CLASS 16July 15 · Thinking Design + Think Print

Class 16: Thinking Design + Think Print. "If you always think the same way, you just create a loop of the same of yourself."

Taught byDominik HeinrichCo-founder, Creative AI Academy
Global Head of Design Intelligence, The Coca-Cola Company

The class in brief

Dom ran a night built around one idea: prompting from your own default thinking just feeds AI back to you in a loop. After this page you can do three things.

Name your own default thinking archetype using Think Print.
Prompt AI in your stretch instead of your default.
Run any brief through Dom's prime-then-loop process, which he calls Infinite Design.

The mediocrity loop

Why this matters · 6:12 PM

Old power was passive. New power is interactive. Augmented power runs in-line with your life.

Dom opened on a three-stage power ladder , landing on "augmented sapient": technology running in-line with life instead of separate from it .

The pivot was automaticity: about 40% of daily actions run on habit , and the brain defaults to autopilot to save energy . The room spent ten minutes naming one thing each person does without thinking, then how they would disrupt it.

The class raised three examples.

Brushing teeth too aggressively.
Checking the phone before anything else in the morning.
Swapping a Candy Crush habit for solitaire, then relapsing into a new app.

Noticing the loop turned out to be the only way out of it.

40%

of daily actions are governed by habit, per the Duke University stat on Dom's slide.

stages on the power ladder: old (broadcast), new (interactive), augmented (in-line with life).

The cost of offloading

The cost of offloading · 6:57 PM

Unexamined AI use bolts a prediction machine onto an autopilot brain.

AI processes data roughly a million times faster than the human brain . And the deck's own numbers make the tradeoff concrete .

61% of knowledge workers report less mental friction, which sounds like relief until you notice the other two numbers moving the wrong way.

Those two numbers are a 2-5 point IQ decline and a 42% loss of divergent thinking. Dom's answer is not to unplug, it is meta-learning: knowing how you personally learn, so AI augments that instead of replacing it.

2-5

points of measured IQ decline, the Reverse-Flynn Effect, National Institutes of Health, 2023.

42%

loss of divergent problem-solving tied to cognitive offloading, MIT Sloan's AI Productivity Paradox, 2025.

The framework · 7:01 PM

The six Think Print archetypes

Centristfocus and control

Grounds in facts first, wants clarity before moving.

Patternistsystem and repetition

Sees breadth: what else can these same pieces build.

Precisionistprecision and order

Goes deep, polishing one version until it is flawless.

Imagineervisionary, inspiration

Opens the world as wide as possible before narrowing anything down.

Interpretermeaning and influence

Reads between the lines, translates subtext into direction.

Activatoraction and execution

Turns the idea into one concrete next step.

ThinkPrint Types, all six archetypes on one slide

Each archetype pairs with an opposite stretch . Either write your own prompts as your stretch, or assign the stretch role to AI and answer as your default yourself.

The worked example was a video game: an Imagineer alone gets an endless, beautiful open world. The same idea pushed through the Activator lens gets a tight, iconic core loop about smashing boxes as a cheeky monkey in a dense city .

The exercise · 7:13 PM

Take the test to find your default and your stretch

The test

Take the short or long Think Print test, then post your result to the class Miro board.

The test lives at creative-ai.academy/thinkprint. Once you know your default, the method is simple: write your own prompts as your stretch, or tell the AI "act as my stretch, I'll write as my default, answer in stretch mode."

Dom's own habit: sketch on paper before opening any AI tool. The hand-to-brain link changes what comes back.

Awareness is the first move

The default is not a flaw to fix. It is a shortcut the brain built to save energy. The point of naming it is not to abandon it, it is to notice when you are about to feed AI your own pattern back to yourself.

Humans lead, AI follows, the pivot line that closed the lecture

"Humans lead. AI follows" ran as the pivot line out of the lecture and into the test , right before the room went quiet to take the test individually.

The craft · 7:25 PM

From classic design to computational design

Classic designfeature and quality focus

A taxi's color, made easy to spot from a block away. The product itself is the whole unit of design.

Design thinkinghuman needs and experience focus

Uber does not redesign the cab, it redesigns the inconvenience of hailing one.

Computational designsystems and data focus

New York took Uber's ride data together with the city's own accident data and used the two to place protected bike lanes between parked cars and the sidewalk.

The design-thinking-to-thinking-design ladder

The taxi example ran twice, once per rung , then side by side . Dom's framing for where this goes next: "the next generation of innovation is not business, technology, and human, it's actually behaviors, technology, and context."

The close, and the homework

The close · after the break

Prime it and then loop it on the Coca-Cola brief

Infinite Design is a think-and-generate loop laid on top of any design process, never a replacement for one.

Prime it, giving the machine context it doesn't have, using the prompt design canvas: goal, audience, situation, known data, blind spots, what not to do.
Pick the thinking method, your Think Print stretch as the default choice.
Pick the generation tool for each stage.
Run about three design loops, challenging the machine each time before moving on.

In 1985 Coca-Cola became the first soft drink consumed in space, the Coke Space Can; last year marked its 40th anniversary. The provocation: engineer Coca-Cola for zero gravity in a way that redefines the drinking experience and creates uplift back on Earth, through the product, packaging, ingredients, or another route.

Dom's spoiler of the lazy answer: small bubbles that burst in your mouth.

The homework runs the same loop on Miro: prime, apply your stretch, run three loops, screenshot the progression. Dom reviews it for about 30 minutes in the following week's class.

Try this prompt

Quiz me on the six Think Print archetypes and their stretch pairings. Then hand me a small brief and walk me through one Infinite Design loop on it, priming first.

You will know it worked whenit quizzes you on the six Think Print archetypes and their stretch pairings first, then walks a small brief through one full loop, priming before it generates anything.

Methods and prompts

Five methods to take with you

METHOD 01 · THE TEST part of: keep the judgment human

Predict your Think Print before you take it

Guess your own default and stretch first. Comparing your self-read to the actual test result is the real diagnostic, not the test alone.

Where it came fromDom taught this as a way to engage with the Think Print test he had the whole class take live in session 16, a short quiz that identifies each person's default thinking archetype and its paired stretch.Use it whenUse this right before taking the Think Print test, so you can compare your own self-read against the real result.

Same idea elsewhereJudge AI output with your own trained eye (class 1) · The sea-of-sameness check (class 2) · Pick the relationship on purpose (class 2)

Working prompt

My guess at my own default thinking archetype and my stretch, before I take the real test: [my answer]. I answer first, then you check me. Explain what each of the six archetypes optimizes for, and tell me honestly if my guess holds up or if I am flattering myself.

You will know it worked whenit tells you honestly whether your own guessed archetype and stretch actually hold up against the real test, not just a flattering confirmation.

METHOD 02 · THE STRETCH part of: go wide then narrow

Prompt in your stretch rather than your default

Assign the stretch role to the AI and answer as your own default yourself, or write the whole prompt as your stretch. Either way, the output stops rhyming with your own pattern.

Where it came fromDom taught this in class as the practical payoff of knowing your Think Print archetype, explaining that prompting from your own default just feeds AI back a loop of your own pattern.Use it whenUse this when you want a brief pushed somewhere your own default thinking style would not naturally take it.

Same idea elsewhereThe lateral reframe (class 2) · Start a fresh chat instead of one long thread (class 3) · Force the creative partner deeper (class 4)

Working prompt

I am a [your default archetype]. For this brief, act as a [your stretch archetype] instead: [the brief]. Push for what a [stretch archetype] would notice that a [default archetype] would miss.

You will know it worked whenit points out something specific your stretch archetype would notice on this brief that your own default archetype would have missed.

METHOD 03 · PRIMING part of: structure your ask

Prime before you generate anything

Give the machine the context it cannot know before asking it for a single idea. This is the prompt design canvas, condensed into one paste.

Where it came fromDom taught this as step one of his Infinite Design process, condensing the prompt design canvas, goal, audience, situation, known data, blind spots, and what not to do, into a single paste before any generation begins.Use it whenUse this before starting any AI-assisted design work, so the model has the context it cannot otherwise know.

Same idea elsewhereGive it two keyframes instead of one description (class 2) · Speak the language of the internet (class 3) · Write it as a GOLD block (class 4)

Working prompt

Before we generate anything, here is my priming: Goal: [what I'm trying to achieve]. Audience: [who this is for]. Situation: [what's true right now]. Known data: [what I already know]. Blind spots: [what you cannot know from here]. What not to do: [the guardrail]. Confirm you understood all six before we start.

You will know it worked whenit confirms back all six priming fields, including the blind spots and the guardrail, before generating a single idea.

METHOD 04 · THE LOOP part of: reject the first draft

Run three challenge loops

Infinite Design in miniature: generate, challenge, regenerate, three times. The challenge step is where your stretch archetype earns its keep.

Where it came fromDom taught this as the fourth step of Infinite Design, then ran it live in class against the Coca-Cola zero-gravity brief, challenging each round's output before moving on to the next.Use it whenUse this when a first AI-generated idea feels too safe or too obvious and needs to be pushed further.

Same idea elsewhereSet up a debate partner (class 4) · Build cheap and then swap the model up (class 8) · Critique mode (class 10)

Working prompt

Here is loop 1's output: [paste]. Challenge it as my [stretch archetype]: what did this miss, what is too safe, what is too on-the-nose. Give me loop 2 based on that challenge, then ask me if it is ready for loop 3.

You will know it worked whenloop 2 actually responds to the specific gaps the challenge named in loop 1, and it asks before moving on to loop 3.

METHOD 05 · PHYSICALITY part of: context beats prompts

Sketch before you prompt

The hand-to-brain link processes information differently than typing does. Draw or write the idea out first, then describe it, then generate.

Where it came fromDom shared this as his own habit in class, saying he gets better results when he writes or sketches an idea by hand before opening any AI tool at all.Use it whenUse this before prompting an AI tool, when you want the hand-to-brain thinking process to shape the idea first.

Same idea elsewhereControl the input and build gradually (class 2) · Show it your voice and not just the output (class 3) · Evolve the reference instead of restarting (class 5)

Working prompt

Here is my hand sketch of the idea, described in words: [description]. Before you generate anything, tell me back what you understood from the description, then wait for me to confirm before you produce a single output.

You will know it worked whenit restates what it understood from your description and waits for your confirmation before producing a single image.

The shelf

Tools and references

Tools that night

The Think Print test, short and long versions, at creative-ai.academy/thinkprint
Miro: per-student Infinite Design canvases, the prompt design canvas template, a thinking-methods handbook
ChatGPT, Claude, and Gemini as the generation tools per workflow stage; "OpenAI Codex or Claude Design" named for a later stage