The real opportunity isn’t creating more learning in less time. It’s finding out what works sooner.
There is a warning about AI that I hear often in Learning and Development.
AI is fast, but be careful.
AI can create bad content.
We still need instructional designers.
Humans need to protect quality.
I agree with much of it.
Give AI a weak prompt and it can produce a polished-looking course in minutes. Slides, scenarios, quizzes, narration and job aids can all appear before anyone has stopped to ask whether training is even the right solution.
That is a real risk.
But I think we’re spending so much time talking about the risk of speed that we may be missing the opportunity.
What if speed is actually one of AI’s greatest gifts to L&D, provided we use that speed in the right place?
The problem isn’t speed.
It’s moving quickly in the wrong direction.
We’ve Been Slow for a Reason
Traditional learning development takes time.
We meet with stakeholders.
Interview subject matter experts.
Gather content.
Design.
Develop.
Review.
Revise.
Review again.
Pilot.
Revise again.
Eventually, we launch.
There is nothing inherently wrong with that process. Some learning solutions need that level of care, particularly when safety, regulation or significant business risk is involved.
But anyone who has worked in L&D long enough has probably experienced the other side of it.
You spend three months developing something, only to discover after launch that learners needed something different.
Or the business changed while you were building it.
Or the process changed.
Or managers don’t reinforce it.
Or employees understand the material perfectly well, but something else is preventing them from performing.
We sometimes spend months trying to perfect something before the people who actually need it have had a meaningful chance to use it.
That’s where I think AI gives us a different opportunity.
What If We Used Speed to Learn?
Software teams learned this lesson years ago.
They don’t always wait until a product is perfect before letting real users interact with it.
They build something useful.
People use it.
The team watches what happens.
Something doesn’t work.
They fix it.
Something customers thought they wanted turns out not to matter.
They change it.
Something nobody expected becomes incredibly valuable.
They build on it.
The product gets better because reality starts informing the design.
Why shouldn’t L&D work more like that?
Instead of spending twelve weeks debating what a learning solution should look like, perhaps AI helps us create an early version in a few days.
Put it in front of ten employees.
Watch them use it.
Where do they struggle?
What do they skip?
What questions do they still have?
Can they actually perform the task afterward?
What did we misunderstand?
Then improve it.
AI shouldn’t just help L&D create faster. It should help L&D learn faster.
That is a very different use of speed.
Faster Content Isn’t the Prize
This is where I think we need to be careful about how we measure AI productivity.
Suppose an instructional designer previously needed five hours to create a first draft and AI reduces that to one hour.
What should we do with the four hours we just gained?
The obvious answer is to create more.
More courses.
More videos.
More job aids.
More content.
I’m not convinced that’s the best use of the opportunity.
What if some of that time went somewhere else?
Talk to employees.
Observe the work.
Sit with a manager.
Look at the performance data.
Test the first version.
Ask someone to use the job aid while actually doing the task.
Find out what is getting in their way.
Then use what you learn to make the solution better.
The biggest opportunity AI gives L&D isn’t faster content. It’s faster feedback.
That changes the conversation from how much we can produce to how quickly we can find out whether what we’re producing actually helps.
This Makes Instructional Designers More Valuable, Not Less
There is understandable concern about what AI means for instructional design.
If AI can generate objectives, scenarios, assessments, storyboards, graphics, videos and first drafts, what happens to the instructional designer?
I think the answer depends partly on how we define the job.
If most of our value comes from producing learning content, AI clearly changes the economics of that work.
But producing content was never the highest-value part of instructional design.
Understanding people is.
Diagnosing performance problems.
Knowing when training will help and when it won’t.
Recognizing what matters and what doesn’t.
Working with stakeholders who don’t always agree.
Understanding the environment in which people actually perform.
Testing whether something works.
Making trade-offs.
Using judgment.
Those are much harder problems.
Let AI take more of the production work so instructional designers can spend more time doing the work that requires judgment.
That’s not defending instructional design from AI.
It’s using AI to move instructional design toward higher-value work.
Speed and Quality Are Not Opposites
This may be the assumption we need to challenge most.
We often talk about speed and quality as if increasing one automatically reduces the other.
Sometimes it does.
If we simply ask AI to generate a course and publish whatever comes back, we’ve gained speed and probably increased risk.
But there is another way to use speed.
Create faster.
Test sooner.
Get feedback earlier.
Improve.
Test again.
Now speed is helping quality.
The sooner we discover that something isn’t working, the sooner we can fix it.
Waiting another eight weeks doesn’t necessarily make our assumptions more accurate.
Real feedback might.
That is why I don’t think the right question for L&D leaders is:
“How much faster can AI help us build training?”
The more useful question is:
“How do I help my organization build capability fast enough to keep up with the business without sacrificing quality?”
That is a different challenge.
And it requires more than faster course creation.
It requires faster diagnosis, faster testing, faster feedback and faster improvement.
AI Can Get Us to the Wrong Place Faster
The critics are right about one thing.
AI can accelerate bad decisions.
If we misunderstand the performance problem, AI can help us build the wrong solution remarkably quickly.
If we haven’t spoken to learners, AI can make assumptions about them faster.
If training isn’t the answer, AI can produce an impressive training solution that still doesn’t solve the problem.
That risk shouldn’t be dismissed.
But neither should it become the reason we slow everything down.
AI can absolutely help us create bad training faster. But stopping there misses the bigger opportunity. It can also help us find out what’s working faster.
The difference is what we choose to accelerate.
Don’t accelerate assumptions.
Don’t accelerate unnecessary content.
Don’t accelerate a solution nobody has tested.
Accelerate the parts of the process that help us get closer to reality.
Build something useful.
Put it in front of people.
See what happens.
Learn.
Improve.
Then do it again.
That is how AI can help L&D become faster without becoming careless.
And perhaps that is the opportunity we should be talking about more.
Not how AI allows us to produce more learning.
But how it allows us to spend less time guessing what will work and more time finding out what actually does.
Don’t use AI just to create faster. Use it to find out what works faster.
Reflection
The next time AI saves your team several hours of development time, don’t immediately ask:
“What else can we create?”
Ask:
“What can we now learn sooner?”
That may be where the real productivity gain is hiding.
