In our last blog post, we explained what artificial intelligence (AI) is, now we’re going to look at how we’re using AI in our learning solutions.
Natural Language Processing
Natural language processing or ‘NLP’ is a form of AI that allows computers to understand conversation. Traditional computers are like a well-trained dog, they respond to specific commands like “fetch” and “sit”, but if you say something like “Fido, be a dear and pick up the ball for me,” you’ve got no chance. Now, using NLP, computers can understand that kind of conversational command in both text and speech form.
When you’re a baby and people constantly point at your mum and say ‘mummy’ you eventually deduce that this word must refer to this woman, and most of our initial language is acquired like this. The pattern-spotting and learning abilities of AI mean that computers can acquire language and incrementally improve their understanding in the same way as a human.
By examining many examples and spotting the pattern, AI can establish the meaning of a word based on its context and position in a sentence. For example, if you’re talking about fielders and bowlers, you probably mean the game ‘cricket’, rather than the noisy insect. Like a human, it can also learn from its mistakes. If it’s told that it incorrectly deduced that the word ‘cheese’ meant a foodstuff rather than a type of music, it will behave differently next time.
Our chatbot, Flo, uses NLP to understand users’ requests. This makes accessing information and resources a lot simpler. Instead of sifting through layers of poorly designed menus in an LMS, users can just say “which courses are mandatory?” and instantly get the answer they need. Flo interacts with our LXP, Stream, to do this. We use another type of AI on Stream’s dashboard.
In this day and age, we expect a personal experience from our entertainment services. Netflix and Spotify are the shining examples of this – Netflix is constantly suggesting new films and TV shows that you might like based on your previous behaviour. Netflix has 120 million users across 190 countries, and it’s the access to this vast amount of data which makes their recommendations so accurate – if 10 million people watched Die Hard and then also watched Taken, then it makes sense to recommend Taken to Die Hard viewers.
Stream uses similar principles to recommend content to learners – if 500 contact centre operatives found our ‘Introduction to Metrics’ course useful, then we can recommend it to other people in the same position or industry. Of course, we’ve got slightly different algorithms and different ways of delivering our recommendations because Netflix is a platform for entertainment and relaxation and Stream is about self-improvement and productivity. We’re also still learning – our focus at the moment is on constantly improving the metrics that we base our recommendations on – whether that’s role, industry, age, type of content, length of content, or something else we haven’t thought of yet!
Companies are full of knowledge! If you’ve got a question, the chances are that there’s a document or a course about it somewhere. But that’s the problem – it could be anywhere. It could be buried in an intranet, hidden in an LMS, or languishing in a chock-a-block cloud server. Flo uses NLP to give you immediate access to the content you’re looking for and Stream’s recommendation engine gives you the information you didn’t realise you needed. It’s the difference between having a textbook and having a teacher.
AI also allows us to deliver meaningful learning which is grounded in learning theory. For example, with access to a database of content of varying lengths and mediums, we can promote spaced practice – pushing content to learners at specific intervals which combat the forgetting curve and increase recall.
We’ll explain more about Stream’s focus on learning theory in our next blog post.
About the author
Learning Designer, Matt started out his career in marketing but soon realised his talents were better suited to writing content. He’s spent the best part of the year immersed in the world of chatbots and conversational UX, drawing on the dialogue writing skills he developed during his Drama and Creative Writing degree – to put things simply – he’s teaching a bot to speak.