A Learning Team’s Guide to AI

Prelude:?

Thanks for joining me this week, as we dive head first into AI and it's usability for learning teams. Subscribe to my newsletter for bi-weekly articles on how technology can help learning teams.

To say that AI is a hot topic might be the understatement of the century.?

It is no different for learning teams, who are eager to start using the benefits of AI to improve their operations and function as a whole.?

I ran a poll on LinkedIn amongst learning teams, where AI was the single most interesting topic for them with over 58% of the votes.

This article titled “A Learning Team’s Guide to AI” will be a resource for teams to understand the current and future use of AI. It is not a technical dictionary, nor an explanation of AI to developers. It is written by a learning nerd (myself) for learning nerds to understand how they can use AI to improve what they’re doing.?

This article has three parts to it,

1. AI is NOT a learning strategy, it's an Enabler
2. Four key use cases of AI for learning teams
3. Where are we going... My observations on the future of AI

Part I: AI is NOT a Learning Strategy, it’s an Enabler?

AI in learning has always been a hot topic, but a crazy HOT topic last two years.?

This has been accelerated by technological advancements like ChatGPT and L&D looking to embrace AI in new ways. The Large Language Models (LLMs) are getting better, and learning teams are looking for new ways to use AI to improve their function.?

Back in May of this year, we ran a session at Learning Tech in London where hundreds of people came to listen about use cases of AI in learning. Last week, we ran a webinar online with Mike Byrne and Fernando Martin where we had over 250 people joining to listen in about AI.?

AI is the hottest topic for L&D, but beware, AI doesn't replace a learning strategy.

I’m a big fan of Ross Stevenson and his Steal These Thoughts! Newsletter. I read it weekly and he inspired me to start my own newsletter. Last week, Ross talked about how

“AI is an enabler, or a way to reach your strategy faster. It's not the strategy itself”.

Before learning teams dive head first into AI, Ross said they should do the following:

• ? Build a performance-focused learning strategy
• ? Identify the skills the organisation needs to succeed
• ? Maximise your existing learning tech to support and grow employee performance
• ? Develop an onboarding experience to help people be at their best

??This led me to think about an analogy of a marathon where you need to get from A to B. To do this, you need to have a plan for how to run and not burn out (a strategy).

?? AI is a good pair of shoes or some protein-filled food, it’s not the race itself.

?? You need a strategy and a foundation before you start using AI tools.

?? AI tools are great and can be extremely useful (like a good pair of running shoes), but they do not replace the foundational first which comes first.

Your learning strategy should not focus on AI, it should focus on your end goal, increasing business value through your people:

• ?? Attracting and retaining talent.?
• ?? Developing your people capabilities.?
• ?? Creating a values-based culture.?
• ?? Building an employer brand.?
• ?? Motivating and engaging employees.

? AI can help you get to these outcomes quicker, but it's not the end goal.

???? Make AI your friend, a helping hand and an enabler of strategy.

Part II - Four Use Cases of AI in Learning:?

For the purpose of this article, I studied AI research from a technical perspective and equally important, spoke to many learning folks about their current and future use of AI.?

Based on the above, I tried to dissect all the noise and summarise the use of AI in learning into four key pillars (there are more but these are the most pervasive):?

• Use case #1: Content Creation/translation?

• Use case #2: Chatbot/virtual assistant
• Use case #3: AI-powered Deep/Semantic Search
• Use case #4: Skills Intelligence/Tagging

I ran a second poll on these four topics and got votes from learning folks on which of these was the most important topic for them. I ranked the use cases below from one to four based on the highest response answers (#1 getting the highest response).?

Use case #1: Content Creation

Traditionally, authoring of content is a skillset reserved to central teams. Whilst the quality is often high, the speed of creation makes it a bottleneck for L&D operations.

This leaves subject matter experts feeling removed, and reliant on learning teams to get their expertise to the masses.

?? Enter AI - the game-changer for content creation ??

By using AI, you can transform an input source (PDF, free text, links) into learning videos in minutes.

You can auto-translate it into a plethora of languages so users can consume it natively.

You can empower your SMEs to own the creation process and be an active part of your learning program.

When empowered, they will share knowledge in your learning communities, forums and build a whole new learning culture.

Thanks to AI, organisations can save time and money by using this technology that has been trained to create effective and engaging content based on modern learning techniques.

Use case #2: Virtual coaches/chatbots

AI-powered virtual coaches are proactive virtual assistants for users on your platform. They can guide learners through learning activities and work within the learning platform to recommend content, monitor progress, answer content-related questions, and send push notifications related to content or deadlines.

Essentially, a virtual learning coach is like a chatbot, similar to the ones used for customer support on various websites and platforms.

Virtual coaches help learners and make the learning process more streamlined. That’s because, unlike human trainers and admins, they’re available 24/7. A virtual coach can answer many basic questions about learning content thus saving time for learners and admins alike.

Use case #3: Deep/semantic search and auto-tagging

On an AI-powered learning platform, AI analyses both formal and informal learning assets and improves their discoverability through search. It’s like the advanced search engines you see on Google or Amazon. When someone shares a new learning asset or creates new training material, the AI synthesises the information in the asset to produce the most relevant search results.

During this process, AI identifies key phrases and creates tags for assets automatically to make learning content easier to find. Auto-Tagging function identifies key phrases from uploaded learning content and assigns tags automatically, cutting down on the time-consuming task of manual tagging. This helps admins with categorising content properly and also makes the Deep Search function easier to use for learners.

With Deep Search, if you share an interesting article in your LMS and your coworker needs to reference it a few days later, they can just search for it and find it in seconds. This prevents anyone from needing to dig through a bunch of irrelevant search results.

Use case #4: Skills Intelligence/mapping - “Skills in the flow of work”

To successfully implement your skills strategy, you need to define your “source of truth”, or in other words, which system are we going to build our taxonomy from??

Start with the end goal - and build your requirements from there. Then look at capabilities of current systems and where you want to go in the future.

AI can help you as a learning team by taking an existing ontology and mapping it to content that then gets surfaced to learners inside of the learning system. The ontology can be built inside of your learning system (Docebo), from your HRIS (i.e Workday), Talent Marketplace (Fuel50/Gloat) or Skills Intelligence Tool (Eightfold/Techwolf etc).?

AI is interesting here as it allows you to use a specific ontology (inside or outside a learning system) and index/tag content with that specific skill-set. For a learner, they can then find the skills they want to develop and content automatically inside of their learning system. I call this “skills in the flow of work” and it takes the concept of learning in the flow of work but it adds the skills mapping/intelligence on top of this to make it relevant for a user’s upskilling or reskilling journey.?

Part III - Where are we going? My thoughts…?

Observation 1:?

First off, I expect the four pillars below to be incrementally improved using AI. This can go quickly, as the technology is developing at lightning speed.??

• Content creation will be quicker, smarter and more accurate than in its current state (as the model trains and retrains itself).?

• Skills will be smarter and more accessible across enterprise systems and skills will often be added from a “source of truth - SOT” into an external system delivering upon the AI-based skills intelligence from the SOT. Skills in the flow of work will become the norm and the demands from HR/Talent/Learning teams will be higher from vendors/providers.?

• Search will be smarter, and our demands for search will be higher. For learning teams, this is critical as learners’ needs and demands for accuracy and speed will increase. If learners can’t find personalised content recommendations quickly, they will disengage.?

• Virtual assistants/chatbots will become smarter and more personalised. It’ll become a learning coach and learners will demand more from the chatbot. The experience people have with ChatGPT will be the expectation from the learning chatbots, and it’s critical for vendors that the AI engine/brain is intelligent enough to provide a personalised and curated learning experience in a quick manner.?

Observation 2:?

AI ethics will become one of the biggest debates we have seen (and learning teams can’t escape it). What data does companies use to build their Large Language Model (LLM)? Has vendors built their own or used publicly facing LLMs to build their AI intelligence/brain? These are questions that vendors should ask as it will be very influential in your decision-making process.?

My take here is that vendors who have proprietary LLMs will massively benefit as I think there will be a backlash against publicly-built LLMs and more specifically LLMs built from US companies (especially in the EU). We are already seeing huge debates about cloud hosting, GDPR, and AI is next and it’ll be an interesting debacle to observe. I wouldn’t be surprised if certain LLMs will be banned, especially amongst regulatory industries or public-sector organisations.?

Organisations who see this as a risk will consider not adding AI features into their learning project, or will favour companies who have built their own LLM/AI Brain.?

Observation 3: Stay curious, aware, but also sceptical?

Learning teams who have built a strong AI-toolbox will be ahead of those who haven’t.

This one may seem obvious but if you’re not aware of the massive benefits it can have for your learning team, how do you know how to implement it??

As shown in this article, AI can enhance a variety of tasks but the hype can also be exaggerated. It’s your job to cut through the noise and decide what parts of AI you wish to incorporate into your learning strategy.?

This will vary widely if you are training employees, partners or customers, as well as what industry you are in.?

My take - stay curious, aware but also apply a healthy bit of scepticism when it comes to AI.?

I am exploring AI and the possibilities here right with you, staying curious, aware but also applying a healthy bit of scepticism.?

What excites me tremendously about AI is the speed in which the four categories above are improving.?

I’m fortunate enough to work for a learning vendor investing heavily in AI and I can see the incremental improvements day-by-day. Our acquisition of Edugo was like rocket fuel for our roadmap of generative AI features for learning.?

I will update this article once we are ready to spill the beans on all the amazing stuff we are doing in terms of AI.?

AI will transform learning as we know it and make it better for everyone. The question for you as a learning team is how will you use it to transform your organisation and your learning experience you provide.??

Final thoughts:?

This article required more time than usual, from research, speaking to learning practitioners and trying to take a very hot topic and making it personalised for learning teams.???

I am curious to hear your thoughts,

• Are you using AI in other ways than I have shared above??
• What excites you about AI??
• What scares you??
• How do you see AI ethics playing a big role in the future of AI adoption??
• Are you curious, aware or sceptical about AI and its use in L&D??

I am like you, a learning nerd trying to dissect it all and help improve learning experiences - for clients and organisations all over the world.?

Stay curious ??,?

Harald