AI, or Artificial Intelligence, is cropping up more and more in eLearning conversations—who’s using it and how, and what it means for the future of corporate digital learning. As Learning Solutions prepares to explore AI from many angles, an overview of foundational aspects of AI might come in handy. Here, we’ll introduce concepts and trends that are likely to appear in any deep discussion of using AI in eLearning.

Foundational terms and concepts

Artificial intelligence refers broadly to technologies that can learn and perform specific tasks.

More complex tasks entail machine learning, a next-level technology that takes an AI machine or technology and teaches it to make “decisions” based on algorithms, learn from those decisions, and refine its own performance. Consumer examples of machine learning include the algorithms in Amazon or Netflix that make recommendations based on a consumer’s previous choices, their algorithms, and volumes of other data they have aggregated from a variety of sources.

Both AI and machine learning rely heavily on algorithms. An algorithm is a set of steps that a person—or a computer—follows to accomplish a task. You can understand a recipe, or the driving directions your dentist’s website provides as simple algorithms. Other algorithms require vast amounts of computing power. Some algorithms might return a poor answer or even an incorrect one. In some cases, an algorithm that very efficiently produces a “good” answer might be preferable to a “perfect” algorithm that would choose the very best option—but would also require far more time and computing power to reach that decision. Choosing (or creating) the most appropriate algorithm for a task can make the task faster, more efficient, or more successful. An algorithm that improves with use is employing machine learning.

AI in eLearning development

Artificial intelligence already shows up in many areas of eLearning. An example might be a chatbot or quiz program that “decides” which question to ask a learner based on the learner’s previous responses, e.g., adding questions in areas where the learner has had more incorrect responses or skipping to a higher level when the learner consistently enters correct responses. Another common use is in answering simple questions: The AI algorithm can be “taught” basic information that, say, new employees might need. It can be presented as a smart chatbot that “understands” and responds to conversational text messages and questions from those employees. Note that AI applications are not actually sentient and do not think, decide, or understand; they apply logical rules to accomplish tasks that approximate these cognitive tasks.

While building AI into eLearning is going to remain a hot topic in L&D for a while, using AI in the actual eLearning content is not the only way to apply AI to improving eLearning. AI algorithms, and especially those “smarter” machine learning technologies, can help developers create better eLearning.

One way that AI can aid in eLearning development is by improving the classification of elements of content. As digital learners increasingly expect content to be offered in multiple formats and on a variety of platforms, many L&D teams might find themselves “repurposing” content and implementing “plus-one” design. A piece of information might be presented, say, in a video, an infographic, the text file containing the video transcript, and a chat-based quiz that offers feedback.

A developer could try to track all the places a concept is mentioned and explained, then locate the content to reuse it. Alternatively, an AI-based tool could sift through massive amounts of content, identifying and efficiently locating suitable content, thus enabling the developer to quickly reuse content. Automating this process enhances the consistency of materials across platforms while also saving development time.

Where to start: AIaaS and eLearning

You might be wondering where you can get your hands on some AI tools. Fear not; it’s not out of reach for even small (read: resource-poor) L&D professionals. The advent of AIaaS or “AI as a Service” allows eLearning developers to purchase or license algorithms and AI tools and components, thereby avoiding the time and expense of developing their own. These solutions are not applicable to every situation where AI might be useful in a given company’s eLearning ecosystem, but they do offer some enticing benefits, including the ability to add “standard” AI tasks to your toolbox. A standard AI task is one that is part of many well-researched and frequently used solutions. This could be something like identifying specific objects in photos (dogs, for example) or certain logic and decision-making tasks that use similar processes.

Many AIaaS products are cloud-based, potentially requiring that developers—and possibly learners—be able to access the cloud whenever they use the tools. Several well-known tech giants offer AIaaS tools and platforms:

  • Microsoft Azure is a cloud-based set of services that developers can use to build and manage AI applications, such as image recognition or bot-based apps.
  • IBM’s Watson offers cloud-based AI services that can be integrated into your applications; you can store and manage your own data on the IBM Cloud for use with the Watson platform.
  • Google’s TensorFlow is an open-source machine learning software library.
  • Amazon Web Services encompasses myriad services and products, all located on Amazon’s cloud.

AIaaS is not only the playground of the wealthy and enormous, though. Butler Analytics published a list of more than 20 AIaaS platforms over a year ago; the field has only expanded since. When considering using AI in eLearning and eLearning development, it’s advisable to choose a platform carefully. Each of the various platforms has its own strengths and weaknesses; which is the “right” one is highly dependent on the needs of the developers and learners as well as the technical skills of the developers who will be harnessing these tools and services to create AI-enhanced eLearning tools.