Thousands of pocket lamps: how the ia lights the dark corners
In my two previous articles in this series, we explored the effect of street lamp, that is to say that when we tend to look for something where it is practical (under the reverbera) rather than where it is (The Dark Park). In the case of learning and development (L&D), this means measuring easily available measures that are under our control (complements of courses, time spent in training and satisfaction). This last article will highlight (puns) the darker corners of the measurement of the impact of learning, and will show how we can consider AI as a force multiplier in L&D analyzes.
Why is it difficult to measure the impact of learning?
Aligned with several studies and my own experience, the latest research on the ATD on the future of learning assessment has revealed the same obstacles and challenges (1):
- Lack of time and resources
- Lack of access to data
- Lack of skills
- Lack of membership and support for stakeholders
Does that seem familiar? No wonder L&D remains in the well -lit area of the LMS. In order to measure the impact on work, we must get out of the LMS bubble and work with the company, IT, the acquisition of talents, etc. We need many pocket lamps for dark corners. Until now, scalability due to the lack of time and resources seems to be one of the largest barriers.
Technology will not remove all your barriers. Culture, lack of clarity, broken processes, objectives and unclear responsibilities, lack of responsibility, etc., must be discussed by humans before artificial intelligence (IA) can help.
The ability to refer to many places at the same time, quickly and intelligently, to see the whole impact is what data analysis, AI and the promise of automation.
6 AI ways as a force multiplier can help measure and assessment
1. Help with the strategy
Priority, compromise analysis, back design and king's calculations are some of the examples where automation and AI can provide you with advice on what to take first and how success must be measured.
2. Help in design
Even before the measurement and evaluation, you can use AI to help you with the evaluation writing, for example. For learning designers, I built an AI bot that analyzes your evaluation questions and provides a rating, suggestions and detailed comments on the approach. These assistants are now evolving in agents with the capacity to perform actions, not just actions.
3. Transforming satisfaction surveys into questions focused on performance
Working the rear appearance satisfaction surveys in questions focused on information producing information on usable data is another area where AI can help. Another AI service has been trained on the learning transfer assessment model (LTEM) and the design of performance -oriented survey to help create more exploitable data (2).
4. Analyze the data on a large scale and in depth
AI to the extent L&D can help us collect and analyze data on a scale and depth that were not previously practical. When a human analyst could find it difficult to correlate training data with six different performance measures distributed over three systems, an AI -based tool can increase these figures in seconds and punctual models. For example, AI can follow the learning data alongside performance measures over time to identify correlations, compare groups (which have followed the training in relation to which did not), and even analyze qualitative data (such as survey responses or open work product samples) to see how learners apply skills (3).
5. Lack of time and resources
Open text responses, real -time chat interactions or small escape group conversations can now be analyzed to summarize information, find models, categorize responses, predict feeling, etc. Lack of time and resources? Resolved.
6. Immersive dialogue and integrated measurement
During the ATD Technowledge conference in February 2025, I shared a prototype of a 3D adventure where users could interview people of interest on the basis of a given section of best practices. AI characters interacted in real time and they had their short and long -term memory. They shared facts on the world, but also had a conscience of each other. In the end, the AI coach provided a detailed analysis of the interviews. All that I built in a month. My prediction is that this type of immersive activity will soon be available on all decent learning platforms.
A 2025 industry report noted that advanced AI allows more sophisticated approaches to link learning and performance – instead of simply following the complements, analyzes fueled by AI can assess things such as understanding, application and change of behavior, which are “real engines of commercial performance” (3). This means that AI is not dazzled by the reverber: it actively seeks the light of the impact in the dark.
Predictive analysis to provide usable information
In addition, AI can predict and prescribe. Thanks to predictive analysis, AI could underline which employees are likely to benefit the most from special training (so that L&D can better target interventions). He can also help identify if a performance problem emerges that training could help, essentially alerting L&D to a need even before the company requests. In our metaphor, AI could not only shine a light where the keys are located, but even predict where you should look first (“based on past models, the keys are generally abandoned near the park bench”).
And finally, privacy and ethics cannot be ignored – to pass light everywhere should not mean spying employees or violating confidence. The objective is to shed light on the impact, and not to interfere on privacy.
We have technology to really measure what has always worried us: real behavior change and business results in a evolutionary and real -time way. Consider AI as a force multiplier of your impact in the new world rather than a threat to your work in the old one.
A brilliant future: measure what matters in all L&D roles
Getting out of the narrow circle of street light and in a wider and well -lit landscape is not only a good in have, it is the future of L&D. And this requires a change of culture that affects each role in the L&D field:
For educational designers
This means conceiving with a measure in mind. Kiss models like LTEM to make sure that your learning solutions include opportunities to demonstrate the application.
For managers and L&D program facilitators
It is a question of strengthening learning at work and following. You may need to associate with line managers to collect comments on behavior change, or configure post-training contact points (such as refreshments or coaching sessions) which increase the transfer and make it information on progress. Instead of declaring success at the end of the class, you will see your role extend to the workplace: guide managers on how to support new behaviors, and perhaps make light measures such as sampling of work results or the conduct of discussion groups to hear how people apply (or not apply) training.
For L&D leaders
These are strategy and culture. Directing the load in the alignment of learning on commercial objectives. Add the tools and resources (perhaps investing in an LRS, analysis talents, or AI platforms) which allow your team to measure what matters. He will also come across you to educate stakeholders. Define expectations with managers that L&D will bring in business results, not just activity, then hold this promise. Why not use the measurement section and prioritize project requests where stakeholders are ready to collaborate in the measure of the real impact?
For the learning of analysts or scientists of data
Your analysis and installation skills with AI tools will help translate raw data into significant stories. You will experience different methods (A / B tests for training, predictive modeling, etc.) to really understand causality, not just correlation.
Conclusion: AI as a force multiplier
In the end, avoiding the effect of the street lamp in the extent of L&D means having the courage to seek the truth, even if it is in difficult and difficult places. This means negotiating the immediate comfort of an easy metric for the more rewarding gain of a significant metric. Yes, it is more difficult to measure how new software training affected productivity than counting the number of people who have opened the training video. But which one would you prefer to bring to your CEO? Which one really tells you if the training has succeeded?
References:
(2) Surveys on learner and learning efficiency with Will Thalheimer
(3) Measure what matters: connect learning results to business results with AI
At Learnopoly, Finn has championed a mission to deliver unbiased, in-depth reviews of online courses that empower learners to make well-informed decisions. With over a decade of experience in financial services, he has honed his expertise in strategic partnerships and business development, cultivating both a sharp analytical perspective and a collaborative spirit. A lifelong learner, Finn’s commitment to creating a trusted guide for online education was ignited by a frustrating encounter with biased course reviews.