Artificial intelligence is beginning to deliver, mark and even create learning content, so what’s left for L&D? Will AI render L&D obsolete? Not quite; though AI will transform some aspects of L&D, creating new opportunities and new roles, others will remain essentially the same.
We’ve been talking about AI’s role in learning lately. You can read my colleague Peter Exner’s article on AI in learning for an introduction to the topic. But what does AI mean for L&D as a whole? What might L&D look like once AI is adopted, and what will AI offer your organization?
How AI Will Change Your L&D Processes
What is AI?
Before we take a closer look at AI’s implications for L&D, let’s first discuss what AI is, as there is no shortage of definitions. It’s often described as software that can complete tasks that would normally require a human brain. There’s a bit of a problem with this definition, as it often morphs into “something we haven’t figured out how to do yet”. After all, as soon as we build a computer that can do a task, it becomes evident that it didn’t require a human brain after all.
Another definition is “software that can analyze its environment and make intelligent choices.” Key features of AI are the abilities to interpret natural language and learn from experience.
What do you get when you combine software than can analyze its environment, understand natural language, learn from experience and make intelligent choices? The most obvious result is a personal assistant. Ask it a question and it figures out what you meant, searches available information and offers a (hopefully useful) reply. The workplace assistant is certainly one of the functions AI will perform in L&D, but there are others as well.
AI’s Emerging Roles in L&D
AI is beginning to play four interrelated roles in L&D:
- Content creator
- Workplace assistant
- Personal tutor
Let’s look at what these roles will entail, and how L&D will have to evolve to promote training effectiveness.
1. Content Creator
Have you ever googled “the”? My first hit was a local newspaper; its name starts with “The”. L&D research can feel very similar. You read through reams of material in search of information on vapor barriers in commercial construction and finally spot the elusive word ‘vapor’ only to find yourself reading about how humidity at the time of construction affects floor joists. It’s really not how you want to spend your time.
AI has the potential to not only identify material that is actually relevant, but to sort and arrange it as well, providing you with topical, valuable content for your learning event.
The new role of L&D personnel will be to review computer-generated content and provide feedback to help the AI learn from the experience.
2. Workplace Assistant
AI’s role as a workplace assistant is perhaps the most familiar to a modern audience, given the current ubiquity of personal assistants. Workplace assistants will come to both supplement and replace traditional training. Once trained themselves, they will provide immediate, accurate responses to most workplace questions: the epitome of microlearning, with all the advantages it entails.
Microlearning involves breaking learning into small chunks, ranging from a 10-minute module to a single question asked in the office. It’s a just-in-time solution that empowers learners, improves completion rates and increases retention and application of knowledge.
When a workplace assistant is used alongside traditional training, it will also contribute to the positive evolution of the traditional learning material. Questions often asked of the AI were evidently missed or poorly explained in training and will have to be identified and better incorporated into the base training material.
AI assistants require considerable training before they become useful. It’s the assistant’s ability to learn from experience that makes it so valuable, so it needs some experience to learn from before it is deployed. L&D will likely become responsible for initial training. Their role will be to identify quality resources and transfer that information to the AI.
L&D will also provide ongoing oversight. There’s a cautionary tale in the sweet, naïve Twitter chatbot that became a misogynist Nazi in under a day. Workplace assistants may be subject to less deliberate perversion, but other influences can still compromise their quality. For example, AI often relies on users’ ratings of the information it provides to learn the best answers to certain questions. In the workplace, a procedural shortcut that overlooks important health and safety processes may seem simpler, and therefore receive higher user ratings, than the safe, authorized procedure.
The L&D department is the most likely candidate for assuming responsibility for the quality and accuracy of what an AI learns, and ensuring the AI’s teachings remain aligned with organizational priorities.
3. Personal Tutor
Current AI tutors tend to work by helping to explain or demonstrate skills and by identifying concepts learners have overlooked. As AI tutors become more predictive, corporate models will likely begin to offer solutions for building competencies and extending capabilities. A personalized tutor will offer suggestions on what an employee needs to know to become proficient in their current role and, when that is achieved, provide recommendations on how they might extend their capabilities to make the most of opportunities for career growth.
As AI takes on decisions affecting future career opportunities, it will wield considerable power over the lives of individuals and the fortunes of organizations. Like all power, it can be used well or poorly. The Ethical Implications of Using AI in eLearning takes a closer look at this aspect of artificial intelligence.
We’re all familiar with computers that mark multiple choice tests. Peter Exner also mentioned the developing use of AI to grade essays and other written work. But as the AI tutor and the AI assessor develop, their abilities will combine to create a system that identifies learners’ competencies and growth areas, defines skill gaps, and directs learners to the material they need to fill the gap.
Equally powerful will be AI’s role as an assessor. It may wind up making decisions that affect job evaluations, promotions and opportunities for career development.
To some extent, this is already possible. Coursera is an online course system. Courses are developed by instructors and deployed to learners. When the system finds that many learners have answered a question incorrectly, it generates a tip for future learners and alerts the instructor to the problem so they can make appropriate revisions.
AI development will result in L&D departments shifting their focus from assessing learners to assessing learning material and events. Instead of determining whether a learner got an answer right or wrong, they will spend their time investigating what the right or wrong answers mean. If everyone got a question right, is it still important to ask? If too many people got it wrong, is it necessary to explain the concept better, provide more practice, or is the question itself unclear?
As artificial intelligence takes on the traditional L&D roles of content creator, workplace assistant, personal tutor and learning assessor, L&D will need to adapt to ensure that AI performs these roles effectively, safely and ethically. While the core L&D role of evaluating, correcting and enhancing learning materials and events will continue to be central, new roles will emerge for providing AI training and oversight, which will be just as important in the future workplace.