📅 2025-06-23
🕒 Reading time: 4 min
🏷️ Creative 🏷️ AI Image Generation 🏷️ ChatGPT 🏷️ Claude 🏷️ Gemini
"We have to create dozens of banners every week."
In the afternoon sunlight streaming through Baker Street, our visitor was the clearly exhausted Sportia Promotional Manager. In her hands were colorful banner samples.
"We operate sporting goods stores nationwide and sell through e-commerce platforms. There's always some sale or campaign running..."
Her voice carried the particular rush of modern marketing environments.
"The in-house image production workload has reached its limit. Plus, sudden decisions like 'let's do an emergency sale this weekend' come down, and each time we work late into the night to meet deadlines. It's an endless cycle."
Her exhausted expression told the story of structural problems plaguing modern creative workplaces. Work that should be creative had become time-driven labor.
"A fascinating modern dilemma, Watson," Holmes observed, studying Sportia's promotional banner inventory by the fireplace. Sale banners, seasonal product banners, social media banners, store display banners—an incredibly diverse range produced without rest.
"'Deadline' is always creativity's enemy," Holmes shook his head. "Creation should naturally be a process requiring time to mature, yet modern business won't allow it."
Examining the details revealed the problem's structure. While content varied, composition and color patterns followed certain rules. Product categories, sale types, target demographics—these combinations determined patterns to some degree.
"Here lies the thread to our solution," Holmes's eyes gleamed. "Time for generative AI, our new 'creative partner,' to enter the stage."
"If 'what should be created' is standardized, then 'who creates' need not necessarily be human."
Holmes stood and began drawing creation process diagrams on the whiteboard.
"Traditional creative processes assumed 'creating something from nothing.' But promotional banners often amount to 'optimizing combinations of existing patterns.'"
His proposal was revolutionary:
AI-Powered "Draft Creation" System - Image-generating AI automatically presents 3 patterns with different compositions and colors - Staff select and make minor adjustments as needed - Accumulate past banners and sales performance as training data
"More fascinating still," Holmes turned around, "training AI on past sales performance enables developing AI assistants that understand 'designs that sell.'"
I was impressed. Indeed, patterns that humans intuitively recognize as "good design" must contain analyzable rules.
I opened my investigation notebook to organize this new creative process's possibilities.
Category | Traditional Human-Led | AI Collaborative New Model | Future Integrated Type |
---|---|---|---|
Keep (Values to Preserve) | • Dynamic brand tone for sports • Fast-paced promotional culture • Intuitive field sensibility |
• Human final judgment • Brand consistency maintenance • Flexible client requirement response |
• Essential value of creativity • Human-like sensitivity • Strategic thinking capability |
Problem (Challenges to Solve) | • Image production over-dependence and ad-hoc response • Management complexity from sale type increases • Creator burnout syndrome |
• AI training data quality management • Ensuring stable quality of generated content • Adapting to changing human roles |
• Reconsidering creativity definition itself • Optimal AI-human division design • New creative evaluation standards |
Try (Next Experiments) | • AI automatic image "sketching" generation • Format + color pattern knowledge systematization • Phased AI introduction to production process |
• Sales-linked design learning system • A/B testing automation for effectiveness measurement • Creator training curriculum renewal |
• New job categories for human+AI collaboration • Higher-dimensional creativity expression • Business outcome-linked creative work |
"I see," Holmes nodded with satisfaction. "This isn't mere efficiency—it's creativity redefinition."
"If pure '0→1' creation is unnecessary, then 'AI-powered high-speed reproduction' is most efficient."
Holmes contemplated deeply while studying AI-generated banner proposals on the monitor. The screen displayed color combinations no human might conceive.
"This color scheme suits summer sales," he murmured with a smile. "Interestingly, I've become not a 'creator' but an 'appreciator.' I'm experiencing the fundamental shift of becoming the 'proposal recipient' in creative work."
I sensed profound implications in those words. Traditionally, creators generated something from nothing. Now, they were transitioning toward "curators" who select optimal solutions from countless options.
"Human true value is shifting from 'creating' to 'judging,'" Holmes continued. "AI mass-produces, humans select—this may be the new form of creation."
As evening light streamed into the office, I reflected on the client's final question.
"But can we really call this 'creation'?"
Holmes's response was clear.
Creative work is evolving from 'making' to 'choosing.'
Like museum curators who create moving experiences by selecting from countless works for exhibitions. Like film directors who weave stories by choosing optimal cuts from vast filmed material.
AI draws, humans decide.
In this new creative process, human work ascends to pure "judgment." And that judgment may be where true creativity manifests.
The changes occurring at Sportia's promotional department preview the future of the entire creative industry. Technology moves hands in place of humans, while humans concentrate on higher-dimensional decision-making—such a future approaches.
The "temptation" of generative AI may actually be "liberation" guiding humans toward more human-like creativity.
"True creation may be selecting the single answer from infinite possibilities"—From the Detective's Journal