While you're still working on the AI shorts, the AI feature has gone gold on Hunan TV.
On August 31, a television series with no real actors and no real-life footage was shown on the big screen.
The characters, scenes, and character portrayals were all completed by AIGC, with the first season consisting of 30 episodes, 40 minutes each, for a total of 1,200 minutes. After the first five episodes of "Fruit Mountain Story" aired on Apple TV, the series simultaneously landed on Hunan TV and the "The Drama Look Good" large-screen slot.
How can we not say that the capital market is already boiling?
On August 31, the AI video, short drama games and cultural media sectors collectively strengthened, with Apple's supermedia, Chinese online, "20CM" in Hong Kong culture shut down, China's broadcasting industry shut down, the BONA film industry shut down, and many others.
It's not hard to see why the market is so excited, because if AI can stabilize the long-running series, the platform's classic IP, copyright libraries, and storytelling, they can be transformed into content at a lower cost and faster rate.
But from technical demos to real TV series, there is a real threshold in the middle.
AI can make 30 episodes of long content, but can't get the audience to follow up on 30 episodes with a question mark.
One summer, from 60 minutes to 1200 minutes.
AI video has started to accelerate the trend towards longer content, which is actually happening this summer.
On July 23, AIGC became the first AIGC network storyline film to be licensed and officially streamed by AIGC in China.
It led the way across the compliance threshold for AI-length content, with AI-generated characters and scenarios supporting an hour-long narrative, as well as completing submissions, audits and platform releases.
On August 22, the anime adaptation was released in Japan. The two films were each over an hour long and were directed by Guo Yusung-yu as artistic director.
Compared to the original AI movie, the anime's anime is set in front of a more difficult audience.
The 2014 anime series has been airing for 12 years. The show is familiar with the look, sound and behaviour of waitresses, winters and monkeys, and AI has found a very specific place here.
In addition to authorizing the use of personal imagery and voice acting, the AI allows the three characters to return to their early youth and help the team achieve high-cost scenarios such as dreams, exotic beasts and magical warfare.
Unnamed, a former butler who succeeds in "slimming back to the first season", is still reactive and behaves abstractly; Winterfell and Monkey still clash with his familiar destiny. The AI version complements the previous lives of several people, creating a resonance with the real-life version of Monkey's Soul Shift.
In the real-life version of the anime, the ten-year-old manga and the AI version of the anime revolve around a butterfly dream.
For the Old Man, familiar actor voices, character relationships, and storytelling undermined some of the cracks of AI imagery.
Also, AI has not changed the core part of the anime's soul-transition. Little Mercy Day continues to be driven by the fascination, regret and choice of the story-mappers, and the butterfly dream continues the habit of going one step further than the original story.
The market feedback was also more positive than many anticipated. The two films were released in three days, grossing over 3 million yuan at the box office; it surpassed the 6,600 mark on the Hot Movie Chart, and became the highest-grossing film of 2026 on the Hot Movie Chart. The Soul Swing trailer confirmed the possibility of viewers paying for AI long content.
From the swinging of the soul to the swinging of the West, the change is not only reflected in the passage of time.
From single web storylines to series, from web platforms to television golden archives, from post-production to "side production, side auditing, side broadcasting", AI-enabled video production and broadcast systems are becoming more and more complete.
In just over a month, domestic AI video connectivity has crossed three thresholds of continuous production of certified broadcast, classic IP commercial verification and feature films.
What's up with the content?
Currently, post-war western tourism still consists of a large number of short shots. Short content changes, mainly due to characters, scenes, costumes and style of footage, can now be preserved as a sustainable reusable asset.
Models such as Seedance 2.0 and Seedance 2.5 have improved multi-modal reference, character consistency, and local editing capabilities. Creators can first create character libraries, scene libraries, and then continuously generate, filter, and modify footage.
AI video production is a process of production that can be managed and modified through the use of luck-driven "draw cards".
Another important change comes from the long video platform.
Entertainment, Tencent Video and Apple TV already have a lot of IP, copyrighted material, scripting director resources, auditing experience and broadcast channels. They don't need to build the video industry from scratch, they just need to embed AI into the original production chain.
As a result, the competition for AI feature films is also changing.
The ability of video models is still important, and every link is essential for mature IP and creative teams, including platforms that ultimately deliver content to the audience.
After 12 years of accumulation of personalities and emotions, it is a case in point that the entire system behind Apple TV, from production, auditing to broadcasting on the Internet, laid the foundation for the development of the entire industry.
At the same time, the mechanism of "side production, side auditing, side broadcasting" further amplifies the advantages of AI.
Real-life scenes or plays need adjustment, which often means replaying, replaying, time and budget will also increase step by step. The advantages of AI plays in this regard are very prominent, and subsequent content can continue to be adjusted based on audit opinions and market feedback.
Do you think AI can win the audience 40 minutes?
There is still a huge gap between how much content AI can produce and how much content viewers want to see.
In the first quarter of 2026, the Chinese Association of Internet Audiovisual Societies published a guideline for the creation of short dramas, which shows that about 128,000 short dramas were uploaded to the industry in a quarter, of which about 122,000 were AI short dramas, accounting for more than 95%.
AI has become one of the most hard-working producers in the film industry.
But by the 2026 Spring Festival, the number of streams of AI dramas is about 50 times the number of live-action dramas, and the total is only 1/25th of live-action dramas.
Production wins and attention loses.
In short videos, beautiful images, exaggerated mirrors or characters like Fallout can attract a click. But the demands of the audience for a feature-length and a short-length play differ, and the mentality of "throwing your head is watching" is clearly not true in the field of feature-length drama.
AI faces, stiff movements, inadequate syntax, rhythmic drag and repetition of footage, all the disadvantages of AI video have been amplified over time.
At the same time, public confidence in AI content remains low.
Following the broadcast of the western tourist attraction, the commentary section said: "This must be a script written by AI".
In fact, the post-apocalyptic western trilogy is based on the eponymous fantasy novel of the same name from the end of the Qing Dynasty, with 40 books in total, and is often listed as one of the three sequel books to the original trilogy. Here, "All AI Production" refers mainly to the images and characters portrayed by AIGC.
As long as the work is labeled "all AI", the problems with the plot, lyrics, rhythm and character shaping are almost all attributed to AI.
This "stress" is not without reason. Early AI videos accumulated the impression of cheap, rigid, streamlined production. So when viewers learned of the use of AI in a play, they would be more sensitive to looking for anomalies in the picture.
Maybe the model doesn't do a good job of making an appearance for the time being, but with what shots and how the play should go, the production team will finally be applauded.
Real-life photography does not lead directly to "a good play", and using AI naturally cannot be the only explanation for a bad play. Producers cannot simultaneously create a dissemination theme with "all AI" while hiding behind "technical constraints" when the work is criticized.
Since the AI feature has gone on to become a TV Gold Record, it should receive the same ratings as the live-action series.
And part of the feedback that's been generated by the souls of the dead has shown that technology can be a tool for IP to continue to grow while familiar writers, actor voices, personality relationships and emotional nuclei still exist.
This may point to a more exploratory direction for AI. While trying to hide traces of technology, AI can also attempt to enter real-world videotape-heavy voids.
The game allows the characters to return to the way they were twelve years ago, as well as to present their imaginations of the past, subject to scenes, special effects and budgets. After the game, the western traveler chooses a magical theme, as well as the ability to play the AI to create monsters, fantastic landscapes and fantasy worlds.
The first full-length animated computer-animated feature film in the world.
At the time, CG technology was difficult to naturally present skin, hair and other organic materials, and human characters were easily hard and plastic-sensitive. But computers were particularly good at showing blocks, metals, plastics, and the geometry of rules.
Thus, Pixar chose "Toys" as the lead character. CG's plastic feel for humans was flawed, while the bass light years and toys were just fine, and the technical limitations were perfectly combined with the story.
The early days of new technologies were difficult to outperform the old ones on all indicators, and the most effective approach was to find a category of content that particularly suited it.
Now, instead of expecting the audience to say, "It's not like AI", perhaps we should strive for another exclamation: "A story like this, AI is the right thing to do".