We hear this wave of panic all the time now, that AI is coming for writers and a lifelong talent in creative writing, especially for those who do it for their living, is a thing of the past. Algorithms can write poetry, draft novels, and evoke human emotion in a matter of seconds. I decided to test this theory, but I didn’t want to test the machine against my current 26-year-old self.
I wanted to run a chronological experiment.
I took the exact topics I wrote about in my childhood notebooks at ages 7, 10, 11, and 22, and fed them to a Gemini AI pro version, Then, I compared its billion-dollar neural network to the scribbles of a kid. What I found wasn’t just a lesson in writing; it was a glaring exposure of exactly what a machine lacks.
The first test was from when I was seven. The topic was simple: Toys.
I asked the AI to write a short poem about toys. It gave me a perfectly rhymed, incredibly boring inventory list. Here is what the AI spit out:
‘The wooden blocks are stacked in rows,
The painted soldiers stand in line.
A wind-up train in circles goes,
A teddy bear with button eyes that shine.
They wait upon the playroom floor,
For little hands to bring them life anew.
Unlocking magic through the door,
In painted worlds of plastic, red and blue.’
It was mathematically metered, safe, and ended happily. But here is what I actually wrote in July 2007:
‘When the night comes beckoning
And the sun bids farewell
The toys come to life
On their way down town
The party lights are turned on
And pop music sounds
As the dolls bring lemonade
And cake from the town
Every one clears up the tables
And sit down to eat
But sadly then the sun comes up
And the poor toys have to stay hungry’
An AI trained on millions of children’s stories will always default to a magical ending where toys are brought to “life anew.” But even at seven, I inherently understood narrative tragedy. I understood that time is an antagonist, and that stories are infinitely more compelling when there is unfulfilled desire. A machine doesn’t let the toys starve; a real child knows that magic has limits.
By age ten, the topic was “Family.”
When I gave this prompt to the AI, it immediately defaulted to philosophical metaphors. It tried desperately to sound profound, comparing a family to a “sturdy tree” and a “harbor.” Here is the AI’s attempt:
‘A family is a sturdy tree,
With roots that run so deep,
A harbor in the stormy sea,
A promise that we keep.
We share our laughter and our tears,
Beneath the same roof’s dome,
Through all the changing of the years,
The heart will find its home.
A guiding light when days are dark,
A shelter from the cold,
Together we will leave our mark,
As memories unfold.’
And here is what my 10-year-old self actually cared about:
‘Oh dear mummy
You are so sweet
You are the only person
Whom I will do anything to meet
Oh dear daddy
You are so courageous and brave
Away from the cruel world
Me you protect and save
Oh dear granny
You are so old and wise
You always comfort me
When you hear my cries
Oh dear grandpa
You are so kind and strong
Whenever you go to the market
To buy my favourite fruits, You take me along
Oh dear brother
You are very cute and nice
But the awful thing is
You keep me awake with your cries
Oh dear Timmy
You are the most adorable dog I have seen in my life
You are faithful and obedient
I hope I have the best master been’
To an algorithm, love is an abstract concept like a “guiding light.” To a ten-year-old, love is highly specific: it’s your grandpa remembering your favorite fruit and taking you to the market. Furthermore, an AI writes about generic family warmth, but a real kid writes the unvarnished truth: “You are cute, but the awful thing is you keep me awake with your cries.” The machine mimics emotion; the child documents reality.
At eleven, I wrote about seeing snow for the first time. The AI gave me a stock-photo description. It observed the snow from a safe, theoretical distance:
‘The sky turned pale, a muted gray,
And chased the autumn wind away.
Then silently, without a sound,
A million crystals hit the ground.
Like fragile stars of frosted white,
They danced within the fading light.
I held my breath and caught a flake,
And watched the frozen world awake.
It wrapped the trees in silver lace,
And left a chill upon my face.
A pristine blanket, soft and deep,
That sang the weary earth to sleep.’
But my 11-year-old self wrote from inside the snowstorm:
‘As an icy layer formed
On top of the pond
Small white icicles
Formed on my nose
I shivered as I saw
That a wet blanket of snow
Had covered the garden
Sparkling, pristine and slippery
I build a snowman
And little snow angels
I played till I froze
And reluctantly had to go in
I took a sip of hot cocoa
As I sat by the fire
And remembered the glorious evening
The evening I saw snow.’
An algorithm does not have a nervous system. It doesn’t know what it feels like to have your breath freeze into icicles on your own face. The AI described the snow as “soft and deep.” I described it as “sparkling, pristine and slippery”—because a kid actually playing in the snow knows you are going to fall down.
For the final test, I skipped ahead to age 22. I didn’t just give the AI a one-word prompt. I gave it a massive, highly specific structural blueprint based on a poem I wrote about violence against women.
I ordered it to use my exact refrain (“The [woman] could have made it home safely; / Almost…”). I told it to write about marital rape, a baby in a gutter, and an acid attack contrasted with the holy immersion of the Goddess Durga.
The AI followed my instructions, but the result was exhausting. Because I told the AI in my prompt to “create a sharp cultural antithesis,” the machine literally dropped the phrase “a sharp cultural antithesis” right into the middle of the poem.
Here is the AI’s bloated version:
‘Dust settles on the banyan leaves as she walks through the twilight smog,
Deaf to the rustle of the dry Lantana scrub,
Blind to the sour, fermented breath of country liquor,
As an intoxicated whisper slithers out from the tangled bushes.
The young teenager could have made it to her home safely;
Almost…
Swirling ice against crystal, the bass thumping against her ribs,
She spins in scarlet silk, a modern-day princess lost in the music,
Tasting the sweet citrus of her crafted cocktail,
Completely unaware of the bitter drop of oblivion,
And the cold, sinister smirk carving the bartender’s face in the shadows.
The woman in the red dress could have made it to her home safely;
Almost…
Tiny, dimpled fingers wrap tight around her uncle’s calloused thumb,
Toddling toward the sugar-spun promise of a brightly colored toy,
A trusting, innocent joy swallowed by the thick humidity of the afternoon,
Replaced by a horrific, sickening silence,
Where a small, stained shroud now floats among the waste in a choked monsoon gutter.
The innocent baby could have made it to her home safely;
Almost…
Gold bangles stack heavy with the pride of a pure, arranged union,
Neighbours whispering in bitter envy at the lavish feast and jasmine garlands,
Yet Indian society wraps its hypocrisy in heavy Kanjeevaram silk,
As the law turns a deliberate, blind eye to the suffocating terror of marital rape,
And pristine white sheets are violently painted crimson.
The newlywed bride could have made it to her home safely;
Almost…
Scripts memorized, top of her class, her posture a steel rod of ambition,
She pushes open the frosted glass door of the casting room,
Believing the spotlight awaits her rehearsed, flawless monologues,
Stepping unknowingly past the threshold into a predator’s velvet-lined lair,
Where talent is just raw meat for the men holding the keys.
The confident acting student could have made it to her home safely;
Almost…
Devout men chant hymns, immersing the clay idol of Goddess Durga into the holy Ganga,
Worshipping the divine feminine, a sharp cultural antithesis,
To the sudden, caustic wave of acid thrown by the hands of mortal men.
The flesh dissolves in a searing, unthinkable hiss,
Melting away the proud nose that resembled her dad, and the gentle eyes of her mom.
The devoted daughter could have made it to her home safely;
Almost…
Do not tell me it is the hour, or the skirt, or the unchaperoned walk,
For the rot is in the soil, the poison spinning in the very axis of this planet.
How many more hollow, echoing almosts must we stack against the sky,
Before the earth itself becomes a home for women?
Look into the sun and confess the bitter, bleeding truth:
It is not a crime for a woman to simply exist
It didn’t know how to subtly weave a contrast into imagery; it had to explicitly announce what it was doing. It was terrified of empty space, stuffing every line with heavy adjectives—”heavy Kanjeevaram silk,” “tangled bushes,” “fermented breath”—completely destroying the breathless dread of my original piece.’
Here is my original 22-year-old poem:
‘The veil of twilight caressed the golden sunset glow
Her footsteps quickened, furrowed brow
Then from the bushes an eerie intoxicated whisper
The young teen could have made it home safely;
Almost…
She twirled around in her red dress, a princess in her head
Sipping on her cocktail, with a mind free of dread
Little did she notice the bartenders sinister smirk
The young woman could have made it home safely;
Almost…
Baby eyes lit up, innocent joy
As she held the uncle’s finger, who had promised her a toy
Days later the gutters cradled a little shroud afloat
The baby girl could have made it home safely;
Almost…
How lucky she is, envious voices chime
Her parents beam, riding high on societal pride sublime
Blind like the law to the white sheets that turned crimson that night
The new bride could have made it to ‘her’ home safely;
Almost…
Brimming with confidence, the topper of her acting school batch
She swept into the casting room, her competitors were no match
Unaware of the predator whose lair she had entered
The budding actress could have made it home safely;
Almost…
The Pujo is complete, the idol of Durga the men immerse
The waves of the Ganga causes the goddess to submerse
Elsewhere a caustic wave melts away the nose that resembled her dad, her mom’s eyes
The bold woman could have made it home safely
Almost…
How many more almosts must it take
For the world to stop calling it her mistake
How many more almosts must pass through
Before this planet becomes ‘almost’ a home to women too
How many more almosts before we insist
That it’s not a crime for a woman to simply exist!’
This experiment taught me the greatest, unspoken secret about generative AI in art.
If you want a flawless, perfectly rhyming poem that sounds like a corporate Public Service Announcement, the AI will write it in three seconds. A massive sigh of relief for people struggling with basic grammar, spelling or mindless clerical typing. But if you want to capture the visceral human emotions of terror, the fantasies of a child brain nurtured by literature as varied as Enid Blyton books and tinkle comics, or the simple yet familiar annoyance of a wailing younger brother, you are on your own.
AI can mimic structure perfectly. However thats precisely what it is in, 2026 in its present version- a flawless mimic. It lacks what makes a human experience of reading a book so vividly enjoyable, the truth that this was someone else’s literal lived experience, not a mere concoction of words strung together by a robot. As my 7-year-old self proved a long time ago, a machine cannot simulate a soul. Will it soon be able to do so with more improved versions of technology and billions of dollars being spent on improving LLMs to make them more humane- well only time can tell.
My final conclusion- well AI is indeed a valuable supportive tool just as a chisel is in the hands of a master sculptor. I too did use AI to help with my formatting and generating images including for this very post. However the imagination, the emotions and the shared thread of what makes art palatable and humanly relatable – that soul shall always belongs to a beating human heart rich with the experience of the highs and lows of human existence.