AI Has the Potential to Reshape IT Training and Certification
At the end of February, I wrote a related story for GoCertify titled "The Upcoming Role of AI Is Everywhere." Since then, I haven't been able to get this topic out of my head. Lately, I've been pondering the ways in which AI is already and will be changing the ways that training and certification work. (Note: For the balance of this article, I'm going to lump training and certification together under the catch-all label "traincert.")
I see AI's fingerprints all over this business sector. The Statista graphic below shows that traincert is big business in the United States, and certainly draws in the kinds of revenues that could (and, in fact, do) justify bringing AI to bear. Check it out:
Source: Statista — Training industry in the United States
What Can AI Do? Plenty!
In fact, AI can provide all kinds of interesting insights and information for training developers, training delivery and marketing, and trainees alike. Each of these areas has a different focus and will no doubt put AI to slightly different uses as a result.
Let's explore them in the order just presented, and dig in a little deeper. In general, what I say in the sections that follow for training goes more or less the same way for certification tests and labs or simulations, unless I specifically observe something added or different.
AI for TrainCert Developers
In testing and in maintenance mode, AI can provide information about training material modules, test items, and supporting materials. It can target items where students are having trouble understanding or applying the content effectively or efficiently.
This can come from analysis of time spent on the materials, error rates on related questions, labs, and practice materials, as well as from the frequency and type of questions or comments on those selfsame materials. In general, AI works well to let training developers establish and maintain various levels of understanding and comprehension.
As students interact with self-paced and online training materials, training delivery platforms can acquire rafts of statistics and measurements about them, which then become grist for the AI mill. I have to look at this as an unalloyed benefit, because it can only help make things more consistent, while also making sure they work as well as possible for their intended audiences.
This approach has been driving certification testing for years. That said, such testing has primarily relied on statistical and psychometric analyses and feedback to enable continuous development and continuous improvement — but now there's plenty of room to bring AI into that picture too.
AI for TrainCert Delivery and Marketing
AI is already heavily used on sites like Amazon and Newegg, which spend lots of resources collecting data about shopping habits, purchasing clusters, and trending products and materials. Based heavily on machine learning, it's what lets those platforms recommend specific items in categories (e.g., best USB 3.2 or USB-C cables) and present you with "people who bought this item usually bought these other items" kind of data.
Guess what? The same thing works for large collections of training materials. Thus, a big platform provider can tell somebody who signs up for a class on Python that he or she might benefit from some Scrum training, as well, to better understand how to develop in an Agile environment.
I expect this will make an increasingly big difference for platform providers and their customers. The benefits will be clear as the former build bigger and more focused collections of traincert offerings and the latter spend more time learning from them.
AI for Traincert Consumers
The sky is the limit here on what AI can do for learners, but much of that territory is at present unexplored. I could see learners getting feedback on their learning styles and techniques, and thereby obtaining materials tailored to fit them better.
I could also see generation of drills and supplementary learning materials, where leaners need extra help or outright remediation to prepare them to fully understand and appreciate what they should be learning to master various topics, subjects, concepts and so forth. Skills assessments and inventories can be sharpened.
Learners can then focus learning time on areas where they need to get exposure, develop understanding, and build skills and experience. They would no longer have to work through an entire set of objectives for the subject matter, even if they're already test-ready for some subset of these objectives going into training.
Practice tests and further assessment will keep them on their toes with familiar materials, while they develop and improve in areas where time and effort will do the most good. For some, this will shorten the learning cycle; others new to various topics or certifications won't notice anything different.
As I said, I see the influence of AI on traincert (and education overall) as huge and transformative. If we can put this powerful toolset to good work, then it has the chance of improving learning for all parties involved. It also shows promise of improving the efficiency and utilization of learning platforms and resources, which could deliver interesting (and occasionally, time-saving) impacts on the learning experience.
I look forward to learning more about this, and talking to major learning platform providers and operators about how they're putting AI to work in their environments.