Original video in Chinese.
Key Takeaway
- The rise of an AI subculture: text-to-video is driving the deconstruction of authority; abroad, Bigfoot challenges real footage, while in China Sam Altman PUA’s historical figures and challenges narrative authority. Like Bilibili barrage comments and bizarre remix videos, the margins are challenging the traditional.
- Timing of technical maturity: Google/OpenAI have solved the problems of voiceover, consistency, and camera movement, turning video from a toy into a medium of expression; it is closer to the public than tools or images, and power is shifting toward creators.
- Milestone significance: a key signal that AI is moving toward the masses; entrepreneurs and investors must pay attention. Looking ahead to next year, there will be IP revenue sharing and massive battles between real people and virtual characters. This marks the first AI subculture in three years of revolution.
If you are an AI entrepreneur or investor, then you absolutely need to pay attention to the AI subculture that has emerged this year. This subculture is being driven by text-to-video technology. I think it is an important signal that AI is moving toward the masses, and a key milestone.
Have you noticed that in the past couple of days, there have been a lot more AI-generated workplace videos? The main character is Sam Altman, who is wildly PUA’ing Newton, Darwin, Zhuangzi, and many other historical figures. The lines are all the workplace jargon we know so well.
For example, PUA Li Bai by saying that his “suspected to be frost on the ground” user data still isn’t enough, and he should run an A/B test to see whether “suspected to be snow on the ground” will get a higher click-through rate.
Or PUA Lu Xun by saying that his A Madman’s Diary is sufficiently critical, but lacks constructiveness and does not convey positive energy.
I took a look, and the data for this kind of video is all very good. Judging from the number of likes, each video gets several hundred thousand views.
That’s the situation in China. In fact, abroad, this direction had already gone viral a few months ago. One representative work is the Bigfoot video. The creator used these animals to act out and satirize stories from our daily lives, and it also got huge results.
Why do I say they represent an AI subculture?
I think whenever a subculture rises, there is one obvious characteristic: the deconstruction of authority.
What authority do Bigfoot videos deconstruct? The authority of reality.
They are shaking the foundation of “image = fact.” At the time, Google launched Veo 3, pushing the effect of text-to-video to a truly realistic level for the first time. The implication of the Bigfoot videos is: what you used to believe as evidence, I can now generate with one click. This is a fundamental challenge to our perceptual system and trust in media.
What authority do videos of Sam Altman conversing with historical figures deconstruct? The authority of narrative.
They are shaking notions like “historical figures = sacred and inviolable” and “ideas = serious texts.” The implication is: the authority and grand narratives you revere, I can place them in an office and casually joke about and parody them. This is a challenge to cultural idols and ideological orthodoxy.
If this feels hard to understand, think back to when Bilibili emerged, and barrage comments and bizarre remix videos rose up. It was exactly the same.
What authority did barrage comments challenge? The “interpretive authority” of works.
Before Bilibili, the way we watched film and television works was “broadcast-style.” TV stations and streaming platforms aired them, and viewers could only passively receive them. The tone, emotion, jokes, and tear-jerking moments of a work were all preset by the creator. That was a kind of authority.
After barrage comments appeared, they stripped creators of the absolute right to interpret their own works. The audience’s real-time emotions and secondary interpretations became part of the work itself, and even influenced the final meaning of the work.
Bizarre remix videos are even more aggressive; they directly open fire on the content itself and the authority behind it. I won’t go into that here.
So you see, the rise of every subculture is a dynamic process in which the margins challenge the center, and the new challenge the traditional. And “deconstructing authority” is an inevitable way for these subcultures to fight for room to exist and for legitimacy.
It’s been almost three years since this wave of the AI technology revolution began. This is the first time a subculture capable of representing AI has emerged. So why now?
Because the technology is ready.
Among various AI technologies, only text-to-video is closest to the public. Think about it: generating text is still mainly a tool function, and it is still far from mass cultural frenzy; generating images—static images—are inherently weaker than video in narrative and emotional impact.
To truly explode at the mass level, it has to be video. And this year, Google and OpenAI have solved the three key problems of text-to-video:
First, adding sound to video. This is equivalent to moving from “silent film” to “talkies.”
Second, maintaining consistency in a character’s appearance and voice. That means the birth of characters;
Third, including camera movement, multi-shot transitions, and shot design in the video. That is the foundation of storytelling.
Once these three problems are solved, AI-generated video crosses the C-end threshold and goes from a “technical toy” to an “expressive medium” that anyone can easily pick up. The power of content creation begins to shift from “people who possess production skills” to “people who possess brilliant ideas.”
I saw the latest news: Sam Altman is considering an IP revenue-sharing model to bring in more traditional partners. By next year, AI-generated video may well become a gigantic showdown between IP and real people, virtual and reality.
OK, that’s it for this episode. If you want to understand AI, become a super individual, and find like-minded people, come join our newtype community. See you next time!