Describe the task.
- Give it context.
- Tell it who your learners are, what the performance gap is, what constraints you’re working with.
- See what comes back.
- Then edit it. Because you will need to. And notice what you had to change, because that’s where you start to get good at this.
That’s it. That’s the whole entry point.
You don’t need an uninterrupted hour. You need one task and a willingness to see what happens.
One more thing
If you’ve been watching from the sidelines while the field moves and quietly wondering whether you’re falling behind, I want to be clear: you’re not. You’re at exactly the stage most in-house L&D professionals are at right now. And there’s a version of this that doesn’t require you to rebuild your workflow, compromise your standards, or become an early adopter.
It just requires you to close one of those seventeen tabs and try.
Inside #IgniteLearning this month, the focus is on exactly this: AI as a real workflow tool, built around the work in-house L&D professionals actually do. Free 7-day VIP trial. https://zps.circle.so/checkout/ignite-learning