Stella Collins provides reassurance and guidance on those looking to use artificial intelligence in their learning design and delivery

Neuroscience and AI both attract attention in L&D – some good and some bad. You might think AI is the newer kid on the block, but it’s been around longer than you think. Both disciplines have borrowed and learned from each other for years.

Why does this matter to us in L&D? Because our primary role is to improve performance by helping people to learn and change their behaviours in response to organisational needs – that poses challenges that both AI and neuroscience are already helping to solve.

Three pressing issues in learning design and delivery

In a recent industry survey, three of the most consistent challenges for people who design and deliver learning were revealed. Which ones do you recognise in your role?

Lack of time

Learning designers need to work at speed and to tight deadlines to meet business needs but current design tools and processes often can’t keep up with the pace of change. Delivery challenges happen when learners and their managers can’t commit sufficient time for effective behaviour change.

Deciding objectives and aligning with business requirements

You know this one – a senior manager comes up and asks you for ‘a course on negotiation’ and expects you to wave your magic wand and solve their challenge. You do your best to analyse the problem, but stakeholders find it difficult to answer the question ‘What do you need your people to do and why aren’t they doing it now?’ After all, they might be part of the difficulty. So, we end up creating a course to feed their short-term needs or simply because they have more authority.

Company or industry-specific requirements

All too often subject matter experts suffer from the curse of the expert and deliver too much information. It’s not their fault – learning isn’t their specialist area. Once they’ve delivered ‘the knowledge’ they think their job is done and don’t have time to validate, support or evaluate. Or they simply create ‘training’ themselves – with yet more slides!

The survey found that people were looking to generate quality, impactful learning programmes faster.  

The problems are ones we’ve all faced for years. So will AI solve all of these problems, and kick L&D further from the corporate attention span?

Build the right prompts into your system and AI can generate the content plus the additional elements people really need for behaviour change

Fortunately, AI isn’t that smart (or at least not yet). Your human skills are still vital to add the ‘magic’ that makes learning enjoyable, creates long-term impact and returns the investment of time and energy.

AI can help to solve problems like these only if you’ve got strong learning science baked into your prompts. Otherwise, you’re in danger of fuelling cognitive overload with more content, produced much faster.

What can AI and the science of learning do together?

AI can create content faster of course, but this won’t solve the problems outlined above because tools like ChatGPT tend to create generic content. After all, they are just looking to see what is the most likely pairing of the next words they find on the internet – if the words ‘learning’ and ‘styles’ are frequently linked AI has no idea that it’s not a useful concept. The prompts you enter define what comes out so ‘Garbage In = Garbage Out (GIGO).

If you’ve already experimented with AI tools, you know the frustration when the tool returns exactly what you asked for – but not what you intended. Good prompt engineers are like gold dust but good learning designers remain essential.

Use learning science to guide the AI

What people need are rich learning opportunities where they get guidance, can experiment, practice and transfer their learning into the workplace with appropriate support and validation. Build the right prompts into your system and AI can generate the content plus the additional elements people really need for behaviour change.

Use AI to support learning transfer

As learning specialists you know that content alone isn’t sufficient for behaviour change. AI can be trained to create and deliver other resources like spaced repetition questions to beat the forgetting curve.

Iterate quickly

AI makes laborious processes like content and question creation faster so you can iterate quickly. Once the process is faster you can optimise and update learning journeys more often to boost impact and stay relevant to the business context and needs.

Align stakeholders on outcomes faster

It’s much easier to align with stakeholders when you can all see examples of expected behaviours and you can prototype activities. With the right prompts, AI can help you get these nailed faster. And, once clear on objectives you can prototype programmes aligned with these and share them with stakeholders early on. It’s easier to iterate and deliver value when you get learning design right from the start.

Spend more time evaluating and acting on meaningful learning data

Good technology helps you collect data, good learning science means you know what data you need and AI helps you analyse that data (and frees up your time to do so!) to create insights.

Empower non-specialists to design great learning

With systems that combine the benefits of AI with good learning science and built-in transfer objectives you can support Subject Matter Experts (SMEs) to design learning interventions that avoid the curse of the expert, reduce cognitive overload and focus on behaviour change rather than simply dumping information. More time for you to focus on your specialist areas.

Increase personalisation

AI and neuroscience combined can personalise learning. People no longer have to follow every step of a training programme whether they want it or not. Instead they can prove to the smart system that they already have certain knowledge and skills and can skip to the parts that are relevant for them.

Does AI render learning design dead?

Of course not! When AI frees up the time for you and neuroscience provides the solid framework you need for quality, you can:

• spend more time with your ‘customers’
• ditch the happy sheets and get valuable insights from the data
• influence organisational levels of transfer
• up-skill yourself and your team
• dive deeper into the technology that supports performance.

What are you waiting for? Combine AI and neuroscience to achieve quality at speed!

Stella Collins is co-founder and chief learning officer at Stellar Labs