Artificial intelligence didn’t break into mass use through labs or enterprise tools. It happened through cartoons.
Weeks after OpenAI gave ChatGPT-4o the ability to craft stylized pictures, many of them resembling Japanese animation, usage exploded. That shift — more than any press release or research breakthrough — may have pushed AI past its turning point with the public.
Sam Altman spoke about the surge at a recent TED event. He gave no fixed number but hinted the base has doubled since the upgrade. Before, he’d mentioned 500 million people used it weekly. Now it may be closer to a billion. Or around 820 million. He suggested something similar (that is "10 percent of the world").
The numbers are hazy, but the direction is not. Growth began just as users started producing images echoing works by Studio Ghibli. What began with soft-lit fantasy characters quickly turned into other viral themes. Lately, people have been designing mock-ups that resemble high-end toys or action figures.
That style shift brought up a harder question. If users create things based on artists’ visual signatures, who gets paid?
Altman didn’t give a solid answer. He said the company might, in time, connect certain prompts with opt-in payments. Artists could sign up and receive small cuts. Nothing’s live yet. But the system has blocks in place that stop copy-paste imitations.
That part matters. Some missed a legal angle: the images aren't always illegal. In Japan, Article 30-4 allows AI developers to train on copyrighted work without permission. Copying an entire character or film shot would still cross the line. But generating a dreamy look, or putting someone’s face into a hand-painted frame? That’s still allowed, unless it's sold.
Under Japanese law, recreating an artistic style is not infringement. Matching a specific drawing is.
That’s where the line stands for now.
Image: DIW-Aigen
Read next: Google Chrome Blocks Cross-Site History Tracking Exploit Hidden in Link Colors
Weeks after OpenAI gave ChatGPT-4o the ability to craft stylized pictures, many of them resembling Japanese animation, usage exploded. That shift — more than any press release or research breakthrough — may have pushed AI past its turning point with the public.
Sam Altman spoke about the surge at a recent TED event. He gave no fixed number but hinted the base has doubled since the upgrade. Before, he’d mentioned 500 million people used it weekly. Now it may be closer to a billion. Or around 820 million. He suggested something similar (that is "10 percent of the world").
The numbers are hazy, but the direction is not. Growth began just as users started producing images echoing works by Studio Ghibli. What began with soft-lit fantasy characters quickly turned into other viral themes. Lately, people have been designing mock-ups that resemble high-end toys or action figures.
That style shift brought up a harder question. If users create things based on artists’ visual signatures, who gets paid?
Altman didn’t give a solid answer. He said the company might, in time, connect certain prompts with opt-in payments. Artists could sign up and receive small cuts. Nothing’s live yet. But the system has blocks in place that stop copy-paste imitations.
That part matters. Some missed a legal angle: the images aren't always illegal. In Japan, Article 30-4 allows AI developers to train on copyrighted work without permission. Copying an entire character or film shot would still cross the line. But generating a dreamy look, or putting someone’s face into a hand-painted frame? That’s still allowed, unless it's sold.
Under Japanese law, recreating an artistic style is not infringement. Matching a specific drawing is.
That’s where the line stands for now.
Read next: Google Chrome Blocks Cross-Site History Tracking Exploit Hidden in Link Colors
