Shifting from "Writing-to-Discover" to "Organizing-to-Discover" (then Write)

Cool Workflows
1

I’ve always been a “writing-to-discover” kind of person—writing messy first drafts, then revising heavily until I figure out what I’m actually trying to say. It works, but it’s exhausting and slow.

Recently, I’ve started experimenting with a different approach, inspired by Lex.page (an AI writing tool I’ve been playing with):

Instead of writing-to-discover, it’s more like organizing-to-discover, followed by writing.

Here’s what that means in practice:

  1. I start by throwing all my notes, ideas, and research into one place.
  2. Then, rather than trying to write a messy draft right away, I use Lex’s AI features to organize my raw materials, identify connections, and clarify the core idea upfront. I also use ChatGPT and its friends, too. It’s good to get different perspectives, and they often see things I can’t (forest v.s. trees kinda thing).
  3. Once I’ve got a solid idea of what my key points are, I start writing my first draft.

It sounds simple - and probably obvious, too! - but the mental shift has been huge. I’m finding that once I’ve organized first, the writing itself flows far more easily—clearer, faster, and with way less mental friction.

And, yes, I often ask AI to write the first draft, and then I edit the ar5e of it.

(I brace myself for the moral outrage, but, I write for a living, and my writing helps people, and I think it be unethical to not use AI to write more and help more people. I’m sharing this because I think a lot of people are missing out on using AI, because they assume it’s just going to produce more slop, and they’ve not yet used to write in a good, decent, and honorouable way.).

4 Likes
2

BTW: I wrote this post, above, using this approach, and a mix of lex.page, my fingertips and voice, and chatgpt. I was inspired by this article, written by the people at lex: The Lex Method — Lex

You don’t need to use lex.page to work this way, though I find it very helpful for working this way.

3

What did you throw into Lex that it organized to write this post?

(I’m neither enthused nor outraged. As you describe it, It reminds me of having an assistant organize clippings and archive drawers, which also had tradeoffs.)

4

Funnily enough, I chucked their article Iits in the first reply above) into ChatGPT. It felt rude to feed their content into their app, weirdly cannibalistic. I think that’s how mad cow disease got started. So I used ChatGPT, then cleaned up its output using Lex (which is an editor).

Just before that I started writing an article about how aggressive / impossible deadlines often spark amazing innovations, and I kicked off in lex. I used Superwhisper to brain dump about the world fair in the 1850s and how they announced an impossible date, and that lead them to creating the crystal palace and the amazing engineering necessary to build it. Current methods wouldn’t work.

I then pasted some research about he crystal palace that I’d gotten from perplexity into lex, and I told lex’s chat that this was a brain dump, and could it help me find the key points and organise them.

It said this

Looking at your brain dump, there’s a fascinating core idea emerging: The Crystal Palace wasn’t just built quickly - the aggressive deadline actually forced a complete reimagining of what was possible in construction and architecture.
Here are some powerful angles you could develop:

  1. The Counterintuitive Truth: Most people think impossible deadlines lead to corner-cutting and mediocrity. The Crystal Palace shows the opposite - the deadline forced them to innovate beyond what anyone thought possible.
  2. The Hidden Pattern: The real story isn’t about meeting a deadline - it’s about how constraints killed conventional thinking. They couldn’t use traditional methods (brick and mortar), so they had to invent new ones.
  3. The Leadership Lesson: The courage to set and stick to an “impossible” deadline. Most leaders today would have extended the timeline when they saw no site, no design, and no funding.
  4. The Innovation Trigger: The specific ways the deadline pressure led to innovations:
  • Prefabrication
  • Modular design
  • New assembly methods
  • Novel project management approaches
    What I find most compelling is how this challenges modern project management wisdom. Today, we’re told to be “realistic” with deadlines, to add buffer time, to avoid pressure. The Crystal Palace suggests we might be doing it wrong.

And I just loved this (below, its point 2) because that was what I was trying to say all along but I didn’t have those words and it would have taken me hours to find them:

The real story isn’t about meeting a deadline - it’s about how constraints killed conventional thinking. They couldn’t use traditional methods (brick and mortar), so they had to invent new ones.

I then went to ChatGPT, pasted in Lex’s summary and the Wikipedia page on the crystal palace, and asked it to pluck out the facts I needed to tell this story, and then to draft an article.

The article is amazing, but dull, so I will use lex’s editor to clean it up later on (it’s nice to type in, plus it does a whole lot of Grammarly stuff, but better, I think).

That was the process, so far.

I’ve been thinking about the crystal palace story and writing an article about it, and this got me a load of traction very quickly.

So, it’s done a lot more heavy lifting for me than an assistant could.

2 Likes
5

This is interesting… thank you. I too have switched methodology, or at least become more efficient, effectively because I now have a research assistant and writing collaborator. I’m still the boss, and it’s still my words, but I have a tool to rapidly unblock “stuckness” and suggest illustrations which I can make my own. The point 2 summary you give is a great example of drawing out a central concept. AI seems particularly good at that.

It’s also great for repurposing: I rapidly made a 90 second video script from a 15 minute talk.

I would like to write a lot more for the same reasons as you, and I might try your/the Lex approach for a while. I write in Obsidian using the Smart Composer plugin and OpenRouter and it works pretty well.

1 Like
6

Ooooooooh. Smart composer looks interesting. I just tried it, and I will tinker some more. I’d rather type in obsidian, than lex, but every time I try to use it … it doesn’t stick. Don’t know why.

Btw: I just wrote something with lex, and then I got fed up, and told its chat AI, “Has this got really boring all of a sudden?’, which is how I felt, and below is its reply. Take a look at the bit in bold. I didn’t expect it to answer my question like that. I left it for a bit, and it was correct.

Let me be direct: Your piece isn’t boring - it’s thought-provoking and well-structured. You’ve taken a powerful historical reference and connected it to a relevant business challenge. The evolution from literal “deadline” to modern project management frames the discussion perfectly.

What I particularly like:

  • The stark opening with the Civil War reference grabs attention
  • The clear contrast between past and present
  • Your concept of “collaborative urgency” is compelling
  • The practical middle ground you advocate for

The piece maintains engagement throughout by mixing historical context, modern application, and actionable insights. It reads like an experienced consultant sharing hard-won wisdom, not a dry management lecture.
If you’re feeling it’s become boring, that might be from looking at it too long (a common writer’s experience) rather than any actual issue with the content. Your audience will be coming to this fresh, and the message is both clear and impactful.**

Writing with AI can be fascinating and invigorating at times!

1 Like
7

Gemini told me that a recent writing idea was a bad one and my approach didn’t hold up. I was annoyed… but it was right :slight_smile:

Just like having an assistant, everything AI suggests needs to be evaluated, but it does provide insights which are impossible if you’re working on your own.

8

That’s so true! I’ve found that with human editors too. Sometimes they’ll make a suggestion and I’ll instinctively go grrrrrr and want to reject it, but it turns out they’re right. Occassionally, I’ll stick with what I wrote because even though they’re correctly, I decide thst my words are better.

One of my editors once suggested I remove the two hidden rude jokes in my business novel, Rolling Rocks Downhill.

I’d only written one, but she pointed out a second accidental joke that was absolutely filthy. It was so well hidden I didn’t even know it was there! I deleted that one, but I kept the deliberate one, and apart from my editor… Only two or three people have ever mentioned it, which is disappointing :slight_smile:

1 Like