How can AI further accelerate the transfer of skills?

The recent advances in artificial intelligence continue to amaze us. Each week that goes by is accompanied by a new innovation, on the generation of images, text or other things. But is the e-learning sector really taking advantage of these innovations? Is it making the most of it to achieve its primary objective, which is to transform practices in the field? That's what we'll see in this article!

E-learning prioritizes the organization of information and not the transformation.

If we analyze the uses of generative AIs made in the e-learning sector, we observe 2 dominant uses:

First use: the “reorganization” of top-down information. The information from the initial text is reorganized in the pages of an e-learning course. The titles of the paragraphs become the titles of the pages, the main ideas are addressed first... You got the idea, here we speed up the transposition of content by simply filling in the boxes in the destination format (generally kilometers of one-pagers).

There are a few actors who stand out and propose to automatically generate exercises. But between empty text that can be solved by having a few notions of grammar and multiple choice questions for checking knowledge where the correct answer is the most detailed, most only offer superficial formats, which will not help the learner to transform his practices.

Second use: to facilitate access to this information. This is generally done through adaptive learning but also through chatbots based on generative AI, to which the learner can ask questions to better understand a point of the lesson (see an example). These technologies are powerful but have a limit: as fine as the adaptation they are capable of making, their impact on transformation depends on the effectiveness of the educational formats offered.

However, these remain superficial and therefore... Not very transformative!

To transform practices, aim for learning transfer

This observation should not surprise us, it is a symptom of the “managerial” approach that dominates in e-learning. Training is seen as a bureaucracy (we talk about learning). Management systems) and for good reason: for years the challenge was to make a large number of content accessible to a large number of learners.

But anyone who works in the field can see it: “pushing content” does not solve the real problems that employees in the field encounter. For the simple and good reason that information is not training and even less transformation.

To enable a transformation of practices, it is necessary to aim for what cognitive sciences call learning transfer, that is to say, the ability of a learner to apply what he has seen in a context different from that of the training... Like his workstation!

Cognitive science research offers 3 ingredients to boost learning transfer:

• Have a high proportion of exercise in your training : we are not talking about reserving a meager 10% of e-learning for multiple choice questions, but about having the learner apply each important concept they discover.
• Propose simulation exercises that are closer and closer to the field: to work on each skill first in isolation and then in interaction with each other. If we take the example of sales, we will have exercises where a salesperson must distinguish the important points from the secondary points of a prospect's message, practice responding to objections, and then simulate an end-to-end exchange with a prospect.
• Give tailor-made feedback: For each action of the learner, he must receive precise, specific, benevolent feedback, which contains, if necessary, an area for improvement.

Each of these points contributes significantly to the transfer of learning, but it was previously difficult to integrate them into training without investing too much time. To propose relevant scenarios, you need both a good mastery of pedagogy AND field expertise. To give tailor-made feedback to each learner, you must take the time to write them down.
‍

Educational AI finally makes the transfer of learning into e-learning possible

But this question of time and expertise is now being solved by generative AI. In fact, in our opinion, this is its greatest promise: Educational AI finally makes it possible to transfer learning to e-learning. To do this, it is necessary to move from a “managerial” vision of training to a “transformation” vision.

This new vision involves rethinking both the formats we want to prioritize and the success metrics of training courses. It is with these biases in mind that we created our Educational AI.

Its operation is simple: it transforms any text, regardless of its initial format, into effective e-learning with numerous applications, tailor-made feedback... Anything that improves learning transfer. It also adapts to your context so that these situations are close to the experiences of your learners: the challenges of the apprentice salespeople we were talking about earlier will not be the same if they are in store or on the phone with a prospect. Our educational AI will script the exercises differently depending on these situations.

“In my 20-year career, I have never seen such a speed of media coverage of content, with such final quality” Boban Milic, director of the CFPC

At the end of the year, our Educational AI will already have 1 year of existence and we have been able to quantify its impact. It makes it possible to transform 150 hours of training in 3 months, into rich e-learning with high-impact activities. In terms of time, it is 10x faster than the market standard for equivalent e-learning.

We are also in the process of developing new AI-based activities for: the corrected activity, which generates tailor-made feedback on business productions and a virtual coach who helps each learner to move on to practice in the field.

So transformation is already within reach: ready to move up a gear?
‍