The development of any new phenomenon is fraught with twists and turns – old eras, old things, and conservative people all create obstacles. As a new form of artistic expression, AI naturally encounters various predicaments. AI doesn’t just have its cool aspects; dark clouds loom over it.
The first difficulty is AI’s “original sin”—training data scraped without permission and training processes that fail to share profits. This leads to the moral high ground occupied by anti-AI activists, which has consequently created a large number of anti-AI people, especially prevalent within and outside the art industry, primarily in foreign countries.
Voices opposing AI emerge constantly. From large language models to paintings to video, people always criticize AI artists for various reasons – from copyright infringement to poor quality, from lacking soul to being inhuman. I won’t say that the wheels of history roll forward relentlessly because no one can guarantee they’re on the historical train rather than the tracks.
I’ve been opposed by quite a few anti-AI people myself. Their rhetoric is surprisingly consistent as if they’ve all memorized the same script. Although I’m highly annoyed by anti-AI people, as long as they don’t come to my comment section, it’s fine. Where there are new things, there are people who oppose them.
While we forge ahead, some people always want to stay in the past. However, the birth of AI has permanently shattered past creative methods, making people reconsider what technology and art are. I will always believe that new forces will defeat the old era.
Besides anti-AI people, many platforms don’t fully support AI creation either – I won’t elaborate. While AI lowers the barriers to creation, it inevitably leads to uneven quality. Some people think AI is producing cyber garbage.
This relates to AI’s current technical limitations. Everything is still in its early stages—too many shots cannot be made, and too many emotions cannot be expressed. So the films produced, even when edited artistically, all look like mashups, though they’ve moved beyond the early PowerPoint feel.
Currently, AI-produced content with realistic styles has a strong uncanny valley effect. People find it hard to accept videos that look like real people whose expressions and movements aren’t entirely human. Although Veo3 has made significant progress, it’s stronger at text-to-video. For real storytelling, we still rely on image-to-video.
Combining AI with traditional 3D works better currently. Without this combination, too many camera movements can’t be achieved. I would like to know if future developments like world models or further advances in multimodal models will improve this situation. I’m not blindly promoting AI because I’m a practitioner—I know its limitations.
Then there’s AI’s cost-reduction and efficiency-improvement capabilities, which aren’t as cheap as some people imagine. AI costs are higher than current market live-action short drama costs, and production cycles are longer. There’s no real cost or time advantage, although there is a price advantage compared to high-effect sci-fi movies.
AI can demonstrate its cost advantages when depicting sci-fi, fantasy, and other imaginative stories. This is undoubtedly a huge opportunity—authors who previously wrote fantasy stories can finally have an affordable way to adapt their work for film and television.
Then, difficulties are encountered when forming AI teams. Detailing requires high personalization, with each member needing specific directorial abilities. It’s not just demanding for directors. Since I haven’t formed a team yet, this is speculation on my part. AI is somewhat similar to traditional CG production—it requires stable teams.
Everything is still in the pioneering stage, with nothing obvious, especially AI short dramas’ monetization capabilities. But in the second half of this year, we’ll see quite a few well-produced AI short dramas. We can look forward to them and try to participate.
Will audiences watch AI films? This depends on whether we’re still stuck in the past stage of making low-budget bad movies. Doing serious work and telling stories, which must be said, is something we can achieve in the AI era. Many other creators I know also don’t want to use AI to make bad movies.
Whether we’ve made bad movies in the past or not, every new AI film is a new adventure. Imaginations that were impossible in the past are now becoming reality. We are pioneers, pathfinders, and creators of a new world, and no difficulty will make us retreat.
Follow me and explore AI filmmaking together.
