Notebook LM: The AI Assistant That Makes CPD Actually Work
The Study Partner I Wish I Had
When I transitioned from delivering road and rail projects to marine infrastructure, I knew I had a steep learning curve ahead.
There’s no crash course for this kind of shift — especially when you're not a grad anymore. As the project manager for a $100M+ wharf rebuild, I couldn’t afford to wing it. I had to ramp up, fast.
Like most engineers, I learned on the job. I asked my boss dumb questions, and even dumber questions to AI. I even clicked through dense PDFs, watched numerous videos and hoped things would stick.
But one tool actually helped things click: Notebook LM by Google.
🧠 What is Notebook LM?
Notebook LM is an AI-powered research assistant. You can upload a bunch of sources — PDFs, blog posts, videos, lecture notes — and it helps you understand, summarise, and navigate that content.
It’s part note-taker, part tutor, part podcast host.
What sets it apart from other LLMs is that it stays grounded in your material. No hallucinations. Just synthesis, structure, and sharp answers from the stuff you give it.
I used it to break down PIANC’s Design Principles for Dry Bulk Marine Terminals — a dense guideline. Normally, digesting something like this would take days. With Notebook LM, I had a grasp of the core ideas in hours — and remembered more of it.
🚀 Why is this a game changer?
Imagine you’re learning a new engineering concept. You’ve found:
3 technical videos
4 blog posts
A textbook chapter
In the past, you’d jump between tabs, take messy notes, and probably forget most of what you read. Notebook LM:
Summarises the core ideas
Links related concepts across sources
Prepares quizzes to reinforce learning
And (my favourite) can generate a podcast version of the content
Yes — it reads your sources back to you. I listened to marine design principles on my daily commute and racked up CPD hours.
Side note: Listening to it was equal parts impressive and unsettling. Have a listen and let me know what you think (forgive the audio watermarks). I think it’s funny how they generate almost forced uhms into the conversation to mimic human conversations.
🎧 Me, on a train, headphones on, listening to the driest fake podcast ever.
(Prompt: Generate an image of me on a train wearing headphones. I have a thought bubble which features a dry bulk commercial ship unloading at a dock. The style is playful and cartoonish, in a studio Ghibli style.)
⚠️ What Notebook LM isn’t
Notebook LM is built for focused learning, not free-form chatting. Unlike general AI tools, it only answers questions based on the documents you upload — no guessing, no filler.
So why not just use ChatGPT?
Because even if a chatbot was trained on similar material, it won’t reliably reflect the exact version or nuance in your sources. Notebook LM stays within the confines of your source material, which makes it far better for studying standards, guidelines, or technical references without hallucinations.
Use it like a research assistant: great for absorbing info, revising, and connecting concepts — not for brainstorming or problem-solving.
📚 Why this matters for CPD
Whether you’re a grad or a senior engineer, the problem’s the same: you need to learn fast, and you need to retain what matters.
Notebook LM won’t replace deep reading or hands-on experience — but it compresses the learning curve and gives you tools to stay engaged, not just absorb.
And that’s key: it helps you stay in control of your learning. Too many AI tools make it easy to delegate the hard thinking. Notebook LM rewards curiosity — not shortcuts.
With Notebook LM’s quiz feature, I was able to seriously reinforce my understanding of the key concepts and retain the information
If you’re managing your own development or diving into a new field, this is the AI assistant I wish I had years ago.
🔗 Try it out
https://notebooklm.google/
(It’s free and browser-based)
💬 What do you think?
Tried Notebook LM? Got a learning workflow you swear by?
Reply or comment — I’d love to feature how other engineers are using AI in a future issue.
This saved me so much time