Prompt Shared by a Domestic Creator Spreads Rapidly
AI Users Apply It to Memes and Profile Pictures
OpenAI Adds "Scribble" Style Template to ChatGPT

"Draw the attached image in the world's most trivial, sloppy lines (...) Enough, just draw it however you want."


A scene from the TV entertainment show 'Infinite Challenge' reimagined in a 'trivial' image style. Garden Rush thread

A scene from the TV entertainment show 'Infinite Challenge' reimagined in a 'trivial' image style. Garden Rush thread

View original image

AI prompts that create so-called "trivial" images are heating up among AI users. At the end of last month, Korean creative director "Garden Rush" shared a command on his social networking service (SNS), calling it "the most trivial prompt in the world." He generated images of popular scenes from the TV variety show "Infinite Challenge" and cats, using only black lines—like something roughly drawn in Microsoft Paint with a mouse. Contrary to the recent AI image generation trends that compete for pixel-level sophistication, he simply told the AI, "Just draw it however you want," leaving the decision entirely to the system.


This kind of command spread rapidly across Threads and X (formerly Twitter). Users have been transforming and sharing various memes and profile pictures in the trivial style. Even Sam Altman, CEO of OpenAI, reposted one of these related posts and changed the profile pictures of OpenAI and ChatGPT’s official social networking service accounts to ones in this trivial image style.


On May 2 (local time), OpenAI also added a "scribble" style to ChatGPT’s official image template. Selecting this template inputs a prompt that asks the AI to "redraw the attached image as clumsily, messily, and embarrassingly as possible." When users attach an image, the result is a rough, crude version. On May 4, Garden Rush shared this update on Threads, saying, "I never thought it would come this far," and added, "It’s all thanks to everyone who enjoyed it." The chatbot "Grok" from xAI has also introduced this prompt as a template.


OpenAI generating Instagram feed images in a 'trivial' image style. OpenAI Instagram

OpenAI generating Instagram feed images in a 'trivial' image style. OpenAI Instagram

View original image

Previously, in March of last year, the release of the ChatGPT-4o image model sparked a major trend of generating images in the style of Japanese Ghibli animation. During the first week after release, approximately 700 million images were generated—roughly 1,200 images per second. According to app analytics firm Appfigures, ChatGPT recorded about 12 million new installs in the month following the launch of the image model.



Meanwhile, leading AI companies such as Midjourney and Google Gemini have recently enhanced their image generation features, intensifying competition in this market. OpenAI released ChatGPT Image 2.0 last month, addressing the persistent issue of broken text in AI-generated images. As each company aggressively introduces price cuts and discount promotions, the competition to secure their ecosystems is becoming increasingly fierce.


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

© The Asia Business Daily(www.asiae.co.kr). All rights reserved.

Today’s Briefing