AI has exploded onto the scene. It can write, summarise, translate, generate images and fix code. Note-taking apps are adding AI features at speed.
So it's a fair question: if AI can do all that, why bother with a note-taking system at all?
I've spent a lot of time on this question, partly because I'm building a note-taking app myself. Here's where I've landed.
The race to the middle
I think AI creates a "race to the middle". Subpar creators who use it will improve. Above-average creators will use it to speed up and settle for average. The people who stand out will be the true experts who keep their craft and don't hand it over wholesale.
The threat isn't that AI can think better than you. There are plenty of humans who already can. The threat is volume. It's the equivalent of the number of competitors in your field multiplying a thousand-fold overnight.
In that world, what makes you valuable? Depth. Your own understanding, your own connections and your own point of view.
You can't outsource deep thinking
Imagine you copied somebody else's notes onto your computer. Would you have learned anything? Of course not. Using AI to create notes, reason about them and do the thinking is no different.
Think about what happens when you go for a walk after working on a problem. Your brain keeps chewing it over in the background. It compares, reframes and connects. But it can only do that if you've actively engaged with the problem first. If the AI did the work, there's nothing for your subconscious to work on.
If knowledge management were a video game, letting AI do the thinking is like watching an AI play the game for you. Not much fun.
Where AI genuinely helps
I'm not anti-AI. I've used AI-generated images throughout my book because my goal was to teach, not to produce artwork. It's a tool that served a purpose.
The test I use: AI should help you with your goals, not replace them.
- If your writing is holding you back from sharing deep ideas, AI can be a huge benefit.
- If you want to condense long articles and your reading skills are a barrier, AI can help.
- If you're stuck, tell it what you currently understand and ask it questions. Where are the gaps? What should I explore next?
But if you want to be an excellent writer, AI can't do the writing for you. If you want to be an excellent reader, it can't do the reading for you. And if you want to think deeply about a subject, it can't do that either.
It's rather like the preprinted messages in greeting cards. They made it accessible for people who couldn't find the words. That's great, until you want to say something truly personal.
Auto-linking: a warning
Many apps now offer to link your notes for you. It's tempting, but I'd be careful. Links created out of context can be misleading, and, more importantly, the act of linking is where the thinking happens. Delegate that and you remove yourself from the process. You become less aware of what's in your own system, and less able to spot what's new.
What I think comes next
I believe there will be compelling uses: an assistant that challenges your ideas, finds references, and fact-checks; a tutor that explains, tests and personalises your revision. The best versions will support your deep thinking, but not influence it.
A practical rule
Next time you reach for AI in your note-taking, ask yourself one question: am I using this to think better, or to avoid thinking? If it's the first, carry on. If it's the second, close the window and write the note yourself.
There's more to the story
This is the shorter version of one chapter. In Atomic Note-Taking, I go further into where AI fits in knowledge management, what I think the future of note-taking looks like, and how to choose a note-taking app for your own style: local, cloud or hybrid, Markdown or WYSIWYG, long notes or short.
