5 ways AI is being used in learning
Posted on Sep 13, 2017
Artificial Intelligence (AI) is the next big thing, but how can you use it to create a better learning experience?
Look at any list of disruptive technologies and AI is likely to be at the top, it’s the latest tech buzzword to take over the news media.
As an L&D professional, what should you be paying close attention to? And what’s safe to ignore?
Our round up of the more useful applications of AI for learning includes many of the leading examples from academia and adult education that will be filtering their way into the workplace learning environment as AI becomes more widespread.
What is AI?
AI computer systems are able to perform tasks that would normally require human intelligence. A large part of most AI systems relies on machine learning in order to perform those tasks.
In machine learning, a computer program is trained to recognise patterns, or complete an act, like identifying a person's face, playing a game of chess, or responding to a learner’s request for information. To train the system, you expose it to as many examples of the task as possible, there are different ways to train using different types of input data.
After being given goals for the data, some AI systems train themselves by creating their own tasks. Google’s DeepMind AlphaGo project played itself millions of times in the build up to it’s match against the best human Go player, Lee Sedol.
The availability of data and relatively inexpensive access to massive computing power means AI systems are becoming a more viable option for any task that a computer can be taught to do.
Learning applications of AI
Any repetitive task that is improved by performing it over and over is a good fit for artificial intelligence. Here are some of the ways AI is being used in learning right now:
1: Teaching assistants
Natural language recognition, through text and voice, is one of the early success stories of AI, it allows more natural communication between humans and computers.
Conversational interfaces are being used in many areas of higher education, and there are opportunities to use the same techniques in workplace learning.
Georgia State University created a chatbot to help handle incoming student’s enquiries about enrolling in their course. In the pilot study, where half of the students were first contacted by the chatbot, and half were left to use normal methods, the chatbot group had a 21.4% lower rate of drop out over the summer before enrollment.
Georgia Tech has already had AI teaching assistants helping students on their Knowledge Based Artificial Intelligence (KBAI) course for three semesters. What started as a secret experiment has turned into standard procedure for the course and they’re seeing increased engagement from students as a result.
Allowing employees to ask basic questions and get answers immediately from an AI system will free up L&D time for answering more in depth queries, increase responsiveness and make better use of existing resources.
2: Artificial mentoring
Not all AI-based training is new, in 2010 the US Navy ran a pilot study to train new recruits in IT skills with an AI tutor. They first assembled a group of expert tutors, ran intensive one-to-one training sessions, then re-created the experience with an AI ‘Digital Tutor’. Competitions between recruits trained using the normal blended approach and those who had the Digital Tutor revealed better performance from the Digital Tutor group in every case.
This same principle, learning from an expert, is being used by Duolingo’s chatbots. Using an AI-powered chat interface, with support for text and voice conversations, you can learn French, Spanish and German in a natural way.
As the costs of machine learning and AI tools fall, it’s becoming possible for workplace L&D teams to take elements of current training and automate them through AI.
3: Generating and augmenting content
Creating content from source material and subject matter experts is one of the time consuming parts of building digital learning. AI systems can be trained to extract the important information and put it into a human readable format suitable for digital learning.
Content Technologies Inc. (CTI) are the world's largest publisher of higher education material, they use AI to create content based on all the articles and books that are available on a subject. Outsourcing the more straightforward preparation work gives skilled learning designers more time to create an engaging learning experience.
AI is also being used to augment existing content, automatically adding links to relevant material to videos for example. Introducing fresh information to existing content without valuable design and development time helps extend the life of your courses.
Content creation is one of the biggest challenges of effective workplace training, requiring many different skills to do well. AI has just started impacting on the more basic end of these skills and there’s sure to be more advances in this area as AI continues to grow.
Good learning developers create courses that can be accessed by everyone, and AI can be a powerful ally in their quest. To see how useful AI can be for partially sighted people take a look at Microsoft’s Seeing AI app.
The same sort of image recognition tools are already used on consumer platforms like Google photos to identify people and objects to make it easier to find relevant images. The photos or graphics used in your elearning courses might already have been chosen by AI in the background of your favourite stock image libraries.
The most common problems for people with disabilities were discussed in an impromptu Twitter survey, the top issue was: Lack of captions for video. YouTube have offered an automated closed caption system since 2009 which seeks to solve the issue with AI.
Next time you create a piece of learning content, consider letting AI create alternative text for your images and a caption your video, it’s not perfect, but it’ll save you time and help reach your whole workforce.
One of the most visible ways AI is being used today is in personalising experiences. Services like Netflix use machine learning to generate better recommendations for viewers, and have even developed their own specialised machine learning programs to help.
Modern digital learning lends itself to a personalised approach by allowing course creators to track the interactions learners make during the course. In trials of a personalised learning platform, 70% of students preferred it to a one-size-fits-all approach and engagement rose by 72%.
New tracking standards like xAPI and bespoke solutions, based on web standards will allow courses to be more closely targeted at learners needs in the future.
As AI becomes better and more affordable, more of the basic tasks of creating a learning programme will become automated. By getting familiar with the various ways AI can help your existing workflow you can adjust to a new future in which it’s embedded in all the work you do.
You may also be interested in
Learning expert, Donald Clark makes the case for AIDonald Clark has been creating and writing about all types of learning for over 30 years. His popular blog features a mix of learning theories...
Donald Clark has over 30 years of experience in learning technologies, we spoke to him about the latest emerging tech and how it will impact the workplace learning industry. We cover a range...
Nigel Paine is the former boss of the BBC’s training and development operation and headed up the broadcaster’s award-winning leadership programme. Now, a thought leader, speaker and author on...
Join the Sponge team
Great people are fundamental to Sponge. We employ many industry-recognised leaders and are busy growing the next generation of leaders. We don’t just look for talented people who can do a particular job – we look for people who share our passion and values.Read more