Meet Alex (AI): Your Super Keen L&D Sidekick (10 minute read)

How Best to Manage Alex

Alex is a curious collaborator, eager to learn and contribute to any piece of work. They are willing to get stuck into any task you ask of them, however, while they are a sponge for information, they are not an L&D expert. They don’t have a great short-term memory, and they take the information they read on face value. For this reason, your management style needs to be catered to maximise the value you get from them.

You might be familiar with the Situational Leadership Model (Hersey & Blanchard, 1969) and adjusting your style of leadership to the situation or person, or you might be familiar with the Skill/Will Matrix (Lansberg, 2003). These can be used as tools to enable leaders to adjust their style based on the skill and will of their team members.

Alex’s L&D Use Cases

Alex can perform lots of roles, some are listed below, but the portfolio of tools available is expanding every month; however, these can be quite niche. I have also noted what type of Gen AI could be used in each case. For example, a chat AI can be used for most of these use cases, like Microsoft Copilot or Chat GPT.

• Ideation (chat AI). Alex is great when you don’t have a clue where to start on a topic or need a nudge for ideas to solve a problem. They can also be used to bounce ideas and provide a critical evaluation based on certain criteria. Try Alex, they may prompt the right neurons to get you moving.
• Persona generator (chat AI). At the scoping phase of your project, you may be looking to use some personas to help and needs analysis and development of a solution to plug a knowledge or skills gap. They could create a generic job description if your organisation doesn’t have any. This can also be used for learning design and assessment design – using the persona to create scenarios, exercises and assessment examples. You could also expand this to target audience analysis, as a part of a learning needs analysis, to help identify and use research to improve the design of learning.
• Learning needs analysis (chat AI). There are some sophisticated tools developed in this area, but they need to be paid for; however, Alex’s large language model would have scraped the internet for potential industry standards and other examples of similar projects to help get you started – quickly! For example, this could be delving deeper and expanding learning outcomes to create learning objectives and key learning points for a specific topic. They can also be used to evaluate plans and products based on project delivery criteria. You can use all of this to save time for your SMEs by providing them with a rough draft analysis instead of starting from a blank page.
• Lesson/assessment planner (chat AI). Give them a format to write a plan, along with learning objectives and detailed requirements, and watch them go and deliver a reasonable first attempt at a lesson or assessment plan. SMEs and learning designers can ‘chuck out’ or adapt their outputs before going into the learning development stage.
• Content creation (scenarios, examples, images, video and narration) - a chat AI like Copilot can do most of these, less video and audio generation. This will require a mixture of AI generative tools. Remember, chat AI can also be used to create prompts for these other types of AI tools. For example, they may provide a half-decent knowledge check to use at the end of a lesson or e-learning product just by using the lesson plan and standards expected of the learning audience. They may also provide a script and video plan based on the same information.
• Initial product pilot (chat AI). Usually, pilots occur in three stages, a concept pilot (a paper-based exercise by the L&D team and SMEs), individual volunteers from the target audience, and then a full pilot with a cohort of the target audience. However, you can usually iron out quite a few kinks just by feeding Alex some assessment criteria and seeing what they find. This could be for a scoping report, assessment strategy, lesson plans, examples, scenarios – to name a few.
• Enhancing accessibility (chat AI or video generation AI). Gone are the days when creating written versions of documents or closed captions on videos took so much time. This is done almost instantly alongside development of content. Script narration is also getting pretty good, even on free platforms. For example, a chat AI could create a script and then you could use Microsoft ClipChamp's text-to-speech function, along with royalty free b-loop music. Combine this with some PowerPoint skills and you could have a quick and cheap way to produce videos. You could also go further and combine this with some of the cheap AI video generation tools like InVideo AI.
• Research and information gathering (chat AI). Alex can conduct rapid research and has the ability to absorb vast amounts of information. They can help identify relevant studies or sources to help you use reasonably up-to-date knowledge and models. 

Coaching and Guiding Alex: Effective Prompts

Creating the right prompt, information, and requirements is critical to getting the best out of Alex. Prompting is a big topic and an art, which will need to be covered in another blog to do it justice; I am experimenting on a concept of a prompt pathway template to get an L&D team from learning outcomes to a fully developed learning solution with minimal time and cost, but let’s consider some basics you need to give Alex:

• Be specific: Clear instructions yield better results. Set parameters and outputs like response length, tone, referencing, expected output formats or standards. 
• Context matters: Provide relevant info to guide Alex: business context, business objectives, target audience, relevant industry bodies or government policy - to name a few. Giving Alex a persona is also effective.
• Show examples: Illustrate what you’re looking for. Alex learns by examples, like us.
• Variety hour: Iterate and try different prompts, be it language or style. Alex loves challenges - like a puzzle-solving prodigy. It is a learning curve. I recommend recording your prompts and the outputs. You can also ask Alex, or one of their other AI colleagues, to critique their work based on a new persona or set of standards.

Navigating Limitations

Alex is far from flawless:

• Bias alert: Large language models inherit biases from the internet. Stay vigilant. Alex might unknowingly favour certain topics or perspectives.
• Quality control: Always review what Alex produces. It’s like proofreading a friend’s essay. Sometimes it’s poetry, other times, it’s gibberish formed from ‘hallucinations’ where they couldn’t find anything, so they made something up.
• Human touch: Alex complements us but doesn’t replace our expertise.

Recommended approaches when working with Alex

Here are some other approaches to consider when working with Alex:

• Critical thinking hats: Encourage questioning Alex’s output. “Is this accurate? Is it fair? Can it be corroborated or assured using an SME?”
• Attribution etiquette: When Alex shines, give credit where due, but decision-makers need to be aware of where the product comes from and what the risks are.
• Plagiarism and Intellectual Property. Remember Alex’s knowledge is from a scrape of the internet. Sometimes what they provide isn’t theirs to give, sometimes you may just need to identify and reference originators – or ask for permission.
• Experiment. Try out new things, Alex may surprise you. The road may be bumpy but embrace the failures and keep learning – you will be rewarded.

Alex (Copilot) helped me plan this post and produce some of the content. It greatly reduced time and effort, hopefully creating something more engaging than just my words, for free.

If you haven't tried to use Gen AI for L&D yet, I hope this post creates a spark; experiment and give Alex a go. If you have, it would be great to hear your thoughts and use cases in the comments below.