We are no longer amazed. We no longer marvel
A few thoughts on AI and creativity, slow kitchens, bored children, and whether we still know how to wonder
What you’ll read in today’s newsletter:
Imagination means discovery and growth. But how does Artificial Intelligence (AI) affect our imagination and creativity, especially in childhood?
How creative is AI, really? According to a study, more creative than the average person, but not quite as creative as the top 10% of human “creativity maestros”.
What do GenAI and …otters using Wi-Fi on planes have in common? A unique collage by Ethan Mollick shows—within just six frames—how much image generation models have evolved over the past three years.
We often say the EU needs to support AI innovation and entrepreneurship more—but how many of us actually use Mistral?
I. What happened to the wondrous luxury of slowness?
“Do you remember how, in the past, watching someone cook something special felt magical, as if it meant something?” he asked me, completely out of the blue.
And I really did remember. Watching my mother prepare her complex dishes was something I did silently as a child, with the same devotion and curiosity I’d have for a favorite movie. Especially when the kitchen light came in at a certain angle, everything -the utensils, chopped vegetables, colorful spices, her hands- seemed to glow and that was pure magic to me.
I’d stand there watching her, fully aware that I was witnessing a self-contained creation that, in a few hours, would give me a simple, genuine joy through taste, aromas and sight. I quickly wanted to learn how to cook -not as a daily chore, but as a form of applied everyday magic. I still love cooking, and I never get tired of it. In fact, the more exhausted or fed up I am, the more it helps me feel lighter.
I told him all this in response. “But you realize, don’t you, that it’s not the same anymore?” he said. “Few everyday things mean something. We photograph more than we feel. We glance more than we wonder. We rarely pause to ask how or why something is happening or wonder how something can be created”.
It was one of those late-night conversations of ours that always open a door for me toward something I want to explore. We are no longer easily amazed. We no longer marvel. And on top of that, we have so many ready-made answers before we've even had time to ask the questions.
I began to wonder what this means for our creativity. I remembered a book by philosopher Bertrand Russell I first read around age 20 and recently revisited. It said something like this:
A child who never has time to be bored, because their parents constantly keep them entertained, is so absorbed by stimuli that it's hard for them to create anything of their own. All they think about is where the next quick pleasure will come from.
I thought of children of all ages, glued to screens. Of how AI is “always there,” when adults can’t -or won’t- be. I thought of the speed of what happens on those screens.
No slowness. No room to pause. I wondered what those immediate, ready-made answers might do to a child’s creativity; the lack of time and space to imagine, to daydream with open eyes, to wonder and ask questions about what they want to learn and why.
II.
To imagine is to grow
Recently, I watched a video here on Substack, a black and white interview with Ray Bradbury, whose masterpiece Fahrenheit 451 I’ve read four times, devouring it each time like it was the first. In this video, he emphasized the importance of imagination in children:
“The ability to fantasize is the ability to survive. And the ability to fantasize is the ability to grow. Boys and girls at the age 10, 11, 12, 13, spend the most important time of their day—especially at night, before going to sleep— in dreaming of themselves becoming something, being something. When you’re a child, you begin to dream yourself into a shape. An then you run into the future, trying to become that shape.”
Decades ago, that may have been true. Today, more children than we, as a society, can afford rarely daydream before bed. Instead, they consume pre-made content at breakneck speed. And when that blue light goes off, how much energy do they have left to dream about the shape of their future selves—when even in the dark behind closed eyes, they still see the afterimages of games and fantastical metaverse worlds blinking away?
III.
Gaining time, forfeiting creativity
It’s not just children of course. Adults too are developing automatic behaviors.
From many conversations I’ve had recently, I’ve found that when we want to create something, many of us no longer think with the same autonomy. Before writing a business plan, analyzing a budget, choosing a word or title, drafting the perfect email message, we turn to ChatGPT, Claude, or Gemini.
I can’t help but feel that what we gain in time, we forfeit in creativity. That instinct which makes us inventive and original is at risk of going dormant. Instead of using AI as a tool to enhance our creativity, we often assign it the role of a director, tasked with finding new paths for us, so we don’t have to break a sweat using our own right brain hemisphere.
As Nick Cave put it in one of his letters (remember what he wrote in #VII in one of my recent newsletters), maybe technology will eventually replace the chaotic, unpredictable, emotionally charged act of human creation -an act born from the depths of the self.
Cave’s concern isn’t just whether AI can write a good song. Ιt’s the risk of creative inertia, the outsourcing of pain and passion that define art itself.
IV.
So, how creative AI actually is?
The truth is, AI now creates compellingly—whether in writing, music, or even clever and realistic business plans. In writing and music, it’s becoming increasingly hard -sometimes impossible- to tell apart AI work from that of a human professional.
But is it really creative?
I recently read two fascinating studies I’d like to share with you.
The first suggests that AI can be trained to be more creative. How? When its training incorporates distinctly human creative signals: novelty and originality, diversity, surprise, and quality.
In the second study, researchers assessed 14 widely-used large language models (including GPT-4, Claude, Llama, Grok, Mistral, and DeepSeek) using two recognized creativity tests: the Divergent Association Task (DAT) and the Alternative Uses Task (AUT).
Their findings were eye-opening:
No progress: Contrary to expectations, there was no increase in creative performance among these models over the past 18–24 months. GPT-4, in fact, scored lower than in earlier studies.
Above average: On average, all models outperformed the average human in terms of creativity, with GPT-4o and o3-mini ranking highest.
Still not exceptional: Only 0.28% of AI-generated responses reached the top 10% of human creativity benchmarks.
In short, AI has already surpassed the average person’s creativity—but it’s still nowhere near the top 10% of humanity’s creative minds.
Until recently, creativity—especially in the arts—was seen as one of the last strongholds of human superiority over machines. Lately, that belief is being challenged. But we don’t have to panic.
In my view, creativity can be protected and practiced: through open-mindedness, diverse reading (when I felt myself slipping into an AI-only reading loop recently, I reopened André Breton’s Anthology of Black Humor, and it already sparked unexpected thoughts and creative plans), through questioning, doubt, silence and the noble art of staring at the ceiling.
These are the things we must show children. Not by lecturing or preaching (how well did that work when we were teenagers?), but by example.
What do you think? How is AI affecting our creativity?
V.
AI now makes much better... otters
Academic Ethan Mollick, who I’ve written about in previous newsletters, has a daughter. Since she was little, she’s loved otters—those charming, nocturnal, water-loving mammals.
In February 2022, while traveling often by plane, Mollick started using Generative AI to create images of otters—specifically with the prompt: “otter on a plane using wifi.” Recently, he shared a collage here on Substack with AI-generated otter images, from early 2022 to spring 2025.
I think that collage tells the whole story of GenAI’s progress in image creation—in just six images and three years: from a blob of fur in an unrecognizable aircraft tail in 2022, to a fully realistic animal in April 2025.
VI.
Six words
I almost forgot.
At the end of May, I attended the Innovation Northern Greece 6.0 conference.
Many talks were interesting (I shared some takeaways on my LinkedIn profile), but I especially noted six words from Dimitris Gerogiannis, president of AI Catalyst.
Before starting his talk, he asked how many attendees use ChatGPT. Nearly every hand went up. He followed with Gemini, Claude, Grok, DeepSeek—fewer, but still many. Then came his next question—the six words that stuck: “How many of you use Mistral?”
Not a single hand went up.
Mistral is a powerful large language model, developed not in the U.S. or China—but in France. (Try it—it works well!) And it’s not the only European one.
Yet in Europe, we barely use it—perhaps we don’t even know it exists.
And i believe that’s a problem.
Let me explain.
Many of us want the EU to support AI development for business purposes. Many of us believe that regulatory efforts in Brussels are impressive—but rightly, not enough, since countries that fall behind in AI development will also lose their geopolitical and economic edge. Not just because they’ll fail to attract future tech giants or top minds—but because those who don’t speak the “language” of this new era may ultimately lose the capacity to create culture altogether. How can we claim to shape culture with real social impact, if we don't adapt to the world around us? Nevertheless, Europeans often don't use—or even know about—European AI products and services.
What do you think? Should that change? Do you think it can change? How much do you know about Mistral? Do you use it?