Continuous Learning, Human and Machine

Every new phase of every new project brings interesting challenges. The constant problem-solving and adaptation required to keep our courses moving through development can occasionally feel a bit like juggling plates while surfing a rockslide through a snowstorm. It's never boring!

As I write this, our team is wrapping a brand-new course: Machine Learning for Competitive Advantage for Athabasca University’s new PowerEd professional development unit.

If you’re not entirely sure what “machine learning” is, you’re not alone — it’s a big topic and currently, not many organizations are applying it in day-to-day operations. In short, it’s a type of artificial intelligence (AI). Machine learning uses data to answer questions and, the more data it is exposed to, the better it does its job. For example, machine learning could be used to predict consumers’ preferences and help businesses decide what type of products to make available. The more examples of different people preferring different things it encounters, the more accurately it can make predictions.

As someone with an interest in software design and artificial intelligence (AI), I found the course fascinating to work on. However, our team wasn’t only learning about the topic at hand, but also about each other as coworkers, the local technology community, and growing our skills as creative professionals.

For example, I was surprised to learn how much in-house expertise Onlea has in this area. The bulk of the course content was compiled by our own Adriana Lopez Forero, who in addition to being Onlea’s fabulous President, has her Masters in Computing Science and Artificial Intelligence.

We also received input from local experts in machine learning research and business solutions. Did you know that Edmonton is a hot spot for AI technology, not only in Canada but worldwide? I had previously heard that the University of Alberta was doing some interesting AI research, but had no idea the institution was ranked among the highest globally for their contributions to the AI and machine learning fields.


Working on the gathered material as a technical writer and editor was an awesome learning experience for me on several levels. An important challenge in the course was striking the right balance between granularity and accessibility. We needed to be specific enough that learners had a clear understanding of core concepts, without getting too far into the weeds with content that might feel irrelevant to their daily needs or alienate people without a background in computing science.

Working with experts from the Alberta Machine Intelligence Institute (AMII) went a long way towards identifying areas in the script that were not specific enough or might be a source of misunderstanding. AMII is already indepthly familiar with the challenges and potential pitfalls when communicating with business people and members of the public about machine learning technology.

For example, AMII is keenly aware of some of the need for terms that clearly differentiate components of a machine learning project. You may have heard of systems that use machine learning referred to as an algorithm — but this isn’t quite true. An algorithm is a set of rules or steps someone (or a computer) can follow to solve problems, but this is only one component of a machine learning system: the method by which the system learns from data which decisions are “correct.”

To help prevent confusion, AMII has coined a new term — Question Answer Machine, (QuAM) to describe a system trained by feeding data to a machine-learning algorithm to make decisions or predictions. If your organization was applying machine-learning technology in their day to day operations, the general term for the system you’d be working with would be a QuAM.

These kinds of distinctions don’t always seem hugely important, but for my role they are essential. Most of what I do is take information provided by others and shape it to better fit the needs of learners. Terms that are vague or concepts that are easily confused with each other are roadblocks to clear communication, and each time we find and remove one, it makes the whole course a little better.


Working through these types of challenges are particularly important for a course that covers an exciting but not well-understood technology. Machine learning is well into its latest trip on the hype cycle roller coaster, and anyone doing research on their own faces a slog through speculative fluff pieces, dense jargon, and technology evangelism to find information that is not only accurate and objective, but also relevant to their particular needs.

I don’t always get it right, but fortunately, I’m not alone. Onlea has a diverse team of professionals working to remove barriers to learning in different ways, with animations, videos, graphics, better interfaces, and strategic use of interactives like quizzes and exercises that challenge and engage learners, helping them to check in with their own progress while also reinforcing important concepts.

Looking at the end results, I’m hugely proud of the work we’ve done. The goal of any course we make is to not just to help learners understand concepts and pass tests, but to make a real difference in their ability to make decisions and do things in the real world. I think we’ve hit the mark on this one, and I look forward to seeing it out in the wild.

If you’re interested in checking out the Machine Learning for Competitive Advantage course or any of Athabasca’s University other PowerEd professional development programs, visit

Share this

Mel Guille

Mel Guille