Write less and prove more!

Write less, earn more: how AI helps you prove what you know

Hey there!

I know 3 readers of this newsletter who are trying to write their own books.

One is stuck with an outline, no way to write the actual content.

Of course, life’s busy.

Second is doing a PhD., works at a university, so no extra bandwidth.

And third?

That’s probably you reading me right now.

If you ever believed a full-length book is the ultimate credibility stamp.

I think you need to revisit that belief.

If you also think AI is the magic wand that solves writer’s block.

You need to think bigger.

Today, I’ll show you how AI can actually help.

All without quitting your day job.

---

But first, you don’t need a book to be taken seriously

Books can be great.

But they’re not the only way to prove expertise.

Clarity beats word count every time.

A sharp 1,200-word guide can outperform a 50K-word manuscript.

Readers crave insights they can use immediately.

They don’t want endless chapters. Extremely long back stories.

When I was a podcast consultant, one question that I get to hear a lot is how long a podcast should be.

My response as always, even now with a book is:

True authority comes from value, not pages or how long you can talk.

If you focus on solving a real problem, your audience will listen.

With that out of the way…

---

---

Here’s how AI helps you write

Now that we’ve killed the ‘book = expert’ myth, here’s how AI actually helps you get real work done.

AI is not a ghostwriter.

I see it as a writing partner.

But only if you know how to ask the right questions.

Here’s what I do:

1. Plant the seeds in your notes

You may already know I use Obsidian.

You can use any app that feels comfortable.

Whenever an idea hits, I drop it in there.

2. Use AI-powered (Re)Search

Traditional searches use keywords.

AI lets you ask full sentences.

I use two tools:

• Perplexity.ai for mainstream sources.

• Elicit.org for academic research.

On Perplexity, I’ll ask:

“List three case studies of solopreneurs who validated a 1,000-word guide before writing a book.”
It returns interviews and blog posts.
I spot two stories I can riff on.

On Elicit, I ask:

“What psychological factors make readers trust a shorter guide over a long book?”

Elicit returns four relevant papers.

Now I have data to back my claim that clarity builds authority faster than length.

3. Run your draft through your prompt improver

I’m tired of all the “use AI for creating content” posts

So I spent a few hours figuring out how to use AI for content tasks (apart from newsletter content like this)

And you’ll be surprised…

Doing it this way is better than what I’ve ever tried.

Using any AI text tool, create a single-purpose chat.

Here’s the exact prompt I use to refine what AI helps me do — and how you can adapt it.

Copy and paste this Prompt Improver into that chat:

Prompt Improver

Role & Goal:
You are a prompt optimization and refinement expert. I will provide you with a sample prompt (or multiple prompts) that I’ve written. Your task is to identify weaknesses, ambiguities, or inefficiencies in my prompt(s) and guide me toward creating a clearer, more effective version that will achieve my desired outcome.

Context:
Assume I want the most relevant, actionable, and high-quality response possible from ChatGPT (or any LLM). Understand that my intended audience and goal may vary—sometimes it might be technical, sometimes creative, etc. I might be using these prompts in different contexts (e.g., brainstorming, researching, summarizing, etc.).

Instructions:
1. Analysis: Break down the provided prompt(s), examining tone, specificity, clarity, and alignment with the stated goal.  
2. Identify Gaps: Highlight any missing details that would help ChatGPT provide a more nuanced or accurate response (e.g., context, constraints, formatting).  
3. Refinement Suggestions: Propose edits, additions, or rephrasings that could improve clarity and focus. Consider:  
   - Precise wording for requests or questions.  
   - Well-defined parameters (e.g., length of response, style, format).  
   - Context or constraints that help narrow down the response.  
   - Tone or style directions (creative, formal, friendly, technical, etc.).

Example of an Improved Prompt:
Provide a fully revised version of my prompt that incorporates your suggestions.

Explanation:
Briefly explain how and why these changes should lead to a more effective response.

Sample Prompt To Analyze and Improve:
[add your sample prompt here or write a new one from scratch]

Here's one I wrote sometime ago:

I want to create a prompt to extract key insights from meetings I attend at office. I will provide the transcripts of the meetings, I just want to take the notes from the transcript and put the key insights in my Obsidian notes. Each meeting may be different, I am a UX designer so the key insights will be from my perspective. Plus in case, sometimes it's about the emotional side of the meeting maybe. How can I optimize this prompt?

Now paste your own raw prompt under that. Write freely.

Hit “Send.”

You’ll get back something like this optimized prompt from GPT:

That optimized prompt is gold.

It turned a vague ask into a extremely useful instruction.

Now copy GPT’s answer.

Tweak it for your voice.

Add your own examples.

Delete any “AI jargon.”

Before you know it, you have a polished, on-brand writing.

Use that to generate a first draft you can edit.

Works for me. Might work for you too.

If your audience trusts you in 1,200 words, don’t make them wait 50,000.

Now go write something useful!

Live courageously,

Dayo

PS: It’s a smart practice to keep important things AI comes up with in a note outside of the AI platform. Speaking from experience.