The Great Scientific Rewiring

Friends,

Look, I've been thinking about something that's been driving me slightly mad lately: we're all still trying to write instruction manuals in a world that's moved beyond them.

Let me explain.

Last week, I was struggling to explain to a friend how to capture their writing principles in their email. I caught myself doing what we've always done – trying to break it down into rules, steps, guidelines. "Use this tone here, avoid these words there..." You know the drill.

And then it hit me: I was solving a 2025 problem with a 1995 mindset.

Here's what I mean: For basically all of modern history, we've operated under this beautiful illusion that progress means breaking things down into smaller and smaller pieces. 

Want to build something complex? -> Write better instructions. 

Want to scale something? -> Create better processes. 

Want to teach something? -> Make a better manual.

It's the Enlightenment's greatest hit: if we can explain it clearly enough, we can control it.

But something fascinating is happening. Our machines are teaching us that maybe – just maybe – we've been thinking about this all wrong.

I was working on a building a Youtube thumbnail editing powered by a model of my own image, and I caught myself doing something different. Instead of writing out very detailed specs like I would have been advised to do 3 or 4 years ago, I found myself collecting examples:

Screenshots of interfaces I loved. 

Snippets of copy that hit right. 

User interactions that felt magical.

It was less like writing a recipe and more like creating a mood board. 

Less like programming and more like... gardening?

And it feels like its happening everywhere:

The Death of "Just Write Better Instructions"

You know what's funny? We've spent centuries trying to reduce language to rulebooks. (Shoutout to my high school English teachers who made me memorise when to use "whom.") But here's the thing: the latest models didn't learn by memorising grammar rules although that would probably make them very very good at wrting. They learned through recognising patterns.

They developed what we might call "intuition" – that weird thing where you know something's right but can't exactly explain why. You know "the big red ball" sounds right but "the red big ball" sounds wrong, even if you can't quote the actual rule.

From Crafting to Cultivating

I've started thinking about this shift like moving from sculpture to gardening in the words of Dan Shipper. Bear with me here:

A sculptor works through control – every chip of the marble is intentional, every detail explicitly crafted. That's how we've always built software: line by line, function by function, each piece carefully carved out according to our exact specifications.

A gardener? They work differently. They create conditions for growth. They can't control exactly how a plant will develop, but they can create an environment where it's likely to thrive.

What This Means For All of Us

Here's where this gets interesting for anyone building things right now:

1. The future belongs to pattern collectors, not just rule writers

I'm literally keeping a mood board of "vibes" now – examples of things that feel right, even if I can't exactly explain why, things I like and dont. 

2. Your intuition is more valuable than you think

Those gut feelings about what works? They're not just fuzzy thinking – they're patterns your brain has picked up that you haven't reduced to rules yet.

3. Environment matters more than instructions

I've started spending more time thinking about creating spaces where good things can emerge, rather than trying to control every outcome.

I keep thinking about how this changes everything we do. Like, what if instead of writing another process document, we got better at capturing and sharing intuition? What if instead of more detailed instructions, we focused on creating better learning environments?

I'm really curious about your thoughts on this. Have you noticed this shift in how you work? Are you still writing instruction manuals, or have you found yourself gardening too?

Hit reply – I read everything, and I'm genuinely curious about your experiences with this.

Until next week,

Viraj

P.S. If you're interested in this idea of building differently in the AI age, I'm exploring this more deeply in my next few newsletters. There's something big happening here, and I want to figure it out together.

What I got up to last week