For most of us, the arrival of Artificial Intelligence (AI) seems sudden, out of the blue.

Just a few years ago, your friends weren’t automatically tagged in images you posted on social media.  You couldn’t talk to Siri or Alexa.  And pedestrians weren’t getting struck by self-driving cars.  The speed with which AI has begun to transform our digital and physical worlds has been breathtaking.

Like most disruptive forces, we’re not yet sure what to make of AI.  Does it herald the end of human drudgery or the beginning of human obsolescence?  Will AI create more jobs than it destroys over the next 10-20 years?  Is Elon Musk right that the uncontrolled advance of AI represents an existential threat to humanity?

Before we assign sinister or benevolent traits to AI, we should take a moment to dig a little deeper and perhaps uncover AI’s true value.

AI has the equivalent of a large ‘family tree’, with many branches.  There’s been a great deal of investment in some technically very sophisticated branches, such as machine learning.  But when it comes to human learning, I’d argue that there’s a less-appreciated branch of AI with tremendous, immediate potential to improve the way people learn.  It’s called Natural Language Generation, or NLG for short.  Simply put, NLG is computers writing stuff.  It’s the branch of AI that automatically turns data into written language that humans can read, understand, and even enjoy.

Since the printing press was invented, we’ve found ever-better, ever-cheaper ways to get the written word to large numbers of people.  But there have always been two major limitations.  First, humans have always written everything, and writing takes a lot of time.  Second, the words published are uniformly the same, even though the individuals reading those words are uniquely different.  For centuries, it’s been this:  A person spends a lot of time crafting a single set of words, then a great variety of individuals – who may differ wildly from one another – reads those words, and the writer hopes it resonates with each individual.

But thanks to NLG, we can now dramatically speed up the writing process – as in, writing hundreds of pages per second – and we can now vary the words written in each case.  NLG allows us to tailor written language depending upon certain variables, so that not every person reads the exact same thing.  This means the more we know about someone, the more we can tailor written content so that it’s hyper-personalized and relevant to that particular individual.

My company, CredSpark, is an interactive assessment platform enabling learning companies, marketers and media firms to ask questions of their learners and readers, in order to generate insights and catalyze action.  CredSpark is a proud partner of Degreed, and we’re also among the first companies use to NLG in a new way:  To generate personalized recommendations to individuals based upon what they’ve told us about their knowledge and interests.

Our initial work with NLG personalization has been around professional conferences: how to make a large trade show ‘feel small’ by asking an attendee a few questions and then generating a written recommendation of the sessions, exhibitors, and products most relevant to that person.  The response among attendees has been extremely positive.

But we’re equally excited by the possibilities around personalized recommendations for professional learning.  Imagine you’ve arrived at a website with a long list of learning resources: articles, videos, webinars, etc.  Instead of having to spend lots of time filtering and scrolling, what if you could simply answer a few questions and have a ‘short list’ of the most relevant resources tailored exactly to your knowledge and needs?  Even better, what if it wasn’t a list, but a narrative, providing you with context around why these particular resources are relevant to you, thereby giving those recommendations real meaning?  That’s what NLG + assessment can deliver.

We think this is the true power of AI in learning:  The ability to deliver individually-tailored learning guides containing only the most relevant resources, wrapped in a narrative that conveys meaning and value to the learner.  People learn best when highly engaged, and there’s no better way to engage learners than with learning plans which reflect their unique identity and needs.  Further, such personalization can support learning among people with varying levels of prior familiarity, and who learn at different paces.

It’s our hope that Natural Language Generation, combined with interactive assessment, will be widely adopted to scale the delivery of personalized learning journeys, thereby making each learner the hero in the narrative of her own advancement.