The Real Cost of AI: Copyright, Creativity, and the Future of Literature-Big Journeys by Kavishala
Can artificial intelligence write the next great novel—or is it quietly reshaping creativity in ways we don't fully understand?
Artificial Intelligence has transformed the way we create content. From blog posts and marketing copy to poetry, novels, and film scripts, AI writing tools are becoming increasingly sophisticated. Some believe they democratize creativity, while others worry they threaten the future of human expression.
In a compelling conversation, poet and lyricist Puneet Sharma explores one of the most important debates surrounding AI today: not whether AI can write, but what it costs to build these systems in the first place. His perspective goes beyond technology, focusing instead on copyright, ethics, originality, and the future of literature.
Before We Talk About AI's Future, We Must Talk About Its Foundation
Most conversations about AI begin with productivity.
Puneet Sharma believes they should begin with copyright.
According to him, many large language models have been trained on enormous collections of books, articles, poems, essays, and creative works without directly compensating the writers whose work became part of that training data. He argues that this raises a fundamental ethical question before we even discuss AI's capabilities.
His concern isn't simply technological.
If literature helped train intelligent systems, what responsibility exists toward the people who created that literature?
The Ethics Behind Artificial Intelligence
Technology has always changed how humans work.
The Industrial Revolution automated physical labor.
Computers automated calculations.
The internet transformed communication.
AI is now beginning to automate parts of creative work.
Sharma acknowledges that technological progress is inevitable. However, he argues that innovation cannot ignore the rights of creators. Without ethical foundations, technological advancement risks becoming exploitation rather than progress.
This discussion is increasingly relevant as publishers, authors, artists, and AI companies continue debating copyright laws worldwide.
Can AI Replace Human Creativity?
One of the biggest misconceptions surrounding AI is that it can simply replace writers.
Sharma offers a more nuanced explanation.
Large language models learn from enormous collections of existing text. Because they absorb both exceptional writing and mediocre writing, their output often reflects the statistical middle rather than genuine originality.
In other words, AI doesn't invent literature from lived experience.
It predicts what words are most likely to come next.
That distinction matters.
According to Sharma, AI-generated writing may satisfy many readers, but truly extraordinary literature demands emotional depth, personal experience, and artistic risk—qualities that emerge from human life rather than probability calculations.
Why AI Often Feels Convincing
Many AI-generated poems and stories sound polished.
That's because they imitate familiar patterns.
Sharma explains that if an AI model learns from millions of examples, it naturally produces writing that resembles the average quality of its training data.
For most readers, that average may feel impressive.
However, experienced readers who have spent years engaging with great literature often notice subtle weaknesses in language, rhythm, emotional progression, and originality.
The difference is similar to hearing a technically correct performance versus witnessing genuine artistic expression.
Literature Is More Than Well-Arranged Words
Throughout the conversation, Sharma repeatedly emphasizes curiosity.
His writing emerges from asking difficult questions rather than accepting convenient answers.
He argues that simple answers rarely reveal the whole truth because reality is inherently complex. That same philosophy shapes literature.
Great poems are not merely collections of beautiful sentences.
They are explorations of uncertainty, contradiction, memory, identity, grief, love, and imagination.
For the whole conversation, check out:
