{"slug": "why-ai-cant-replace-the-creative-struggle", "title": "Why AI can’t replace the creative struggle.", "summary": "A novelist who used an LLM to draft a sequel chapter found the output true to her voice but ultimately insufficient, concluding that AI cannot replace the human creative struggle. The author, who also works at Journey, argues that while AI can generate viable options, the most consequential decisions about character, emotion, and meaning remain the human's responsibility. She advocates for principles that keep human creative intelligence in the driver's seat, emphasizing the value of divergence, process, and tactile experiences over AI's pattern-converging optimization.", "body_md": "This spring, I sat at my desk to start my second novel. Instead, I got writer’s block. It’s a quaint affliction nowadays. Large-language models (LLMs) will generate five viable paths for a character in 10 seconds, making this torment feel almost elective. A few days later, a friend, who’s also a fiction writer, recommended a subscription-based LLM designed for hacks like me, who are fumbling around in the fog.\n\nI paid my dime, uploaded my first novel, and, to test the waters, tried an audacious prompt: “Write the first chapter of the sequel.”\n\nDear reader, it’s been almost four years since OpenAI released ChatGPT, yet the LLM’s output still felt equal parts magic and mind crime. The 15-page passage was true to my voice, pacing, and dialogue. Yet once I traversed that uncanny valley, I realized the most consequential decisions remained: What did I want my characters to do? How should they feel? What did I want readers to take away? For all the words on the page, these questions were mine alone to grapple with.\n\nMy experience ties to a question that many people, but particularly designers, are asking right now: What’s the optimal way to apply creative intelligence—that ineffable thing that we humans possess and practice—to outputs of [artificial intelligence](https://www.fastcompany.com/section/artificial-intelligence)? Since that strange spring day, wrestling with that question has led me, thankfully, not toward an existential ditch, but toward a tight set of principles that keep my creative intelligence in the driver’s seat.\n\nAt this moment of human-machine inflection, these principles buoy and center me in my work. The good news is you don’t have to be a struggling novelist to apply them to your work.\n\nFirst, we need to remember how AI goes about its labor versus how we do. AI optimizes toward a goal using pattern recognition. It’s all about converging on a viable answer to a query. When we apply creative intelligence, we look for patterns, too, but we also free-associate. We diverge. In that divergence we probe the flat-out weirdness of our brains and find something that we’d *choose* to struggle with.\n\nI’m getting at the distinction between process and product. It’s why AI-generated design often fails to create an emotional bond. That doesn’t mean AI can’t stimulate and force us to consider fresh paths.\n\nReal meaning, though, flows through a human’s application of their creative intelligence. Whether it’s building out a pitch deck or sketching concepts for a restaurant, there are innumerable dead ends, discarded ideas, and early scribbles. From this slurry emerges the confidence to make decisions forged in human emotion. Yes, AI can abbreviate the process, but deprioritize it at your own peril.\n\nAn LLM generated a viable chapter. That hasn’t spared me hand-wringing over story arc, character motivation, or conflict. In the intervening months, the AI-written chapter has assumed the role of provocation, not blueprint. It sent me reeling in the best possible way. I don’t feel any less of a writer; rather, the exercise affirmed where I should apply what creative intelligence I can claim, and now I have a clearer sense of where this book is going.\n\nA thorny creative brief, a paragraph that’s just not cohering, a stubborn shadow on the stage’s proscenium—these struggles test our nerves. They send us on wrenching journeys of self-doubt. But we wouldn’t embark on them if they weren’t so damn satisfying.\n\nWhen my colleagues at Journey design an experience, they’re considering how it will make people feel across all their senses. That sometimes requires stepping away from the laptop and taking a walk through a park, or visiting a gallery, or painting a chair—not in the vain hope that the muse will speak, but because staying close to nature and tactile crafts leads to better work. Nerves can be all too real, and so is the satisfaction of calming them through actions that reaffirm our humanity.\n\nMuch of the discourse around AI, particularly in design, implies a division of cognitive labor that’s largely correct: The AI does the automatable things, the human deploys their creative intelligence. We stake out the emotional ground we want work to land on and point the AI that way.\n\nStill, it would be naive to ignore how good AI models are getting at this calculation. Whether it’s *Billboard*[ singles](https://www.theguardian.com/technology/2025/nov/13/ai-music-spotify-billboard-charts) or [artwork catching millions](https://www.bbc.com/news/articles/cpqdvz4w45wo) at auction, novelty is wearing off and acceptance is settling in. This requires us to focus on what feels undisputedly, inseparably ours, and claim it loudly. What that reclamation looks like will vary across industries, but in my writing, it means that I may use AI to sharpen my thinking and provoke me, but my name must earn its place on the jacket cover.\n\nTo await some grand convergence between creative intelligence and AI is counterproductive—because it’s already happening. Rebecca Winthrop, the director of an education-based AI project at the Brookings Institution, [has found](https://www.nytimes.com/2026/05/27/opinion/writing-creativity-ai.html) in her research that AI “can make it difficult to distinguish where the user’s thinking ends and the bot’s begins.” That boundary grows blurrier with every release from the big AI platforms.\n\nInstead, why not exalt what makes us human? Exercising creative intelligence, even toward an AI-generated output, reminds me that I possess embodied experience and emotional awareness—in short, that along with 8 billion fellow travelers, I am alive.\n\nAI is a remarkable story, but don’t ever mistake it for the protagonist in yours.\n\n*Andrew Zimmerman is CEO of *[ Journey](https://journey.world/).", "url": "https://wpnews.pro/news/why-ai-cant-replace-the-creative-struggle", "canonical_source": "https://www.fastcompany.com/91588485/why-ai-cant-replace-the-creative-struggle", "published_at": "2026-08-12 12:00:00+00:00", "updated_at": "2026-08-12 12:36:57.330405+00:00", "lang": "en", "topics": ["generative-ai", "large-language-models", "artificial-intelligence"], "entities": ["OpenAI", "ChatGPT", "Journey"], "alternates": {"html": "https://wpnews.pro/news/why-ai-cant-replace-the-creative-struggle", "markdown": "https://wpnews.pro/news/why-ai-cant-replace-the-creative-struggle.md", "text": "https://wpnews.pro/news/why-ai-cant-replace-the-creative-struggle.txt", "jsonld": "https://wpnews.pro/news/why-ai-cant-replace-the-creative-struggle.jsonld"}}