Is this the future of Digital Education?

When learning science meets AI

Author

Ross Stevenson
February 11th, 2025

📨 Subscribe | 🤖 AI Training | 🚀 Courses

Today’s Thoughts ☠️

Hey there 👋,

We’re back! And… the sun has reappeared in my dark corner of the world.

Some might call this a win-win.

We both know the world is littered with AI tools. Even as I type this, my mac is annoying me about trying out “Apple Intelligence”, aka Apple’s jump into AI-powered features.

I care not for them because they have little impact on my work.

That’s the big challenge, right? What is actually useful for our work.

After weeks of testing in my lab (or normal people would say home office), I’ve found a tool worth sharing with you. This one comes direct from Google. They’ve decided to fuse learning science principles with an AI model to move into the lucrative education space.

Today, we’re exploring this tool to see what it does, how it works and uncover it’s applications for the future of education (+ L&D).

Get your tea or beverage of choice ready, 🍵.

We've got lots to discuss!

 TL;DR 📰

  • Google wants to be your personalised learning assistant

  • What happens when learning science meets AI

  • How you can use Google’s experimental tools for L&D

 THE BIG THOUGHT 👀

Is This The Future of Education with AI?

will ferrell anchorman GIF

I know the feeling

The big G or Google, as normal people would say, has entered the AI for education space with a new model called LearnLM. 

Original name, I know. 

But before we roll our eyes, let’s take a look at what this actually means for the future of education (and by association, L&D).

You’ve probably heard of OpenAI’s ChatGPT, Anthropic’s Claude, and Google’s Gemini. Well, LearnLM is part of the Gemini family, but with a special twist - it’s been fine-tuned specifically for education.

For the non-techies, fine-tuning means taking an AI model and giving it additional training in a particular field. In this case, Google has trained a version of Gemini on learning science principles to create an AI that supports personalised and adaptive learning. 

So, while this isn’t built specifically for workplace learning, it’s a peek into the future of education across our society. 

Saying that, in the next few weeks I’ll be sharing new AI features that will impact workplace learning, and probably, reshape what we consider as performance support ‘in the flow of work’.

I know it all moves so fast, but I’m going to do my best to keep bringing you the most useful stuff.

For now, let’s discover how our children are going to tackle homework, learn new concepts, and shout at us, “Well, AI says you’re wrong!” when we question how they arrived at an answer.

Learning Science meets AI

According to Google, LearnLM incorporates several key learning science principles:

  • Active learning – Encouraging practice and effort by providing timely feedback.

  • Cognitive load management – Structuring information in a way that prevents overload (I’ll believe it when I see it).

  • Adaptability – Adjusting responses based on user interaction (vague, but okay).

  • Curiosity stimulation – Keeping users engaged by sparking interest and motivation.

  • Metacognition support – Helping individuals reflect on their progress and plan next steps, essentially nudging them to think critically about how they learn.

Sounds promising.

I’m not a learning scientist, btw, so this could all be very good or just very good marketing.

This model is only in the research phase for the general public, so we can only use it in Google’s AI studio. The goal, right now, is to test this across traditional education establishments.

However, it is being used across a lot of Google’s existing products today, just in a very low-key way.

🥷Shadow AI at its best

Google has already integrated LearnLM into several of its tools, including:

  • Google Search

  • YouTube

  • Google Classroom (Side note: People actually use Google Classroom? 🫤) 

You’ll also find it in something called Gems, Google’s version of custom AI assistants.

One notable example is Learning Coach, an AI-powered assistant designed to support users by offering explanations, answering questions, and providing guided learning experiences.

If you use Gemini, you can access it now.

And then there’s YouTube, where Google has rolled out an AI feature that lets users ask questions and get explanations while watching educational videos.

So, next time you’re deep in a ‘How to fix your sink’ rabbit hole, don’t be surprised if AI starts breaking it down for you.

Oh, if you’re wondering about the ‘Shadow AI’ reference. This is a term I use for when people are using AI in existing products but are totally unaware of it.

Two experimental tools for learning

Beyond the familiar platforms, Google is working on two new experimental tools powered by LearnLM:

  1. Illuminate – This is a stand-alone version of the ‘podcasty’ feature within NotebookLM that converts research articles into audio-based conversations. If you’re familiar with NotebookLM, you’ll know it’s a useful research and analysis tool. Illuminate allows you to create those popular audio overviews with any content.

  2. Learn About – A new Google Labs experiment that aims to bring learning science, high-quality content, and AI-powered discussions into one experience. Think of it as a Google Search that’s fine-tuned for educational exploration.

I signed up for both waitlists months ago and still haven’t gained access (Google, if you’re reading this - help me out!).

L&D use cases for LearnLM

So how can you put LearnLM to work? 

Here are a few ideas:

  1. Experience Design: As LearnLM is built on learning science principles, we can use it to help both uncover blindspots and improve ideas for digital and physical learning experiences. Just tell LearnLM your ideas and what you want to achieve, and have a chat.

  2. Explain it like I’m 5: The built-in “Concept Tutor” is an easy way to get participants to not only understand complex concepts but also promote their critical thinking in the process. In the demo video below, you’ll see me work with LearnLM in this way, and I must say, I was impressed with how much it was challenging me.

They’ll be more, no doubt. So get experimenting!

Accessing LearnLM

If this sounds interesting and you want to test LearnLM yourself, you can find it in Google AI Studio. 

This platform lets you experiment with different AI models, including LearnLM, and even build custom tools with Google’s API. One of my other favourite uses of the AI studio is uploading videos for it to analyse, summarise and find the best stuff for my context.

The best part? It’s completely free - you just need a Google account. 

You can learn how to use it and see some use cases by joining me in the video below ↓

📝 Final thoughts

Google is making moves in the AI and education space with LearnLM. 

Not surprising as the industry is worth billions of dollars.

We know it’s built on solid learning science principles, but as always, we need to see how it actually performs in real-world applications.

While still experimental, it’s an interesting tool to explore, especially for us in education and L&D.

Whether it becomes the AI learning assistant of the future remains to be seen. But hey, it’s free to try, so why not play around with it?

Let me know if you give it a spin.

→ If you’ve found this helpful, please consider sharing it wherever you hang out online, tag me in and share your use cases.

👀 ICYMI (In case you missed it!)

Till next time, you stay classy, learning friend!

PS… If you’re enjoying the newsletter, will you take 4 seconds to forward this edition to a friend? It goes a long way in helping me grow the newsletter (and cut through our industry BS with actionable insights).

And one more thing, I’d love your input on how to make the newsletter even more useful for you!

So please leave a comment with:

  1. Ideas you’d like covered in future editions

  2. Your biggest takeaway from this edition

I read & reply to every single one of them!

 TECH THOUGHTS 💾

The Problem-Solving Framework I Use To Build Solutions For Real Problems

Although much of the business world is consumed with becoming AI-ified.

You must not forget the power and need of your human thinking. Everything I build, write and publish goes through a vast amount of human driven thinking frameworks (I hope that makes sense).

AI is good at a bunch of stuff, but so are you.

Here’s something I hope helps you create great products by asking better questions.

P.S. Wanna build your L&D advantage?

Here’s a few ways I can help:

🙋‍♀️ Want to reach over 4,500 L&D pros? Become a Newsletter sponsor in 2025

Reply

or to participate.