Artificial Intelligence is the talk of the day when it comes to new methods of training and workplace learning. AI is a very wide subject, covering many aspects of technology like big data and the Internet of Things. In the context of training, AI is mostly centered on making sense of big amounts of learner data so that each learner gets personalized suggestions on what comes next along their learning path.
One very important part of AI is machine learning. In fact, AI couldn’t have been possible without machine learning.
In common language, machine learning means that the computer has the ability to learn something without being programmed explicitly to act in a certain manner.
With machine learning, programs learn just like people: through experience and training. It’s like learning how to ride a bike: just because you know in theory what the pedals and the wheels are supposed to do, it doesn’t mean that you can ride it from day one; you need to practice and fall a few times before you get the hang of it. After a while, you can enter bike races, or simply overcome the challenge of riding it through the urban jungle.
Going back to the idea of machine learning, you feed the program all the data you have on a subject, but it’s an enormous task to think about all the possible answers. Machine learning means that the program will learn how to get all those possible answers, and improve on its own by experience and repetition.
This method is already being used on a big scale by investments and insurance industries, the medical field, and more. Whenever you see and click on a Google suggestion, you are seeing the results of machine learning. The Google algorithm could not output the relevant data of what many other people are searching for, and make those suggestions, without machine learning.
It can come up with solutions a lot faster for the mere reason that it can access and parse a huge amount of information in a very short time. Not to mention that when it learns something, it stays. People learn new things all the time, but we also forget many of them, or simply choose to ignore them. An algorithm will not forget or ignore a solution due to feelings or fond memories on a subject, so it’s safe to say it will offer better and faster results over time.
But since this is a blog about e-learning and online training, the following question begs an answer: what can machine learning do for those of us who are creating or attending training sessions and learning every day a little more to improve ourselves?
I really think that machine learning is a highly valuable asset for the training industry and here’s why.
What exactly can machine learning do for L&D?
Well, I have two words for you: personalized learning. Everyone is talking about why it’s important to offer personalized learning experiences for trainees, and the only way to achieve this is through big data. And with the help of machine learning, that big data about learners is sorted, patterns are found, so instructional designers can make sense of everything and take the best decisions when creating training courses.
Machine learning can take teaching and learning at a whole new level of customization. Just imagine how it would be if the very first moment someone starts working for your company, a dedicated software will start working alongside with them to gather all the information about how they like to study, their favorite topics and how they do on tests.
I admit it does sound a bit Orwellian but bare with me for another moment. The whole purpose would not be to spy on the new employee but to analyze and process information that would give you very good solutions when it comes to preparing new engaging training materials. Wouldn’t you use this data when creating training materials for those new hires?
People learn in different ways and at different paces, and have our routines for processing and retaining information. For example, my algorithm for studying what I’ve learned in school was this: I read once the full text; then I’d divide it into sections and study each one of them in depth; then I’d go over the whole subject again and do a recap. Sounds simple right?
A machine learning program could pick up on these patterns and algorithms a learner uses and offer important insights and recommendations about how the employee learns best, which materials grab his/her attention first, what’s the best time to introduce new concepts, what’s the retention rate, and so on. It would be like having an army of experts and scientists learning about their way of learning to improve the outcomes.
Every trainer knows how difficult it can get to grab everyone’s attention during a training session and address all the various learning styles the participants have. Machine learning would give you the opportunity to prepare a truly personalized learning experience for those attending.
Just think about online courses. Depending on your company size, there are maybe hundreds of learners accessing the training resources at the same time. It would be quite helpful to get some extra insight about their learning patterns, the time spent on a module, completion time and percentage, or suggestions about extra resources that would benefit the learner, etc. I wouldn’t mind getting automatically enrolled in an Excel class, because the algorithm “watching” my learning pattern noticed I google Excel functions whenever I have to deal with big spreadsheets.
At some level, machine learning is already used in online training with the help of learning management systems. Any good LMS can track learners’ progress, time spent on each module, scores and more, so that the instructors know when it’s time to help a learner or give them access to more information on the next module, or see who’s falling behind on a certain subject and provide assistance.
I’m sure that in the future LMSs will become more and more complex and will be able to customize the learning experience more deeply and easier than it happens today.