The era of "one person, one drama a day" has arrived, but the oversupply of AI short dramas is not the end | Exclusive interview with JURU Jeff

Before the Spring Festival in 2026, the wave set off by Seedance 2.0, a video generation model under ByteDance, is propelling AI dramas towards large - scale implementation. Short drama companies caught in this upheaval may be panicked, anxious, or enthusiastic about new opportunities.

Jeff, the founder of Juriulu, an AI short drama creation platform, predicts that March and April will be a critical juncture for the industry. The production capacity of "one person, one drama a day" will become a reality, and the production cost of commercially - viable short dramas will be significantly compressed. Subsequently, the cost structure will be completely reversed: the proportion of labor costs will plummet from 80% to 20%, and computing power and tokens will become the core costs. An oversupply will arrive as scheduled in two or three months, but this is not a sign of the decline of the track. Instead, it means that the segmented market demand will be fully activated, and the overall scale of the industry will continue to expand.

In his view, this transformation goes far beyond efficiency improvement. The industry logic is undergoing a fundamental change, shifting from "following the list" to "returning to the users themselves", and from "platforms seeking content" to "content seeking platforms". Interactive movie - games are also ushering in a new era.

Amid the complex emotions, how should one seize the opportunities in the current situation? Recently, we had a chat with Jeff, the founder of Juriulu.

Jeff, the founder of Juriulu

The following is the edited dialogue between 36Kr and Jeff, the founder of Juriulu

A certain scale of clusters can support the production and consumption of film and television content

36Kr: What are your core predictions for the AI drama track from 2026 to 2027?

Jeff: I believe that from March to April this year, the production capacity of "one person, one drama a day" will emerge, and we will also launch corresponding technological capabilities. The overall production cost of a short drama that meets the above - average, commercially - viable standard will be compressed to the level that can be completed by one person in one day.

This will bring about a major subversion of the cost structure. Previously, in the cost of a drama, technological costs such as computing power and tokens only accounted for 20%, while labor costs accounted for 80%.

After March and April this year, this structure will be completely reversed: the BOM cost of dramas will drop significantly. The core is that labor costs will be greatly compressed, and finally, technological costs such as computing power and tokens will account for 80%, while labor costs will only account for 20%.

I would like to emphasize again that what I'm talking about here is not top - tier, super - high - quality content, but dramas that meet the commercial standards and are of above - average quality.

36Kr: What does it mean to meet the commercial standards?

Jeff: You can understand it as content that is below super - high - quality but above poor - quality and worthy of commercialization. If you want to evaluate it, there are definitely many flaws. After reaching this level from March to April this year, an oversupply will be achieved in the next two or three months.

36Kr: Does achieving an oversupply mean that this track is not worth pursuing?

Jeff: No. This track will continue to be worth pursuing because individuals have a very strong demand for such personalized entertainment content. The overall scale of the AI drama track will expand, and the themes will become quite rich.

36Kr: That's true.

Jeff: So around May this year, the thinking of the industry in drama production will completely change. Previously, when people produced dramas, they basically focused on whether they could make it into the top ten of the list, studied the themes on the list, and imitated the successful experiences of top - tier dramas, and created with this as the goal.

After another two or three months, when the cost of producing one drama per person per day drops significantly again, the entire industry's business logic must change. Using the old thinking will definitely cause problems.

By then, the core thinking will no longer be to follow the trends on the list but to truly return to the users' tastes, content quality, topic selection, and material selection itself.

Take sports as an example. Previously, users only watched "Captain Tsubasa" and "Slam Dunk". But if it comes to the sport of high - jump, it was too niche. But around May, a certain scale of hobby clusters will be able to support the production and consumption of a film and television content.

Currently, the thinking of people in content production is still centered around how to create a hit. This is the underlying logic. And that so - called "hit" refers to the top - ranked content on the list. Everyone is focusing on the supply for a large - scale population, looking for the pain points and commonalities of the large - scale population. But later, driven by AIGC, people will look for niche hobbies or the pain points of segmented groups.

36Kr: Will there be any changes in opportunities at home and abroad?

Jeff: I think there will be opportunities for the emergence of content platforms with segmented categories and tastes both at home and abroad. Because the supply cost is very low, by the end of 2026, the production cost of a drama may be only 10,000 yuan.

And 5,000 pieces of content can support an independent platform. Vertical content platforms (such as platforms dedicated to art films and retro hobbies) that could not exist due to high costs before will have opportunities.

From "platforms seeking content" to "content seeking platforms"

36Kr: What changes will occur in the purchasing and guarantee logic of content platforms in the next year?

Jeff: From the end of 2026 to the beginning of 2027, except for extremely top - tier content and super - creative studios (only 0.01%), the guarantee and purchasing logic in the industry will basically disappear. Previously, platforms would provide guarantees to CP parties. In the future, not only will there be no guarantees, but CP parties will also need to promote and buy traffic on platforms like Hongguo when they want to launch content, shifting from "platforms seeking content" to "content seeking platforms".

Currently, the industry has started to cancel some guarantees. The core driving factor is the reduction of supply costs and the sharp increase in supply quantity brought about by AI. This trend will be fully implemented from 2026 to 2027.

36Kr: Is the cancellation of guarantees the result of AI promotion?

Jeff: Fundamentally, it is AI, but actually, it is about supply. The reduction of supply costs leads to an increase in supply quantity.

36Kr: How to understand that "guarantees for super - top - tier content will still exist"?

Jeff: No matter what day in the future, there will definitely be guarantees for super - top - tier and extremely high - quality content.

36Kr: What is the cost difference between AI - simulated human dramas and real - person live - action dramas?

Jeff: Previously, the cost of shooting a real - person drama exceeded 500,000 yuan. Now, the cost of AI - created content with a real - person painting style has dropped to less than 200,000 yuan. After March this year, the cost of AI dramas will enter the "ten - thousand - yuan era", and 5,000 - 50,000 yuan can produce a drama. The specific cost depends on the enterprise's organizational ability, the advancement of tool usage, and resource reserves.

Compared with December last year, the cost of AI dramas has been reduced by more than half. And computing power and tokens will gradually become core resources. Just like energy, they will become scarce when demand exceeds supply.

Interactive movie - games are presenting opportunities

36Kr: Besides production efficiency, what new gameplay and theme trends are there in the AIGC content track in 2026?

Jeff: From May to June this year, new things and new gameplay will emerge, and interactive movie - games will become a trend. In the early stage, the plots of interactive movie - games were pre - generated. By the end of 2026 and the beginning of 2027, real - time post - generation combined with AI capabilities will become the mainstream. Users will no longer need to choose plots at the nodes set by the official and can intervene at any time, even determining the plot direction in real - time. The interactivity of the content will be fully maximized.

36Kr: Interactive movie - games are not something new. Why couldn't they be developed before but can now?

Jeff: The core is that the AI technology has reached a certain level of accumulation, and the cost has dropped significantly. Interactive movie - games are indeed an old - fashioned gameplay. The reasons why they couldn't be developed before were: first, the branch plots could only be pre - generated, resulting in limited interactivity; second, the production cost was too high to support the production of multi - branch and personalized plots. Now, AI can achieve real - time post - generation, and the cost has dropped sharply, making the commercialization of interactive dramas possible. The dual breakthroughs in technology and cost are the key.

36Kr: Will there be some new explorations in the model?

Jeff: Previously, it was the thinking logic of the Internet era. In the AI era, it is more like the logic of a factory. There may even be unilateral content platforms where the content is self - produced, just between you and the users.

36Kr: From the user's perspective, people don't seem to have a high acceptance of AI dramas. They all say that they are soulless.

Jeff: When automatic gearboxes and autonomous driving emerged, they also went through a similar process. People need some time to accept new things.

36Kr: Will those short drama shooting bases also be affected?

Jeff: I haven't specifically paid attention to them. To be honest, I'm not enthusiastic about paying attention to who is dying. Instead, I need to spend time thinking about the rhythm of the next era. Because the era is changing too fast, maybe the next one to disappear will be ourselves. I have a strong sense of crisis, so I don't have the time to watch others' elegy.

36Kr: Do you have FOMO (fear of missing out) emotions?

Jeff: I'm more in an excited state. Anxiety is an emotion brought about by a sense of loss of control and the unknown. I'm more like a person who is surfing with passion and having a good time.

I will enjoy the opportunities brought by this era. Enjoy the game. Don't look back when you're old and think that you were on the cusp of the wave but thought too much, were anxious and couldn't sleep, and made mistakes. Now, just don't leave any regrets. It's best to leave a mark, and it's even better to leave a cool one.

The AIGC track has rapid cyclical fluctuations, but both the peaks and troughs are high. Since we have enjoyed the dividends of the industry's rapid development, we don't need to complain about the pressure of the industry's rapid changes.

In addition, I find that people are a bit overly anxious about the iteration of AI models. The iteration of AI video and visual generation models is too fast, and no model can maintain the top position for a long time. After the technical path is exposed, other manufacturers can quickly catch up. In 2026, no single model will dominate, and the model field will always be in a state of blossoming. The key is to choose a partner who can deeply integrate the model capabilities with the capital production process and make the technology truly serve the content.

36Kr: Which enterprises can survive in the future, and what is the core survival logic?

Jeff: First, an efficient production - type team that can control costs, achieve high production capacity, and has the ability to find the needs of segmented scenarios and segmented groups. They can make a profit by "producing in large quantities at a certain level + precise matching". For example, if a team produces 2,000 AI dramas a year and each drama makes an average profit of 10,000 - 20,000 yuan, it can achieve stable profits. Second, super - high - quality creative studios that can create extremely creative content that stands out. This kind of content will always have a market.

36Kr: I guess many real - person drama companies have come to cooperate with you this year, right?

Jeff: Yes, there are many. The overlap between the people who used to make comic dramas and those who make real - person dramas is already very high.

36Kr: Is there still an opportunity for long - video platforms under the AI narrative?

Jeff: All video platforms have opportunities because equality has been achieved in terms of supply.

Small teams should focus on segmentation and core training of "from 0 to 1" capabilities

36Kr: Before the Spring Festival, you predicted that 2026 would be a very difficult year for real - person short dramas. Now it seems that your prediction is coming true.

Jeff: At that time, I said that 80% of real - person short dramas would be replaced by AI. Many people argued with me, but now no one does. Of course, it's impossible to be 100% replaced. Just like there are still people using Nokia even when the iPhone has reached the 17th generation, there will definitely still be people insisting on traditional production methods.

36Kr: Will both large and small companies have a hard time?

Jeff: In the previous cycle, the larger the real - person drama company, the more difficult it is now. Such companies have large shooting teams and heavy assets, making it difficult to turn around. On the contrary, small and medium - sized teams that previously had no real - person drama resources can quickly seize the opportunities of AI dramas. The supply units in the industry are shifting from large - scale centralized ones to small and medium - sized decentralized ones.

The core of small and medium - sized teams is to abandon the thinking of "creating a hit and making a fortune" and turn to "small - cost, high - frequency, and precise matching". Even if a drama only makes a profit of a few thousand or tens of thousands of yuan, stable profits can be achieved through the accumulation of quantity.

36Kr: So, in the AI wave, how do you think people should deal with the impact?

Jeff: I think the core is to test your ability to keep moving forward. This will challenge the founder's cognitive actions, abilities, and determination. This is a very anti - human and tough state.

Many people said that Juriulu had a smooth progress last year. In fact, we have to fight a new battle every once in a while. If we don't win, we will fall behind. You see, making comic dramas with AI was a new battle from 0 to 1, and then we moved on to making comic dramas with agents. When we established an advantage, I realized that we needed to make AI - simulated human dramas. When we started making AI real - person dramas, we actually couldn't see a clear opportunity, but we just did it. In so many battles, if we didn't keep up in any one of them, we would be out.

36Kr: You are in the tool - making business. How should content - making companies establish their survival opportunities?

Jeff: One way is to find segmented needs and scenarios. You need to have high efficiency and large production volume. The key is your ability to discover opportunities. Once you find an opportunity, take advantage of it and then move on to find new opportunities. Another way is to create extremely high - quality, super - top - tier, and creative content.

36Kr: What changes will occur in the relationship between tool providers and platforms in 2026?

Answer: For tool providers, the opportunities for new players are slim, and the industry space will become narrower. The core work of top - tier tool providers will become doing "arduous tasks" - such as guiding customers to use tools, solving implementation problems, and providing accompanying services, rather than simply engaging in technological R & D.