Last week, I was feeling completely overwhelmed. My mind was a mess of half-formed thoughts. I found it impossible to focus on any single idea.
And then I thought of Dumbledore…
In the Harry Potter books, he uses a device called a Pensieve. It’s a stone basin where he can deposit excess thoughts and memories, pulling the silvery strands directly from his mind. He explains:
“I sometimes find that I simply have too many thoughts and memories crammed into my mind... It becomes easier to spot patterns and links, you understand, when they are in this form.”
He pulls the chaos out to see the connections.
And I realized this fantastical idea provides the blueprint for a profoundly effective creative workflow.
A lot of founders and marketers are drowning in half-formed thoughts, brilliant shower ideas, and nagging questions. We let them swirl around until they evaporate.
But what if you had a Pensieve?
What if you could pour that chaos into a system, and have an AI dive in to find the gold you’re too busy to see?
How to build your AI-powered Pensieve
The biggest lie I bought into early in my career is that you need to “come up with” new ideas. You don’t. You are already sitting on a mountain of them. But the problem is a lack of a system to capture and strategically filter them.
This workflow fixes that.
It’s a simple, repeatable process for turning your raw thoughts into a strategically prioritized content pipeline. This is how you build an AI to reflect your best thinking back at you with a clarity that’s difficult to achieve alone.
Step 1: The weekly brain dump
Set aside 15 minutes, once a week. I do mine on Thursday afternoons usually, but find whatever works. Open a single, clean document. For those 15 minutes, you will write without stopping and without judgment.
Do not filter. Do not edit. Do not worry if it makes sense.
Honestly, my brain dumps are a mess. They’re embarrassing. They’re filled with half-formed sentences and petty frustrations. And that’s the point.
Just dump everything that’s in your head. Questions you have. Snippets of conversations. Frustrations with your industry. A weird observation you made. A half-baked product idea. An article you disagreed with.
⏱️ Time: 15 minutes, no more
✅ Success looks like: A messy, unfiltered document of 300-500 words that would embarrass you if anyone else read it
Your brain dump might look like this:
Why do all B2B SaaS blogs sound the same? It’s like a single robot wrote all of them. So boring. A customer an onboarding call said they felt “paralyzed” by our competitor’s dashboard. Interesting word. The “Jobs to be Done” framework feels too academic. Is there a simpler way to explain it? Just saw a marketing agency promise “guaranteed viral content.” What a f*cking scam. Read a thread on how Gen Z hates polished corporate-speak. They want raw, behind-the-scenes stuff. How could we do that without looking messy?
This document is your Pensieve. It’s the raw material. And now the real work starts when you hand it over to your intern (guess who that is!)
Step 2: Extracting strategic themes
Copy your entire brain dump. Open your AI assistant of choice (Gemini, Claude, etc.) and give it this upgraded prompt.
Prompt:
You are a qualitative researcher specializing in thematic analysis and strategic positioning.
I’ve pasted my unfiltered weekly brain dump below. Your job is to perform a layered analysis:
LAYER 1: Surface Themes
Identify 3-5 explicit topics I’m thinking about.
LAYER 2: Underlying Tensions
For each theme, identify the TENSION or CONTRADICTION I’m wrestling with.
LAYER 3: Strategic Implications
For each tension, explain:
- Why this tension matters for my audience (not just me)
- What unique angle this gives me that most people in my space aren’t exploring
- What blind spot this reveals about my market
Output format:
Theme: [Name]
Tension: [The conflict]
Strategic Opportunity: [Why this matters]
Competitive Blind Spot: [What others are missing]
Here’s my brain dump:
[paste your dump]⏱️ Time: 5 minutes to run the prompt, 2 minutes to read the output
✅ Success looks like: 3-5 themes with clear strategic implications, not just topic summaries
The AI acts as an external processor for your own thoughts, doing exactly what Dumbledore did: spotting patterns and links you would otherwise miss. (It’s surprisingly good at this, which can be both helpful and a little creepy). From the example above, it might come back with:
Theme 1: The Authenticity Gap. You are repeatedly circling the conflict between polished corporate content and the demand for raw, authentic communication.
Theme 2: The Promise of Simplicity. You are frustrated with complex jargon and frameworks, seeking simpler ways to solve customer problems.
Theme 3: The Deception of Guarantees. You feel a strong sense of injustice toward empty marketing promises and unethical claims.
Suddenly, the chaos has a structure. But not all themes are created equal…
Step 2.5: Business impact scoring
Before you start generating content angles, you need to know which themes actually matter for your business.
Take your 3-5 themes and run this prompt:
You are a business strategist. I’ve identified these themes from my thinking this week:
[paste themes with tensions and opportunities]
For context, here’s my business positioning:
- Product: [Your product/service]
- ICP: [Your ideal customer - be specific]
- Current challenges: [List 2-3 business challenges you’re facing]
- Revenue goal this quarter: [X or describe qualitatively]
For each theme, score it on three dimensions (1-10):
1. Strategic Alignment: How well does this theme connect to my core positioning and ICP pain points?
2. Competitive Differentiation: How likely is this theme to produce angles my competitors haven’t covered?
3. Revenue Proximity: How directly does this theme connect to actions that drive revenue (newsletter signups, product trials, consulting inquiries)?
Provide scores and brief justification. Then rank the themes by total score.
Tell me which theme I should prioritize THIS WEEK and why.⏱️ Time: 3 minutes to run, 2 minutes to review
✅ Success looks like: A clear ranked list showing which theme has the highest strategic value.
Now, before we move on to steps 3+, here’s where it gets interesting….
Alright, let’s play a delightfully dark game.
There’s a mental model from writer Luca Dellanna called a Premortem. Most teams wait for a project to crash and burn before they conduct a post-mortem to figure out what went wrong. A Premortem is the opposite. You begin by assuming the project has already failed.
And because it’s October and I apparently have a thing for autopsies, here’s the other side of the coin: last week’s piece on conducting the million-dollar post-mortem when prevention is already off the table:
Let’s apply this to your Pensieve workflow.
Imagine it’s six months from now. You’ve been running your weekly brain dumps religiously and have the themes, but your content strategy has failed.
What is the most likely reason it all went wrong?
I’ll bet you my entire collection of sci-fi paperbacks that the answer isn’t a technical SEO error or the wrong email subject line.
The most likely reason for failure is this: Your ideas were boring.
You didn’t have a point of view, so you tried to copy everyone else’s.
And here’s why this matters for the next 2 steps of the workflow: The Premortem forces you to confront the single biggest threat to your work: 🧲the magnetic pull of mediocrity🧲
Steps 3 and 4 are all about systematically preventing the most likely path to failure. You’ll be using my proprietary framework to kill boring ideas before they waste your time.
Okay, back to the Pensieve. Let’s finish building this thing.
Step 3: The Boring Detector (Killing generic angles before they kill you)
Pick your highest-scoring theme from Step 2.5.
Now, here’s where most people screw up: They jump straight to brainstorming content angles. But brainstorming without filtering is how you end up with 35 ideas that all sound like everyone else’s.
Instead, we’re going to use The Boring Detector, a framework I developed to systematically filter out generic angles BEFORE you waste time on them.
Here’s the prompt:
You are an expert content strategist with a brutal intolerance for generic ideas.
I’ve chosen this theme to explore: [paste your winning theme]
Your job is to generate 10 content angles on this theme, but you must run each angle through “The Boring Detector” before presenting it.
THE BORING DETECTOR - Five Tests Every Angle Must Pass:
1. The Copycat Test: Could a competitor with access to ChatGPT write this exact angle in 30 seconds? If YES, it’s boring. Reject it.
2. The Google Test: Does a version of this angle already exist in the top 10 Google results for related keywords? If YES, it’s boring. Reject it.
3. The POV Test: Does this angle require MY specific context, experience, or perspective to write authentically? If NO, it’s boring. Reject it.
4. The Controversy Test: Would this angle make at least 20% of my audience disagree or feel uncomfortable? If NO, it’s boring. Reject it.
5. The Specificity Test: Can I name a specific person, company, or moment that exemplifies this angle? If NO, it’s boring. Reject it.
PROCESS:
- Generate 15 raw angles on my theme
- Run each through all 5 tests
- Only present the 10 angles that pass at least 4 out of 5 tests
- For each surviving angle, show which tests it passed and briefly explain why it’s NOT boring
FORMAT:
Angle: [The content angle]
Boring Detector Results: ✅ Passed 4/5 tests
- Passed: Copycat, POV, Controversy, Specificity
- Failed: Google (similar angles exist, but none with this specific POV)
Why it’s not boring: [1 sentence]
My theme: [paste]⏱️ Time: 5 minutes to run, 3 minutes to review
✅ Success looks like: 8-10 angles that feel sharp, debatable, and unique to YOUR perspective
The AI will turn your single theme into a dozen distinct angles. You might get back ideas like:
Angle 1: “Why ‘authentic’ marketing is the new corporate bullshit: A taxonomy of performative vulnerability”
Boring Detector Results: ✅ Passed 5/5 tests
Why it’s not boring: Directly challenges the trend everyone’s celebrating; requires specific examples of fake authenticity you’ve observed
Angle 2: “The marketing tactics I refuse to use (and why my competitors will hate me for this)”
Boring Detector Results: ✅ Passed 4/5 tests
Why it’s not boring: Names specific tactics and takes an ethical stand; highly specific to your experience and values
Notice the transformation. You’ve moved from a vague, abstract feeling to a list of sharp, concrete, and debatable content angles.
If fewer than 5 angles pass the Boring Detector, your theme might not be differentiated enough. Go back to Step 2.5 and pick your #2 theme.
Step 4: Give it a name worth remembering
You have your non-boring angle. Let’s take Angle 2: “The marketing tactics I refuse to use.”
Prompt:
My content angle is: “The marketing tactics I refuse to use (and why my competitors will hate me for this)”
Your task is to generate 10 title and subtitle pairings for this concept. They must be bold, specific, and create intrigue. Follow these rules:
- No generic clickbait.
- Use strong, active verbs.
- The subtitle should clarify or add a compelling twist to the title.
- At least 3 titles should include a specific number or named tacticThis is how you move from a vague idea to something sharp and compelling, ready for you to write.
BONUS: Step 0 - The monthly Meta-Pensieve
Once per month (I do this on the last Friday), run this workflow to uncover your long-term content thesis.
Collect your last 4 weekly brain dumps and run this prompt:
I’ve been doing weekly brain dumps for the past 4 weeks. Here they are:
Week 1: [paste]
Week 2: [paste]
Week 3: [paste]
Week 4: [paste]
You are a strategic analyst. Your job is to find the DEEPER patterns that emerge across multiple weeks—the obsessions I keep returning to, the tensions I haven’t resolved, the questions I’m circling but not answering.
Identify:
1. The Meta-Question: The one question I keep asking in different ways (even if I haven’t articulated it directly)
2. The Philosophical Tension: The underlying tension driving multiple themes (the “big idea” I’m wrestling with)
3. The Content Thesis: What should my next quarter of content be ABOUT at the highest level? What’s the through-line that connects my best thinking?
This is not about summarizing. This is about finding the signal I’m too close to see.A critical implementation note:
The system only works if you protect the integrity of Step 1. Your brain dumps must be truly unfiltered. The moment you start performing for an imagined audience or self-censoring, you lose the raw material that makes the output unique. Don’t pretty it up. Let it be messy.
This system + your preventative medicine = your ultimate advantage
Your Pensieve is complete. Here’s what you’ve built:
A systematic way to…
capture your unique perspective (the brain dump)
extract strategic themes with business context (the upgraded pattern recognition)
filter for strategic value (Business Impact Scoring)
systematically kill boring angles (The Boring Detector)
package them irresistibly (the title prompt).
PLUS: A monthly practice to uncover your long-term content thesis.
At every step, you’re performing a Premortem, actively choosing against the most likely path to failure.
But here’s the natural next question:
“Great, I have my unique idea. Now how do I write about it with AI without sounding like a robot?”
For this, I want to point you to a brilliant piece by James Presbitero and Daria Cupareanu called “How to Use AI for Writing Without Losing Your Voice.”
The connection to this workflow is perfect: they walk through a 5-step writing process, and the Pensieve system is the engine for that critical first phase: generating ideas that can’t be prompted into existence because they come from your specific context and perspective.
An endless supply of unique ideas is worthless if you can’t articulate them with surgical precision. Their piece is the missing half of the equation.
I showed you mine. Now show me yours: What’s one idea you’ve been sitting on for MONTHS that you’re afraid is either brilliant or completely stupid?
To killing the boring ideas before they kill you,
Chief 🤖 at ROBOTS ATE MY HOMEWORK
🧠 Did I unlock a memory? The Pensieve holds a few other thoughts you might find valuable:
Absolutely excellent stuff Mia. I am flabbergasted by your critical thinking ability, ideas, prompts and stories. You're truly something my dear. I definitely underestimated you.
Absolutely loved your idea! I’m adapting it for the healthcare/Medical Affairs world, it’s such a clever and practical framework. Brilliant work, congrats!