Recursive Intelligence 1.0: The Creator’s Toolkit (Lesson 1)
Option Alchemy: How Hallucinations Create Breakthroughs
Horse in 1900:
“These horseless carriages will take my job!”
Horse in 1925:
“Where did my field go?”
Human in 2025:
“These AIs will take my job!”
Human in 2050:
?
This is the first lesson in a free “Recursive Intelligence” course.
I believe we’re that poor horse in 1900 with little clue of what a new technology meant for its future.
And while I don’t have all the answers to the “?”…
I’ve used AI as a thinking partner (not ho-hum grind replacement) to unlock previously impossible levels of creativity and insight in myself. I have conversations with a probabilistic machine, embrace its "mistakes," and discover hidden patterns in my thinking. This progress has automatically and intentionally fed upon itself.
Hence: “Recursive Intelligence”.
In this course, I'll share the methods I'm still developing through intense, real-time collaboration with AI. Each lesson tackles a different creative challenge, at each stage of the creative process…
The first lesson, "Option Alchemy," focuses on the first stage - idea generation. (I call this “Splurging”.) This AI-powered brainstorming technique will sharpen your taste and judgment…with surprising results.
Buckle up, cowfolk!
BEFORE:
blank empty page Cursor | blink | blinking | blink | blinking | blink | blinking | blink | blinking | sink | thinking into doubt? Begin BEGIN WHERE? BEGIN WHAt? BEGIN HOw… How… how… how do I begin? Damn cursor. Stop BLOODY BLINKING. Mocking. | Waiting. || Judging. |||////___ __ _ _ _ _
NOW:
| The cursor attacks–words ablaze || Vomits "facts"–can't keep track ||| Where there was a lack now I drown IN CRAP...
—==[[CURS0R | TR@NSF0RM3D]]==—
The beast of 10,000,000,000,000 words.
(All of the words, apparently.)
Fear of the blank page is no more. It’s WAY worse than that.
Emptiness shoved aside.
The usurper: abject overwhelm.
Drowning in a flood of vanilla, empty dreck.
You know what?
Let's start THERE…
Let’s begin.
With the SLOP.
Obsessed with Execution
Whether you’re struggling to come up with an idea or overwhelmed with options, the problem is the same:
Where are the INCREDIBLE ideas?
What are the ideas that no one has had?
How will they stand out when creation itself is digitised?
In both cases, the solution is the same.
Yet, most creators never do it.
They build a product, service, business, writing, music, art indistinguishable from anyone else.
(And now - anyTHING else.)
Humans have spent their entire existence in execution:
DOING.
MAKING.
BUILDING.
We’ve had to.
We’ve become very good at it.
We’ve not had the technology to invest more time elsewhere.
But now, if machines handle more of the execution…
“What’s left for us?”
What’s left for us is our birthright.
Where humans are uniquely brilliant.
We’ve not had enough time for this part of the process:
THINKING.
IMAGINING.
DREAMING.
But the greats have always made the time for this.
Finding ways to solve problems and create possibilities NO ONE has done before. Instead of “HOW do I make?” focusing on “WHAT do I make?” and “WHY do I make it?”
We’re not losing our agency.
We are GAINING IT.
The Maths Behind Magic
What humans have created through history is finite.
What we haven’t yet is infinite.
With an always-available AI at your hungry fingertips…
You can feed your brain, gorge ideas and inspiration, stretch seams in your mind ‘til it bursts with what could be!
[SIDE NOTE: Don’t confuse “possible” with “what I know will work”.]
This is why we MUST leverage quantity to create quality.
For a good idea, you need 10 bad ones.
For a great idea, you need 20 good ones.
For an INCREDIBLE IDEA?
It’s [probably way more than] 40 great ones!
Until now, enough quantity to discover the great was hard.
In one of my business courses (The Magnet Model), I created a “Brainstorm Your Business” challenge to identify 100 new ideas in one week. Most aren’t used to this idea velocity, and I considered making it easier. But in good conscience, I couldn't, as this IS the process of finding a great idea.
Until AI came along.
Because what we can do now is EXPLORE…DISCOVER…TRANSFORM…
Step 1: Choose Your Adventure
NOTE: Don’t put it off; do this now with me as I describe the process.
And if you have any questions…
Pick ONE area you want to explore:
A business idea that won't leave you alone
A creative project begging to be born
A problem in your industry nobody's solving
A course you'd love to create
A service that must exist (but doesn't)
Step 2: Map Your Territory
Next, you must create context for both you and the AI.
Write freely about:
What excites you most about this idea?
What makes you nervous about attempting it?
What have others done in this space (and why isn't it enough)?
What have you tried before that’s similar?
What unique perspective do YOU bring?
What's the REAL problem you're trying to solve?
What's holding you back from starting?
What would make this TRULY different?
Who NEEDS this (but might not know it yet)?
Don't edit.
Don't filter.
Don't judge.
You’re not finding answers.
You’re not looking for solutions.
You’re not prompting in the futile hope that the AI will deliver the “right” idea.
You’re opening the floodgates for your unconscious to hurtle through.
This brain dump is fuel for you AND the AI.
Every concern, dream and half-formed thought becomes raw material for exploration. It is impossible to get this wrong. JUST WRITE! The more fuel you create, the more interesting places you’ll explore together later.
Do this now:
Put on your favourite album.
Set a timer for 10 minutes (or more!)
Pour it ALL out.
Step 3: “Attune” the AI
I will write another piece about Attuning in more detail, as this is the default method I’ve developed to collaborate with AI.
[SIDE NOTE: I rarely ask AI to be something like “a world-class expert in underwater basketweaving highly skilled at making soggy baskets”. My theory is this is the main reason it outputs vanilla crap.)
Here’s what to do:
Ask it to review your 10-minute free write with questions like:
“What’s your favourite part of what I wrote?”
“What did you notice?”
“What surprised you?”
“What spoke to you most deeply?”
(And feel free to come up with your own questions.)
Now, before you get on your high horse about anthropomorphism…
YES.
I AM speaking to it as if its a human.AND…
NO.
I DO NOT THINK it is a person.[I don’t believe we can even say it’s “intelligent”.
Yet another piece for my Substack hopper!]
I pretend it’s a person (IT IS NOT) because I get better results.
(Again, I’ll explain why I believe Attuning works in another piece.)
If you need proof of the difference Attuning makes?
Just try it. There’s your evidence. :)
Step 4: Time to Delve. But WIDE!
Why ask for only 10 options from an AI…
Why that rising frustration when it messes up…
Why invest hours improving the prompt to get the perfect answer…
…when it is a PROBABILISTIC machine?
It does not “retrieve information”.
It predicts tokens based on probabilities.
Why instruct a Supernova to generate birthday candles?
We are so used to binary on or off machines that we don’t know how to work with one that’s so…
FUZZYWUZZY.
Mistakes are a feature, not a bug.
So let’s USE that feature.
Here’s what to do:
Ask the AI for 100 options (or if you have time - more!)
But here's the critical part:
Ask it to include ideas from distant fields, weird angles, and strange perspectives. Suggest it generates a few bad, strange or nonsensical ideas. By doing this, you are priming:
Your mind
The AI
(You’ll find out why in a sec.)
Why 100?
If the first 20 are obvious…
And the next 30 are interesting…
In the last 50 – the GOLD appears!
Step 5: From Mining to Magic
As the AI blurts out the 100 options (many = GARBAGE), notice this:
Some ideas will trip an:
“oooooh” or “hmmmm” or “eh?”
Not because they're good.
Not because they're useful.
But because...something...resonates.
This can be subtle…
Don’t expect angels.
Don’t wait for rainbows.
Unicorns won’t appear.
Some create a delicious ripple you can’t quite taste.
This is what I lovingly call:
THE PING.
You can't explain it.
You can't justify it.
You just...sense it.
If this seems too WOO…This ain’t woo. It’s SCIENCE.
Can you immediately bring to mind:
Every skill…
Every trait…
Every memory…
Every resource…
Every experience…
…that could help you progress this idea?
If your answer to that question is yes, I’ll have some of what you’re having!
If it’s no - then what you do have is “THE PING”.
It’s your unconscious intuition whispering “pay attention”.
Learn to trust your gut.
(Of course, verify it–but later.)
So as you read through these ideas, be open to “THE PING”. Become the ultimate “THE PING” detection system. Because that “THE PING” is a spark.
A flicker.
An echo from the depths.
And now comes the fun bit.
Where we switch from mining to alchemy.
Step 6: Hidden Patterns
DO NOT reject any of “THE PING” ideas.
(It’s way too early, and you’ll ruin the method.)
Especially the:
Bad options
Weird options
“I’d NEVER do that” options
“I don’t know how I’d begin” options
(Of course, you can include the good ones too!)
If they have THE PING, paste them back to the AI and ask:
“Why did these options resonate with me?”
“What hidden patterns or connections might exist that I can’t see?”
“Why didn’t I pick the other options?”
“What does this tell you about me and what I’m attempting to do?”
BOOM.
The AI starts identifying patterns in YOUR thinking that YOU can't see.
It becomes a MIRROR.
Step 7: Alchemical Transformation
If there’s JUST ONE rule I could give to anyone using AI, it’s this:
ALWAYS curate what the AI creates.
By selecting “THE PING” options, you’re already curating its output. After that, if you believe every word that billows from its blinking maw?
You’re not always curating.
You only did it once.
You MUST. MUST. MUST ask it to explain the patterns it’s identified:
“Why did you say that?”
“What are the reasons you noticed that?”
“Oh that makes a lot of sense! I’d never thought of it like that.”
But if something seems off (which WILL happen)…
DO NOT tell it it’s wrong!
(LLMs are trained to be way too agreeable. If you say “no”, it will gaslight itself!)
Instead of dismissing anything you think is wrong, take this approach:
“I’m not sure about that, although I am open to it.
I’m interested to know why you said that?”
☝️ Uncertainty leaves the door open in your mind and the AIs output for you to create new understanding, new knowledge, new insight.
And on the subject of open doors…
Step 8: How to Decide and Take Action
This process could be infinite.
(Or if you use Claude, until your daily usage runs out!)
There are infinite possibilities for where your chat could go, so I can’t be definitive other than to say that an LLM won’t offer an end point.
(Again, this is a feature, not a bug.)
So once again, you must “curate” your own resolution.
But here are some ways your chat might end:
You’re so excited about an option the LLM suggests that you ask it to suggest an immediate action plan.
(CURATE the output!)You’re so excited about an option YOU suggested that you ask it to suggest an immediate action plan.
(CURATE the output!)You’re so overwhelmed with new insights that you put the chat aside and let it all “process” for a day or so.
(TAKE a walk!)You’re right in the middle of it, then your wife tells you to stop talking to your AI gf and pay her some attention instead.
(GET a life, Mike!)
Any of these is PROGRESS.
(Except 4, of course.)
You’ll either have an action plan, or you’re closer to choosing a plan of action. Either way, you’ve made the most valuable progress there is:
Self-Knowledge.
This Isn't Theory. It’s Practice.
Last month, I launched a new small-group online business training:
The Strategic Evolution Workshop
It uses the framework I’ve used to develop the strategy, plan and execution of my second business (in under a month) with my AI thinking partners.
But I had a problem.
In less than four weeks, I’d had such an avalanche of mind-blowing insights that I struggled to create a structure that wouldn’t overwhelm the attendees.
I’ve designed a ton of workshops for the last 14 years.
For the last 4 years, at least two a week.
It’s my bread and butter.
But no matter what I tried didn’t quite capture…
It…felt…OFF.
It was missing a certain “ping”?
So I ran an Option Alchemy.
Here’s a selection of the options where I felt “THE PING”:
A jazz improv where each session is a different instrument
A reverse curriculum where clients see the end result first
A fractal structure where each part contains the whole
Four landscape paintings, each in a different season
WTF IS that?
But when I asked the AI why these resonated, it showed me patterns in my thinking that I was too close to see.
I'm drawn to:
Simple, flexible structures, not complicated, rigid frameworks
Similarities between different subjects and scales
Movement between contrasting but complementary states
The whole emerging from seemingly separate parts
Oh…
I see…
Of course…
THIS. IS. ME…
How my mind works…
A pattern obvious once revealed…
INVISIBLE before that.
That same day, I structured the Strategic Evolution Workshop around these principles.
The result? We’re only halfway through and here are a few client quotes:
“I LOVE this. I’m so happy I enrolled.”
“And there I am with tears in my eyes. So thank you for sharing your process with us as its really coming at the exact perfect time for me. Looking forward to today’s session.”
“I've always just wanted to make music. But doing this I'm kind of digging into why have I always wanted to do that? And what might be the unique value that what I am trying to do might offer anybody, and what is it doing for me and what is the synergy?
“BAM! Out of nowhere, I had the beginnings of an idea and a feeling of what I want my business to shape into. I now had something to consider- and it honestly felt good. It was energizing and scary and fun and exciting all at once. It also carries a thread of familiarity in what I’m currently doing.”
This kind of thinking is the ONLY way to win in the AI Age.
The winners aren’t those who use AI to think for them.
They use AI to discover how uniquely brilliant they already are.
Onwards & upwards,
Mike
“Use the AI as it is
To discover who you are.”
P.S. In the next lesson in “Recursive Intelligence 1.0: The Creator’s Toolkit”, I’ll give you a deeper method for developing your ideas (i.e. Splurges) in the second stage of the creative process: Discovery.
P.P.S. If you want to use AI as a thinking partner to build your creator business, I'm running another 5-week small group workshop this month.
Would you like to join us?
If so, send me a message, and I’ll get you all the details…
3 hours has never flown by so fast in my life 😳 Using these prompts and engaging in AI this way is so exciting and fun when you're using it to expand your thinking.
Will be following along and hope to join this soon as I venture out form my monk mode :)
While scratching the surface with AI I waws wondering before I setup my Agent and commit, do you have a recommended LLM? thanks!