iQIYI “Flips the Table”: Why Do Platforms Always Want Us to Watch AI

A speech by iQIYI CEO Gong Yu recently sent shockwaves through public opinion.

The platform hopes to use AI technology to digitize actors' images, enabling them to participate in more content production without increasing filming intensity. But Gong Yu's statement was rather radical: traditional live-action filming may even become "intangible cultural heritage" in the future.

Anxiety over being replaced by AI, combined with the distaste for "assembly line" content, has jointly fueled a strong backlash in public discourse.

But the squeeze of AI content on live-action filming is not just a simple technological imagination. In the short drama sector, the impact and replacement of live-action short dramas by AI avatar comic dramas is happening right now.

In February this year, the Hongguo platform eliminated the minimum guarantee mechanism for small and medium producers, and the number of short drama crews plummeted. At the same time, during the Spring Festival, nearly thirty percent of total short drama views belonged to AI comic dramas.

In the past year, AI content has entered the public eye almost through a "forced viewing" approach.

From AI comic dramas to AI avatar short dramas, from platforms intensively rolling out incentive policies to independent apps and dedicated content sections constantly being launched, platforms are pushing AI content at a pace visibly faster than users’ natural acceptance.

What are platforms rushing for? What are they competing for?

The Content Supply Explosion: Whose Table Did AI Overturn?

The anxiety behind companies like iQIYI is that the "odds" concept is becoming increasingly important in the content industry.

The traditional business model for long-form video platforms is essentially a B2B logic: the platform pays producers for content and then monetizes through C-side subscriptions and brand ads.

Over the past decade, the core of this model has been "betting on blockbusters," investing massive capital in professional production agencies, gambling that they'll make the next big hit like "The Knockout".

Companies like iQIYI assume the greatest financial risk and opportunity cost, but cannot truly control production efficiency, and supply is always limited by the industry's own capacity ceiling.

A failed S-tier drama could mean hundreds of millions in sunk costs.

What's worse, audiences are constantly drifting away, making this "high odds" business increasingly hard to sustain.

According to QuestMobile, long-form video’s share of total user time online dropped from 17.8% in 2023 to 11.3% in 2025.

Short dramas, meanwhile, soared from 2.7% to 10.8%. The gap between the two nearly vanished within only two years.

The business model behind free short dramas is speeding onto the mainstream stage.

Content itself is free; income comes from feed ads and pay-to-unlock features. Platforms are no longer footing the bill for all content, but instead providing traffic channels and taking a cut from every transaction.

For producers, the core metric for short drama ad spend is ROI—ads spent versus user payments—so long as ROI is greater than 1, spending can be sustained indefinitely.

The survival rule in the content industry is shifting from "pile money into content" to "betting on hit rate with volume".

Producers are no longer betting on whether a single work succeeds, but whether, under sufficiently low costs, a few hits will emerge to erase all sunk costs for the whole business.

However, the shadow of "involution" is always overhead.

As supply keeps increasing and competition grows fiercer, the hit rate of individual projects is doomed to steadily decline. Under quasi-industrial production logic, continually lowering costs and expanding scale becomes an almost unstoppable inertia.

At this moment, AI comic dramas born from technological breakthroughs are to short dramas what short dramas once were to long-form video—a new industry migration in pursuit of extreme ROI.

According to data from Juyin Engine, AI can cut average production costs in the comic drama sector by over 70%, boost efficiency by more than 80%, and shorten production cycles by two-thirds.

Traditional 2D animation often costs tens of thousands per episode, while AI comic drama unit costs have dropped to only a few thousand per minute. Examples of "3 people producing 10 episodes in 10 days" have emerged, and even "1 person making 1 drama in 1 day" is being discussed.

AI comic dramas have also satisfied market expectations with numbers.

In the first quarter of 2026, AI avatar dramas had 75 billion views. The views for a single AI avatar drama already exceed the combined total of all AI dramas and comic dramas from last year.

Once content supply grows exponentially past a certain tipping point, companies like iQIYI essentially have no other choice.

Once supply and audiences start to flow, whoever doesn’t pick it up, it will flow to their competitors.

Short-Term Gains, Long-Term Strategies, and Platform FOMO

AI is not just about face-swapping or simple special effects, but a complete reshuffling of production methods.

What platforms are fighting over now isn’t just the content itself, but the entire chain of power surrounding content production.

The first layer of competition is a land-grab at the supply end.

As a new species, comic dramas in the early stages of supply explosion just happen to give platforms an opportunity window to re-divide the landscape. Each platform has rushed to increase split ratios and support, trying to pull high-quality supply into its own ecosystem.

As early as October 2025, iQIYI announced its comic drama cooperation project, offering up to 100% revenue share for exclusive content.

According to LatePost, Tencent Video, to prioritize profits, brought the booming AI comic drama business into the group level.

In February this year, Youku updated its anime revenue split rules, adding a single-project reward of up to 1 million yuan for premium projects; China Literature opened up 100,000 IPs for comic drama adaptation, and set up a special fund of 100 million yuan for creative projects.

The second layer of competition is seizing control over the traffic allocation valve.

Once supply floods the pool, users' attention becomes all the more precious.

Platforms that master user preferences, split ratios, and traffic allocation gain stronger discourse power.

This snowball can theoretically keep rolling: the richer the content supply, the denser the traffic cluster, the higher the ad revenue; the more traffic is concentrated at the head, the easier it is for the next wave of quality supply to follow.

ByteDance is particularly strong in this playbook.

In December 2025, daily ad spend for Douyin comic dramas surpassed 20 million yuan; by March 2026, ByteDance comic drama daily ad spend further surpassed 70 million yuan, overtaking live-action short dramas.

Even if a platform loses in the traffic war, it still retains room for an IP trial run.

Beyond traffic battles, AI avatar dramas are becoming a bridge between core anime fans and general entertainment users.

Major companies have walked the "IP leads, all eat" old map for years, with sizable web novel reserves—ByteDance has Tomato Novel, Tencent holds China Literature, Baidu has Seven Cat.

But for a web fiction work to truly transform into a blockbuster IP, it often must go through the long relay of "novel, comic, animation, live-action film and TV".

The further it goes, the more cash goes in, the longer it takes, and the greater the risk of losing everything with one misstep.

Low-cost AI technology has truly shortened this relay race course.

Not only does it give "male-oriented" and "fantasy" genres, which are hard for traditional TV and film, a visual outlet, but also allows countless IPs that would otherwise gather dust in warehouses a chance to “test the waters.”

What platforms are desperately trying to retain is actually a group of audiences ready to leave at any time. If the other side's content is denser and ad spend more aggressive, this part of the traffic could well be snatched away again.

QuestMobile data shows that in December 2025, average daily time per user for micro-short dramas reached 129 minutes, officially surpassing long-form video.

Even if everyone complains about AI content being unnutritious, or short dramas lowering IQ, traffic numbers never lie.

Platforms are swept up in this upheaval, feverishly pushing AI content — not to stubbornly persuade users to accept a new form, but for fear that if they miss their footing in this supply tsunami, they'll be swept off the boat entirely.

Risk Warning and DisclaimerThe market is risky; investments should be made cautiously. This article does not constitute personal investment advice and has not taken into account individual users' special investment goals, financial status or needs. Users should consider if any opinions, views or conclusions in this article fit their particular situation. If investing based on this, responsibility lies with you.