Where is the headline for L&D?
I’ve been talking to L&D conferences for over 10 years. I have always had the same approach to learning. Talk about what industry leaders are actually doing, talk about what participants want to solve or struggle, attend sessions to understand trends, meet new friends in the lunch or in the hallway, and work , human connection/reconnection about life, the universe. This article (one of two) covers common takeaways at the ATD LearningTechnologies Conference.
Even the TechKnowledge conference is about humans
This year, the theme of ATD’s Tech Knowledge meetings can be summarised by the shirts worn for the session.
Human and kindness (become both)
We humans are complicated. Some of us are even intelligent 🙂 How we treat each other, especially those who may oppose us, are trained to how artificial intelligence (AI) is trained. It plays an important role. After all, AI is trained on human data. Human stories, that is. Humanity is about hidden human stories. Talk to people. Listen to them. You never know until you know their entire story. Be kind. Don’t expect technology (even AI) to solve your human problems.
Why this long intro?
For the first reflection below: first find and understand the problem worth solving, then look for a solution (AI or not). Otherwise, you’ll be tired and overwhelmed by how quickly the tools change in your hands, chasing the problems that can be solved. Many people struggle today, especially in the world of AI. Donald H. Taylor’s annual “What’s Hot in L&D” survey shows the same. AI is the best. [1]
Reflection of the meeting
1. Many people are tired and overwhelmed
The pace of change hurts. Many people are exhausted just reading articles, listening to podcasts, and watching Tiktok. The moment you learn, reflect, apply and share the latest, it may be outdated. It’s no wonder there are very few books on AI.
AI also has an indirect impact on existing technologies. Take a look at the Edtech industry! They know that if they don’t claim that there is AI integration, they’re behind the curve. Creating content is the simplest entry point. Early Adapter has begun implementing live chat functionality to leverage conversational AI. Today we can talk to some pretty realistic 3D characters. In a few years, this feature is on all decent learning platforms.
Continuous changes force us to make decisions. Do I have to wait until I’m ready for a full-scale feature, or do I need to “hack” the current version? Building APIs and using code can become technical debt in the years when vendors are bringing their own.
How to reduce fatigue in change You know you’re not alone
When you’re in your own organization’s trench, it may feel like you’re years behind, and everyone on LinkedIn is an AI and L&D expert. Talk to other people. network. I’m going to the meeting. You can see that everyone has the same problem. You can’t do it all: prioritizing
You will no longer be a master at all trades. That era is gone. Prioritize important things. What is important to you, to your team, and to your organization? Let’s start with the problem
Most of your tasks remained the same. Nowadays, just because there is technology that can solve problems that are not worth solving is not worth it. Follow others in 3 Circle Challenge Circles
People in trench who have similar challenges you deal with motivational circles
People who are six months-a year ahead of where you want to be. People who motivate you to focus in the short term. Inspiration Circle
People who are years away. People dealing with the tsunami of new information, research and technology changes. Use them as buffers. Let us tackle the important things and distractions. 2. AI development: Return to behavioral change?
We deployed the co-pilot and no one is using it.
This recurring theme reminded me of my career experiences in which employees “draft” profiles where they were supposed to fill out information about themselves, interests, skills, and more. What’s the result? A miserable adoption rate. Solution? Essential HR Mandate: People need to complete their profile.
They don’t deploy technology like the red carpet. There is a red carpet so people don’t go to events. The event is there so they walk on the red carpet.
Make “events” meaningful and they will come. Same as AI: Yes, change management is required. Management of behavioral change, that is. Behind it is the whole science! Learning design is not enough to change your mind. Start with self-determination theory, BJ Fogg’s map model, COM-B, or similar foundations. There is something practical about the structure of a rollout plan from the world of game design.
Stage 1: Discover
Find out who has a challenge, problems, and processes. Show us your potential value! First, they have them look at their destination first and before giving them step-by-step instructions on how to get there. Stage 2: Exploration
You got your first buy-in. The motivation is high. The experience is low. This is the first time I’ve used AI. Handheld and serious mistakes may be necessary. Give them an early victory! A small thing that acts as a progression. Provides basic data and AI literacy. Stage 3: Scaffold
Now they use tools to solve problems. Motivations get lower as they burn energy, but their experience and skills increase will help them stay involved. We will support you if necessary, but don’t hold everyone back with a structured live session. Build and share solutions. The challenge of scaffolding with support tools and materials. Stage 4: Learning (and beyond)
You have a champion. Self-efficacy (“I can see myself doing this well”) drives solving for new challenges. Connections and relationships built around solutions. Provide continuous support (this is teamwork. You are not all experts). Experts can help board new employees based on lessons learned. Communities can maintain themselves. Best practices can be saved and shared within an AI system (no need for another SharePoint site). 3. AI Upskills: Where should I start?
One of the most common challenges mentioned at the meeting was to increase the large workforce. I have often found the phrase “skill” and “close the skill gap” to be misleading. What does it take to lift someone to the level? Three things: the level they are at, the level where you want, and the shortest path to connect the two.
But somehow, we often focus only on the desired state. We don’t know where individuals are, so we build one path from wherever we go. Then we force everyone to go back to the end of the road and begin their journey regardless of where they are.
“If I don’t see it, it doesn’t exist” policy is not good to rely on. Employees use personal access to AI to help them resolve problems and get them back to work. That may not be the best secret, but there are many assumptions out there. What is the company’s response? Block external copy paths. Well, there’s an email for that.
Build policies that also provide critical awareness of risk. Next, you can think of skilled things, such as how to assess your current ability to enable the shortest path.
4. Is quick engineering worth learning?
A year ago, rapid engineering was one of the hottest skills. To fast forward to today, there are thousands of acronyms and “frameworks” about how to write prompts. You can also ask your favorite LLM to generate a prompt. Generated AI uses natural language processing, so prompts are not actually “engineering” in the traditional sense. No code required.
My 2 cents means you should learn why, not acronyms or templates. Once you understand why you need to provide a specific context, it’s easy to adjust. What’s certain is that these models continue to evolve. What I learned a year ago might not be necessary today at all. We will focus on explaining the problem rather than the structure and format proposed based on last year’s success.
I always give the model permission to spend time, think, and reconfirm the latest answers. Before that, it often suggested the first and most popular coding, for example, and then it was recently discovered to have been deprecated. Also, don’t forget that you can ask the model to modify the answer. Over and over again. Humans don’t tolerate it, but AI is happy to repeat it.
5. Practical use cases of AI (in the case of L&D))
The most common use case I’ve seen is content generation. Prompt-based instant text and graphic content creation solves efficiency issues. Being more efficient means you can create more content with less effort. The challenge lies on the opposite side of Measurement Coin: Effectiveness. Creating more content faster doesn’t have more impact on your work. In fact, this efficiency leads to “free time” that is often consumed by creating more content. I think we need to use AI in the opposite way: content reduction.
Our session shares AI: personalized learning, coaching, adaptive content, authentic skill assessment, practical implementations for building no-code applications in minutes to quickly iterate, and even more Gaming interactions with AI-driven 3D text interviews (very long, multiple-choice answers).
Outside of learning, check out some of the AI use cases in your business [2]. In the next article, we will continue with the next five themes of the meeting:
Accessibility: Who cares about others? Are you waiting for God Tech? Collaboration Technology is not collaborating. Humans have better results on diverse ideas and personal notes: they live next to Alice. Alice? who [beep] Is it Alice? References:
[1] Global Emotion Survey 2025
[2] Artificial Intelligence (AI) Use Cases and Applications
