[Jihye Choi's Trend] In an Age of Ubiquitous Output, the Human Touch Sets You Apart

Go beyond asking what to automate
Focus on what people can do better
AI brings efficiency... humans bring emotion

The globally successful film "The Odyssey" was made with a production budget of approximately $250 million. A substantial portion of that went toward practical production and on-location filming. A large wooden ship capable of sailing was used as Odysseus's vessel, and the actors learned to sail and row themselves. The Trojan Horse was built as a full-size prop, standing 10.7 meters tall and weighing more than 3 metric tons, and the actors performed inside it. Director Christopher Nolan says that the unpredictability of physical spaces that cannot be controlled as intended adds tension and a sense of realism to the film. Audiences admire this approach to filmmaking because they value the human effort and craftsmanship that go into creating the finished product.

In the past, calling something "as perfect as if it were made by a computer" was a compliment. But things have changed as generative artificial intelligence (AI) can produce text, images, and videos in seconds. The faster and cheaper it becomes to create something, the harder it is to stand out on the strength of the finished product alone. Instead, what a person has experienced and chosen, where they have made a contribution, and what decisions they have made and taken responsibility for become points of distinction. The human touch is becoming a premium in the age of AI automation, and this trend can be seen across a range of fields.

First, imperfection is increasingly appreciated as a distinctive quality in its own right. Instant camera brand Polaroid ran its "The Best of Summer is Analog" campaign in 2026 in South Korea, the United States, the United Kingdom, and other markets. In South Korea, it also promoted the message "Be Pixel Imperfect." Instant photos are not popular because of their image quality. Digital photos can be edited even after they are taken, but with instant photos, there is no undoing the result, even if the image is a little blurry or overexposed. Because they cannot be edited, their imperfections are embraced as part of Polaroid's distinctive character. According to 29CM, the transaction value of Polaroid products rose by more than 114% year over year in May 2025.

Second, the process of creating something is itself becoming content, a trend worth noting. In 2026, electric toothbrush brand Quip released a surreal advertisement featuring an actor wearing a mouth-shaped headpiece and holding an oversized toothbrush. Because the video was so meticulously made, consumers suspected it might be a fake created with generative AI. Quip released footage showing the production of the sets and props and the filming process, declaring, "No AI, just us." The finished ad alone could not convey that it had been made by people; that became clear only after the production process was shown. British streaming service BritBox also conveyed authenticity to consumers in 2025 by fully revealing how it built 11 physical sets and filmed a 90-second ad in a 14-hour single take.

Third, there is a growing trend of emphasizing the human touch as a factor that determines service quality. In 2025, Starbucks introduced Green Dot Assist, a generative AI-powered work support tool, at 35 stores in North America. Instead of having baristas search recipes and manuals one by one, the tool lets them ask AI and get an immediate answer. The goal is not to reduce staffing, but to ease employees' workloads so they can focus more on preparing drinks and serving customers. At the same time, Starbucks is strengthening its "Green Apron Service," which includes welcoming customers, making eye contact, and remembering their names. AI handles repetitive information processing, freeing up time for employees to spend with each individual customer.

Behind these changes are fatigue with and distrust of AI-generated output. In 2025, BetterUp Labs and Stanford University's Social Media Lab introduced the concept of "workslop." It refers to AI-generated work that looks polished at first glance but is so lacking in substance that it leaves colleagues to redo it. Even if AI can quickly produce a report, recipients have to check and revise it themselves if it lacks context and fact-checking. In a survey, about 40% of U.S. office workers said they had received workslop.

Another contributing factor is that AI's capabilities vary from task to task. A joint research team that included Harvard Business School conducted an experiment on AI use with 758 BCG consultants. On tasks in areas where AI performs well, participants who used GPT-4 produced higher-quality results more quickly. On tasks beyond AI's capabilities, however, the group that used AI had a lower rate of correct answers. Because it is not immediately clear when to trust AI and when to question it, the role of people who verify the results and take responsibility for them becomes more important.

This does not mean that we should reject technology. It makes sense to leave tasks that AI does better, such as repetitive work, information searches, and data analysis, to AI. The global brand Stanley 1913 offers a useful example. Stanley uses AI-powered analytics to improve efficiency in its supply chain and operations. However, it has adopted a policy of not using generative AI for advertising materials that consumers see directly. The company uses AI to boost internal efficiency while ensuring that work done by people remains visible at customer touchpoints. Going forward, the competitiveness of companies and individuals will likely depend less on how much they use AI and more on their judgment about what to delegate to AI and what people should do. The question we need to ask now is not only "What should we automate?" but also "What is more valuable when people do it, even though AI could?"

Choi Jihye, co-author of Trend Korea

This content was produced with the assistance of AI translation services.