AI Can Create Training. But Is It Any Good?

Transcript

[00:00] Is AI going to replace learning and
[00:02] development professionals? Well,
[00:04] according to some people, absolutely.
[00:06] Apparently, we can all pack up our
[00:08] laptops, cancel our subscriptions to
[00:10] various L&D tools, and find ourselves a
[00:13] new career.
[00:14] But, here’s the thing. AI can already
[00:17] write training materials, generate
[00:19] quizzes, suggest learning activities,
[00:22] and create presentations, as well as
[00:24] help you design entire training program.
[00:26] So,
[00:27] if it can do all of that, what’s left
[00:30] for us to do?
[00:31] Quite a lot, actually. Hello and welcome
[00:34] to the L&D Academy. My name is Irina,
[00:36] and today we’re talking about where AI
[00:39] actually helps in learning and
[00:40] development, and where it doesn’t. The
[00:43] inspiration for this video came from a
[00:45] comment someone recently left on one of
[00:47] my YouTube videos. They suggested that
[00:50] L&D is a profession that can and
[00:52] eventually will be replaced by
[00:53] artificial intelligence. And this isn’t
[00:56] the first time I’ve heard that argument.
[00:58] So, I thought it was worth exploring
[01:01] because in some ways, I agree. There are
[01:04] parts of our work that AI can already do
[01:07] remarkably well. And there are certainly
[01:10] tasks that I wouldn’t dream of doing
[01:12] manually anymore.
[01:14] But, there is a big difference between
[01:15] automating individual tasks and
[01:18] replacing an entire profession. And
[01:20] understanding that difference is
[01:22] particularly important if you’re just
[01:24] starting your career in L&D. So, let’s
[01:27] look at what AI is actually good at,
[01:30] where things can go wrong, and how you
[01:32] can start using it more effectively in
[01:34] your everyday work.
[01:38] >> [music]
[01:39] >> Before we get into specific examples,
[01:42] there’s something I want to make
[01:43] absolutely clear. AI is a tool. It isn’t
[01:47] the job itself. And I think that this
[01:49] distinction gets lost in a lot of
[01:52] conversations about AI replacing
[01:54] different professions. Generative AI is
[01:57] a software that produces content in
[01:59] response to instructions or prompts. You
[02:02] give it a prompt and it generates
[02:04] something based on that prompt. It might
[02:07] be a piece of text, an image, a
[02:09] presentation, or something else
[02:11] entirely.
[02:12] It’s enthusiastic, it’s available around
[02:14] the clock, and it can produce an
[02:16] impressive amount of work in a very
[02:19] short time. Unfortunately, it’s also
[02:22] occasionally far too confident about
[02:24] things it hasn’t actually checked. And
[02:27] that’s where we come in. Because speed
[02:29] is only valuable when you know what
[02:31] you’re trying to achieve. Let me give
[02:33] you an example. Imagine someone asks you
[02:36] to design a leadership development
[02:37] program. You open ChatGPT and type,
[02:40] “Create a leadership development program
[02:41] for new managers.” Within seconds, you
[02:44] have a beautiful outline. There are
[02:46] modules on communication, delegation,
[02:49] feedback, motivation, conflict
[02:50] management. It looks fantastic.
[02:54] But here’s my question. How do you know
[02:56] it’s the right program?
[02:58] Do these managers actually struggle with
[03:00] delegation? Do they need help giving
[03:03] feedback? Are they managing remote
[03:06] teams? Are they responsible for
[03:08] performance reviews? Are they working in
[03:10] manufacturing, retail, or financial
[03:12] services?
[03:14] What are the actual performance problems
[03:16] you’re trying to solve? AI doesn’t know
[03:19] any of that
[03:21] unless you tell it. And if you don’t
[03:22] know the answers yourself, you’re
[03:24] essentially asking a tool to make
[03:27] decisions you haven’t yet learned how to
[03:30] make. That’s not particularly helpful.
[03:33] So, before you open your favorite AI
[03:35] tool, I want you to try something.
[03:38] Finish the sentence, “I need help with
[03:40] blank so that blank.” For example, “I
[03:44] need three practice scenarios so new
[03:46] supervisors can rehearse asking
[03:48] clarifying questions during shift
[03:50] handovers.”
[03:52] That’s a much better starting point than
[03:53] create some communication training
[03:55] activities.
[03:56] You know what you’re trying to achieve
[03:58] and you can judge whether the output is
[04:00] useful. And that I think is one of the
[04:03] most important skills L&D professionals
[04:05] need to develop when working with AI.
[04:09] >> [music]
[04:11] >> Now, let’s get practical. There are
[04:14] plenty of ways to use AI in L&D and I
[04:17] want to focus on five things I think are
[04:20] particularly useful in everyday work.
[04:24] And for each one, I’ll give you an
[04:25] example of how you might use it. Number
[04:28] one, preparing for a learning needs
[04:30] analysis conversation. If you’ve watched
[04:32] some of my other videos, you’ll know
[04:34] that I talk about learning needs
[04:35] analysis quite a lot and for a good
[04:38] reason. Before you can design an
[04:40] effective learning intervention, you
[04:42] need to understand what people actually
[04:44] need. One of the most important parts of
[04:48] that process is asking the right
[04:50] questions. And this is something AI can
[04:53] help you prepare for.
[04:54] Imagine you’re meeting with a warehouse
[04:56] manager who’s concerned about picking
[04:58] errors. You know the general topic but
[05:01] you don’t know much about the specific
[05:03] processes or working conditions. You
[05:06] could ask AI, “Draft six open-ended
[05:09] questions for a learning needs analysis
[05:11] conversation with a warehouse manager
[05:13] about picking errors. Include questions
[05:16] about expected performance, current
[05:19] errors, tools, and working conditions.
[05:22] Do not assume training is the solution.”
[05:25] And there you have it. Six questions you
[05:27] can review, adapt, and use as a starting
[05:29] point. Now, does that mean that AI has
[05:32] conducted your learning needs analysis?
[05:34] Of course not. You still need to have
[05:36] the conversation, listen to the answers,
[05:39] ask follow-up questions, and identify
[05:41] what’s actually causing the problem. But
[05:44] you’ve saved yourself some preparation
[05:46] time and perhaps more importantly, you
[05:48] can go into the meeting feeling a little
[05:50] more confident. Actually, if you’re
[05:52] interested in this particular topic, I
[05:54] have another video where I walk you
[05:56] through exactly what to do during a
[05:57] learning needs analysis conversation.
[05:59] I’ll link it here and in the description
[06:01] down below. And if you want some
[06:03] ready-made questions to work with, we
[06:05] also have a learning needs analysis
[06:07] question bank available on the L&D
[06:10] Academy website.
[06:11] Number two, organizing feedback. Let’s
[06:14] say you’ve just delivered a training
[06:15] program to 150 employees. You’ve
[06:18] collected evaluation forms, and now
[06:22] you’re sitting there looking at dozens
[06:24] of open-ended comments. Some people love
[06:26] the activities. Others thought that the
[06:28] session was too long. A few mentioned
[06:31] that the examples were irrelevant to
[06:33] their jobs. And one person has written
[06:35] an entire essay about the temperature in
[06:38] the training room. We’ve all been there.
[06:40] Now, you could read through every
[06:42] comment, identify reoccurring themes,
[06:45] and manually organize everything into
[06:47] categories. Or, provided you’re using an
[06:49] approved tool and are authorized to
[06:52] process the comments this way, you could
[06:54] ask AI to help. For example, group these
[06:57] comments by reoccurring issue. Include a
[07:00] supporting comment for each theme and
[07:02] keep conflicting views visible.
[07:04] Within seconds, you’ll have a structured
[07:07] summary. You might see themes such as
[07:09] content relevance, delivery pace,
[07:11] practical application, and learning
[07:13] environment.
[07:15] That’s a considerable amount of time
[07:17] saved. And
[07:19] here’s where you need to be careful. AI
[07:21] is very good at making messy information
[07:24] look tidy. And sometimes, in the process
[07:27] of making things tidy, it can hide
[07:29] important details. For example,
[07:32] imagine 20 participants say the training
[07:34] was excellent, but one person points out
[07:37] that a particular procedure taught
[07:38] during the session contradicts company
[07:41] policy. That one comment might be far
[07:44] more important than the other 20
[07:46] combined. So, don’t just accept the
[07:48] summary. Go back to the original
[07:50] comments and check whether anything
[07:52] important has been overlooked. Number
[07:55] three, brainstorming learning
[07:56] activities.
[07:58] This is probably one of my favorite uses
[08:00] of AI. As L&D professionals, we’re often
[08:03] expected to design learning experiences
[08:05] for topics we’re not necessarily experts
[08:08] in. And while we should understand how
[08:10] people and how to design effective
[08:12] activities, we can’t possibly be experts
[08:15] in every subject. That’s where AI can be
[08:19] a useful brainstorming partner.
[08:21] Let’s say that you’re designing a
[08:22] workshop for retail supervisors who need
[08:25] to improve their delegation skills. You
[08:27] could ask, “Suggest three short
[08:29] exercises where a retail supervisor must
[08:31] explain a task, agree on a deadline, and
[08:34] check understanding. Use realistic
[08:37] everyday retail situations.” AI might
[08:40] suggest a role play involving a busy
[08:42] store, a scenario about preparing for a
[08:44] stock delivery, or a paired exercise
[08:47] where participants practice giving
[08:49] instructions. Great. You’ve got some
[08:51] ideas to work with. But now comes the
[08:54] important part. You need to decide which
[08:57] of those activities actually makes
[08:59] sense.
[09:01] Does it fit your learning objectives? Is
[09:03] it appropriate for your audience? Do you
[09:05] have enough time? Will participants have
[09:08] an opportunity to practice the skills
[09:10] they’re supposed to develop? And how
[09:12] will you know whether they’ve performed
[09:14] the tasks correctly? Because here’s
[09:17] something I really want you to remember.
[09:19] A list of interesting activities is not
[09:22] the same thing as a well-designed
[09:24] learning experience. You can have the
[09:27] most entertaining workshop in the world,
[09:30] but if participants don’t develop the
[09:31] skills they need, what’s the point?
[09:35] Number four, creating first drafts. This
[09:37] is another area where AI can save you a
[09:40] tremendous amount of time. Imagine you
[09:43] need to create a job aid. And if you’re
[09:45] new to L&D, a job aid is essentially a
[09:47] resource people can consult while
[09:49] performing a task. It might be a
[09:51] checklist, a quick reference guide, or a
[09:53] step-by-step instruction sheet. And
[09:56] let’s say that you need to create a
[09:57] checklist for employees handling
[09:59] returned items in a retail store. You
[10:01] already have the approved company
[10:03] procedure. You could give that procedure
[10:05] to an approved AI tool and ask, “Draft a
[10:07] one-page checklist for handling returned
[10:10] items using only the approved process
[10:13] below. Flag any missing information
[10:16] instead of inventing steps.” And within
[10:19] seconds, you have a first draft.
[10:22] Is it ready to be distributed to
[10:24] employees? Absolutely not. You still
[10:26] need to check it against the original
[10:28] procedure, make sure nothing important
[10:30] is missing, and maybe even have the
[10:32] relevant process owner review it. But,
[10:35] you’ve skipped the part you’re staring
[10:37] at a blank page trying to figure out how
[10:40] to structure the information. That’s a
[10:42] huge advantage. I actually think this is
[10:45] one of the best ways to use AI. Let it
[10:48] help you get to a first draft quickly,
[10:51] and then spend your time making that
[10:53] draft genuinely useful because there is
[10:56] nothing more frustrating than spending
[10:58] hours perfecting a document only to send
[11:02] it to a stakeholder who immediately
[11:04] decides to change half of it.
[11:06] Number five, adapting existing material.
[11:09] Finally, AI is particularly good at
[11:12] helping you adapt content you already
[11:14] have.
[11:15] Let’s say you’ve been given a technical
[11:17] explanation of a stock control process.
[11:20] It’s accurate, but it’s full of jargon,
[11:23] complicated terminology, and very long
[11:25] sentences. And your audience consists of
[11:28] employees who are completely new to
[11:30] stock control. You could ask AI rewrite
[11:34] this approved explanation for employees
[11:36] who are new to stock control, define
[11:38] technical terms, keep every procedural
[11:40] requirement, and list the changes
[11:43] separately. Now you have a simplified
[11:45] version that you can compare with the
[11:47] original. You can review the language,
[11:49] check that the meaning hasn’t changed,
[11:51] and decide whether it’s more appropriate
[11:54] for your audience. You can do something
[11:56] similar when adapting materials for
[11:58] different reading levels, creating
[11:59] summaries, or turning lengthy documents
[12:02] into quick reference resources.
[12:04] But again, there is a catch. Sometimes
[12:07] when AI simplifies something, it removes
[12:10] details that are actually essential. And
[12:13] in some industries, those details can be
[12:16] pretty important. Imagine simplifying a
[12:19] safety procedure and accidentally
[12:21] removing a condition that determines
[12:23] when a piece of equipment can be
[12:25] operated. Not ideal. So, by all means,
[12:29] use AI to adapt your materials. Just
[12:32] make sure someone with the appropriate
[12:34] expertise checks the final version.
[12:38] >> [music]
[12:41] >> Now, so far we’ve looked at five ways AI
[12:43] can make your work easier. And you might
[12:45] be thinking, “Oh, that’s great. Why
[12:48] wouldn’t I use it for everything?”
[12:50] Because there are some things AI simply
[12:52] doesn’t know. And some things it can get
[12:55] spectacularly
[12:57] wrong. The first problem is context. AI
[13:01] doesn’t see your workplace. It doesn’t
[13:03] know your employees, your organizational
[13:05] culture, your processes, or the
[13:07] challenges people face every day.
[13:10] Unless of course you provide that
[13:11] information, and even then it only knows
[13:15] what you’ve told it. So, let’s go back
[13:17] to our warehouse example. Imagine you
[13:19] tell AI that employees are struggling to
[13:21] communicate during shift handovers. It
[13:24] might suggest communication skills
[13:26] training.
[13:27] Sounds reasonable, right? But what if
[13:29] the real problem is that the two shifts
[13:32] never overlap?
[13:34] Employees aren’t failing to communicate
[13:37] because they lack communication skills.
[13:39] They’re failing to communicate because
[13:41] the organization hasn’t given them an
[13:43] effective way to exchange information.
[13:47] No amount of communication training is
[13:49] going to fix that. And that’s why
[13:51] learning needs analysis is so important.
[13:54] You need to understand the problem
[13:56] before deciding how to solve it.
[13:58] The second problem is accuracy.
[14:01] You’ve probably heard the term AI
[14:03] hallucinations. Essentially, this is
[14:05] when AI produces information that sounds
[14:08] perfectly plausible
[14:10] but isn’t actually correct. And the
[14:13] particularly annoying thing is that it
[14:15] can sound extremely confident while
[14:18] doing so.
[14:20] It might even invent a research study,
[14:22] misrepresent a theory, or confidently
[14:25] explain a company policy that doesn’t
[14:27] exist. So, whenever you’re using AI to
[14:30] generate factual content,
[14:32] check it.
[14:33] Double check triple check it. Look at
[14:36] the original sources. Consult the
[14:38] relevant subject matter experts. Compare
[14:40] the information against approved
[14:42] procedures and policies. And please,
[14:44] don’t just ask the same AI tool, “Are
[14:47] you sure?” Because that’s not
[14:49] independent verification. You’re
[14:51] essentially asking the same person who
[14:53] gave you the information to confirm that
[14:55] the information is correct.
[14:57] The third problem is data privacy.
[15:01] This one is particularly important when
[15:03] you’re working with employee
[15:04] information. As L&D professionals, we
[15:07] often have access to sensitive data.
[15:09] Performance reviews, employee feedback,
[15:11] assessment results, development plans,
[15:13] and so on.
[15:14] You can’t simply copy and paste all of
[15:17] that into whichever AI tool happens to
[15:20] be open in your browser. You need to
[15:22] follow your organizations data
[15:24] protection policies and use approved
[15:27] tools. And remember that removing
[15:30] someone’s name doesn’t necessarily make
[15:32] the information anonymous. If a comment
[15:34] mentions a particular job title,
[15:37] location, an unusual incident, it might
[15:40] still be very easy to identify that
[15:42] person. So, be careful about what
[15:45] information you’re sharing and where
[15:47] it’s going.
[15:49] And the fourth problem is learning
[15:51] design.
[15:52] This is the one I think matters the most
[15:56] for our profession. AI can suggest
[15:58] learning activities. It can write
[16:00] scenarios, it can generate quizzes, but
[16:03] it doesn’t automatically know whether
[16:06] those things will work for your
[16:08] particular learners.
[16:10] Imagine AI suggests a fantastic
[16:12] role-play activity. It looks great on
[16:14] paper, but you’re delivering the session
[16:16] to a group of employees who are
[16:18] uncomfortable speaking in front of
[16:19] others. Or perhaps some participants
[16:22] need additional language support. Or
[16:24] maybe the training room is so small that
[16:27] everyone is practically sitting on top
[16:29] of each other. I’ve delivered training
[16:31] in rooms where moving a chair felt like
[16:33] a major logistical operation.
[16:36] A brilliant activity on paper doesn’t
[16:39] necessarily translate into a brilliant
[16:41] learning experience. You need to
[16:43] consider the audience, the environment,
[16:45] the learning objectives, and how
[16:47] everything fits together. Not to mention
[16:49] how skilled you are as a facilitator.
[16:52] And that’s where your knowledge of
[16:54] learning design becomes essential.
[16:57] Now, before we continue, I’m curious
[16:59] about something. What’s your favorite
[17:01] way to use AI in your L&D work? Do you
[17:04] use it to brainstorm activities, draft
[17:06] materials, analyze feedback, or maybe
[17:08] something completely different? Let me
[17:10] know down in the comments because I’m
[17:12] always interested in discovering new
[17:14] practical applications.
[17:15] And if you’re finding this video useful,
[17:17] give it a thumbs up and consider
[17:19] subscribing to the L&D Academy. I
[17:21] regularly share practical advice, tools,
[17:23] and resources to help aspiring and
[17:26] beginner L&D professionals develop their
[17:28] skills.
[17:29] And if you know someone who’s worried
[17:31] that AI is about to steal their L&D job,
[17:34] perhaps share this video with them.
[17:37] Right, let’s talk about how to make sure
[17:39] the content AI produces is actually
[17:42] useful.
[17:46] One of the biggest mistakes you can make
[17:47] when working with AI is assuming that
[17:49] the first result is good enough. I mean,
[17:52] sometimes it is, but quite often it
[17:55] isn’t. So, I’d recommend following a
[17:57] very simple five-step process. Step one,
[18:00] give context. Explain what you’re trying
[18:03] to achieve, who your audience is, and
[18:06] what constraints you need to consider.
[18:08] The more relevant information you
[18:10] provide, the better chances of getting
[18:12] something useful. Step two, generate
[18:15] your draft. Ask AI to produce whatever
[18:18] you need. It might be a learning
[18:20] activity, a checklist, a scenario, or a
[18:22] set of questions. Step three, review it.
[18:26] And this is where your L&D knowledge
[18:28] comes in. Does the output actually meet
[18:31] your objectives? Is it appropriate for
[18:34] your audience? Is the information
[18:36] accurate or is anything missing?
[18:39] Step four, improve it. Make the
[18:41] necessary changes. You can do this
[18:43] yourself or ask AI to help you revise
[18:46] specific parts. And step five, validate
[18:49] it. Depending on what you’ve created,
[18:51] this might involve consulting a subject
[18:53] matter expert, testing an activity with
[18:55] a small group, or getting feedback from
[18:57] a stakeholder.
[18:58] Let me give you an example. Imagine
[19:00] you’re designing a workshop on giving
[19:02] feedback. You ask AI to suggest an
[19:04] activity, and it comes back with,
[19:07] “Divide participants into groups and ask
[19:09] them to discuss what makes feedback
[19:11] effective.
[19:12] Okay.
[19:14] There’s nothing inherently wrong with
[19:15] that activity, but let’s say that your
[19:18] learning objectives is for participants
[19:20] to practice giving constructive
[19:22] feedback.
[19:23] Does discussing feedback help them
[19:25] practice giving it?
[19:27] Not really. So, you review the
[19:29] suggestion and realize the participants
[19:31] need an opportunity to actually perform
[19:34] the skill. You could replace the
[19:36] discussion with a role play. Give
[19:38] participants a fictional situation where
[19:40] an employee has failed to communicate
[19:42] important information during, say, a
[19:45] shift handover. Ask one person to play
[19:47] the supervisor and practice giving
[19:49] feedback. The other person plays the
[19:51] employee, and a third person observes
[19:54] using a simple checklist. Now,
[19:56] participants are actually practicing the
[19:58] behavior you want them to develop. You
[20:00] could then ask an experienced supervisor
[20:03] to review the scenario and confirm that
[20:05] it reflects realistic workplace
[20:07] situations.
[20:08] And perhaps pilot the activity with a
[20:10] couple of people before using it in a
[20:13] large workshop. That’s much more
[20:15] effective use of AI.
[20:17] The tool helped you generate an idea.
[20:20] Your L&D experience helped you turn it
[20:23] into a meaningful learning activity.
[20:27] And this is exactly why I think
[20:28] developing a solid foundation in
[20:30] learning and development is so
[20:32] important. Particularly now that AI
[20:35] tools are becoming part of our everyday
[20:37] work. You need to understand learning
[20:40] needs analysis, learning objectives,
[20:42] instructional design, learning methods,
[20:44] and evaluation to make informed
[20:46] decisions about what AI produces.
[20:49] If you’re new to the field, our L&D
[20:51] Fundamentals course covers these
[20:53] essential concepts and helps you
[20:55] understand how the different parts of
[20:57] the L&D process fit together. And if
[21:00] you’re looking for practical tools to
[21:02] support your work, you can also explore
[21:04] the ultimate L&D toolkit, which includes
[21:06] templates, checklists, and other
[21:08] resources for different L&D activities.
[21:11] I’ll leave the links to both in the
[21:13] description below.
[21:16] >> [music]
[21:17] >> Now, I don’t want you to finish watching
[21:19] this video thinking, “Well, that was
[21:21] interesting.” And then carry on doing
[21:23] exactly what you’ve always done. So,
[21:26] here’s a little challenge. I want you to
[21:28] try using AI for three small tasks this
[21:31] week. Nothing complicated, I promise.
[21:34] Experiment one,
[21:35] prepare for a meeting. Think about a
[21:37] stakeholder conversation you have coming
[21:39] up. Give AI some context and ask it to
[21:41] suggest questions you could use. Review
[21:44] those questions and see whether any of
[21:46] them would help you explore the topic
[21:48] more effectively. Experiment two,
[21:51] improve a learning activity. Take an
[21:54] activity you’ve already designed.
[21:56] Explain the learning objective, the
[21:57] audience, and how the activity currently
[21:59] works. Then ask AI to suggest
[22:02] improvements. You don’t have to accept
[22:04] its suggestions, obviously. In fact, you
[22:06] might decide that your original activity
[22:09] was way better.
[22:10] And that’s perfectly fine, of course.
[22:12] The point here is to explore different
[22:15] possibilities. And experiment three,
[22:17] simplify a piece of content. Find a
[22:21] paragraph from an approved learning
[22:23] resource or maybe even a learning
[22:25] policy. Ask AI to rewrite it in a
[22:28] simpler language while preserving the
[22:31] original meaning. Then compare the two
[22:33] versions.
[22:34] Is the new version actually easier to
[22:36] understand? Has anything improved?
[22:39] Or has anything been removed that’s
[22:41] really important? Would it work better
[22:44] for your audience? And here’s the really
[22:46] important part of this little
[22:48] experiment. Keep track of how much time
[22:51] you’ve actually saved. Because we often
[22:54] talk about AI making us more productive,
[22:57] but we don’t always stop to measure
[22:59] whether that’s true. For each
[23:01] experiment, note how long it took AI to
[23:04] generate the draft and then note how
[23:06] long you spend checking and correcting
[23:08] it. And finally, consider whether the
[23:10] result was better than what you would
[23:13] have created yourself without the use of
[23:14] AI. If AI produces something in 30
[23:17] seconds, but you spend 45 minutes fixing
[23:20] it,
[23:21] um perhaps it wasn’t quite the time
[23:23] saver you thought it would be. On the
[23:25] other hand, if it helps you complete a
[23:26] task in 15 minutes that would normally
[23:29] take you an hour, that’s worth knowing.
[23:31] The goal isn’t to use AI for the sake of
[23:34] using AI. The goal is to find the tasks
[23:37] where it genuinely helps you work more
[23:40] effectively.
[23:44] So, let’s return to the question that
[23:46] inspired this video. Will AI replace
[23:50] learning and development professionals?
[23:52] I think the answer is a little more
[23:54] complicated than a simple yes or no.
[23:56] Will AI replace some of the tasks we
[23:59] currently perform? Absolutely.
[24:02] In fact, it’s already happening. There
[24:04] are tasks that used to take me hours
[24:07] that I can now complete much more
[24:09] quickly with AI.
[24:11] And I have no particular desire to go
[24:13] back to doing all of them manually.
[24:16] But does that mean AI can replace
[24:18] everything an L&D professional does?
[24:21] I don’t think so. Because our work isn’t
[24:24] just about producing training materials.
[24:26] It’s about understanding performance
[24:28] problems. It’s about working with
[24:30] stakeholders to identify what needs to
[24:33] change. It’s about understanding
[24:36] how people learn and designing
[24:39] experiences that
[24:41] help them develop new knowledge and
[24:42] skills.
[24:44] It’s about evaluating whether our
[24:46] interventions have actually made a
[24:48] difference.
[24:49] And sometimes it’s about having
[24:51] difficult conversations with managers
[24:53] who are convinced they need a training
[24:55] program when the real problem has
[24:57] nothing to do with training.
[24:59] AI can support us with parts of that
[25:02] word, but it doesn’t remove the need for
[25:04] the professional judgment,
[25:06] organizational understanding, and the
[25:07] ability to work with people. Now, does
[25:11] that mean our profession won’t change?
[25:13] Of course not. I think it will change
[25:15] quite significantly, and I think L&D
[25:18] professionals who understand both the
[25:20] fundamentals of their profession and how
[25:22] to use AI effectively will be in a much
[25:25] stronger position to adapt to those
[25:27] changes.
[25:28] There is something I want to emphasize,
[25:31] especially for those of you who are new
[25:33] to L&D.
[25:35] Don’t confuse knowing how to use an AI
[25:37] tool with knowing how to do the job.
[25:41] Being able to generate a training
[25:43] program in 30 seconds doesn’t mean you
[25:45] know how to design an effective training
[25:47] program.
[25:48] Being able to produce a beautiful
[25:51] presentation
[25:52] doesn’t mean you’ve created a meaningful
[25:54] learning experience. And being able to
[25:57] generate 50 quiz questions
[26:00] doesn’t mean you’re measuring the right
[26:01] things.
[26:02] You still need to understand the
[26:03] principles behind the work. Because
[26:06] without that knowledge, how are you
[26:08] going to recognize when AI gets
[26:10] something wrong?
[26:13] >> [music]
[26:15] >> So, where does all of this leave us? I
[26:18] think AI is an incredibly useful tool
[26:21] for L&D professionals. It can help us
[26:23] prepare for meetings, organize
[26:25] information, brainstorm ideas, draft
[26:27] materials, and adapt content.
[26:30] And when used appropriately, it can save
[26:32] us a considerable amount of time.
[26:36] But, it also has limitations.
[26:39] It doesn’t automatically understand our
[26:42] organizational context. It can produce
[26:45] inaccurate information. It requires
[26:48] careful handling of employee data. And
[26:50] it can’t replace the professional
[26:52] judgement that’s very often needed to
[26:55] design effective learning experiences.
[26:57] So, my advice is simple. Learn to use
[27:00] AI, but don’t neglect learning how to do
[27:04] L&D. Because the better you understand
[27:07] your profession, the better you’ll be at
[27:09] using AI to support your work. And
[27:12] ultimately, that’s what we should be
[27:14] aiming for. Not generating more content,
[27:17] not producing more training programs,
[27:19] but helping people learn, develop, and
[27:23] perform better at work. If AI helps us
[27:26] do that more effectively, fantastic.
[27:29] If it doesn’t, then perhaps we need to
[27:31] rethink how we’re using it.
[27:34] If you’re just starting your career in
[27:36] learning and development and want to
[27:37] build a solid foundation, visit the L&D
[27:40] Academy. You’ll find online courses,
[27:42] practical downloadable resources, blog
[27:44] articles, and plenty of free educational
[27:47] content to help you develop your L&D
[27:49] knowledge and skills. And you’re already
[27:51] working in L&D, but struggling to figure
[27:54] out where AI fits into your particular
[27:56] role, we also offer one-to-one L&D
[27:59] consultations where we can explore your
[28:02] specific challenges and workflows.
[28:04] You’ll find links to everything I’ve
[28:06] mentioned in the description below. And
[28:08] before you go, I’d love to hear your
[28:10] thoughts. Do you think AI will
[28:12] eventually replace L&D professionals or
[28:15] do you think it will simply change the
[28:17] way we work? Let me know your thoughts
[28:18] in the comments down below. Thank you so
[28:20] much for watching, and I’ll see you in
[28:22] the next video.