author | not empty
Unpopular singer Stefanie Sun is "retiring", but AI Stefanie Sun has quietly become the traffic password of the entire network.
"In one night, the number of views reached more than 100,000. After posting less than ten videos in a row, the number of fans has already exceeded 1,000." Xiao Su, who has been an UP host at Station B for less than a month, is one of the behind-the-scenes traders of AI Stefanie Sun , obviously, she is very satisfied with her content data, "It is already very good that the new account can achieve this level."
"In four hours, the number of views on the entire platform exceeded 500,000." Qiqi is also creating content for AI singers on platforms such as station B and video account. The growth of content data continues to stimulate her enthusiasm for creation.
The technological changes triggered by ChatGPT spread rapidly, directly impacting the field of content creation, and even turning AIGC into an important part of the content platform. AI Stefanie Sun and Jay Chou, who emerged on Bilibili and Douyin, have a great tendency to revive the Chinese music scene. It is undeniable that the change of technical tools has once again lowered the threshold of content creation, and this kind of more participatory and even novelty-seeking content will soon flood the entire network.
Image source: Screenshot of station B
However, behind the fanaticism of the whole people, AI-generated content has always had big original problems-copyright disputes and content security.
Kuaiyin founder Li Yan took AI Stefanie Sun as an example to explain, "Everyone will see AI Stefanie Sun covering Jay Chou's songs on the content platform, but this kind of content not only violates Stefanie Sun's own timbre, but also involves infringement of the song. This problem is still very serious in the music industry."
On May 9th, Douyin released an initiative on artificial intelligence-generated content, prohibiting the use of generative artificial intelligence technology to create and publish infringing content, including but not limited to portrait rights, intellectual property rights, etc., and will be severely punished if found; Publishers are also required to prominently identify AI-generated content, plus a unified official "watermark".
The wildly growing AIGC is about to usher in the first baptism of industry self-discipline.
Image source: Generated by Unbounded AI
01
AI Stefanie Sun became popular quietly
"It's very easy to make this kind of content in the past one or two months. AI Stefanie Sun's video has also been popular on station B." In addition to station B, on short video platforms such as Douyin and Kuaishou, AI singers are becoming more and more popular. bigger.
Xiao Su has already summed up the logic of a set of popular AI singers, "Generally, it is the combination of the king and the queen, such as AI Stefanie Sun covering the classic songs of Jay Chou and Tao Zhe." The playback volume on the site is as high as more than 800,000. In addition, Xiao Su said that using the singing style of Tian Wang Tian Hou to cover the popular Douyin Divine Comedy is also easy to get out of the circle because of the novelty and contrast. AI Jay Chou's cover of "Black Plum Sauce" is one of the cases.
300 online viewing, picture source: Screenshot of Station B
AI cover singing is not a new concept, but the transformation of AIGC technology has really fueled the explosion of AI singers. Guo Jing, CEO of Time Domain Technology, explained the creation process of AI singers, "AI singers mainly involve two technologies: singing voice synthesis (SVS, singing voice synthesis) and singing voice conversion (SVC, singing voice conversion). The AI singer directly sings the music score input by the user. The voice conversion is to input the voice of an "original singer" and convert the original singer into the voice of the target singer, which is similar to the voice version of the AI face change. At present, the open source model of voice synthesis and conversion is not It's hard to find, Xiaosu uses the overseas project "so-vits-svc".
Among the crowd of AI singers, the reason why AI Stefanie Sun can stand out from the encirclement is not only Stefanie Sun's unique timbre, but also technical logic. Li Yan revealed that Stefanie Sun's voice was the first voice to be open-sourced. "The open-sourcers first used Stefanie Sun's voice to show their own solutions, which naturally attracted a group of users who followed suit." He said that the earliest time on March 11 , Discussions about AI singers began to rise. In early April, someone open-sourced Stefanie Sun's timbre.
However, despite the high popularity of AI Stefanie Sun's video, Xiao Su does not feel safe. "Every video introduction needs to be clearly marked. The video content is generated by AI tools and is not for commercial use. Infringement will be deleted immediately." In addition to the creator's own active labeling, some platforms will also give warnings at the bottom of the video to remind users to identify AI content. Behind this is naturally the consideration of copyright and security. "Copyright supervision requirements are getting stricter, and I am worried that one day the platform will remove all my content."
Disclaimer for a video profile Image source: Screenshot of Station B
At the same time, how to maintain the popularity of AI Stefanie Sun is also a problem faced by creators. In Guo Jing’s view, the popularity of AI Stefanie Sun is more about consuming the concept of AI. “For most users, the fact that AI can make Stefanie Sun cover Jay Chou’s songs is more important than whether the song itself is good or not. Realized in the AI world This kind of 'lifetime' is very attractive to users, and the result of its popularity is naturally expected."
At the same time, Guo Jing also affirmed that from the perspective of long-termism, AI Stefanie Sun has verified the significance of AI technology as a productivity tool. Expand the diversity of content space".
In addition to copyright anxiety at the creator level, the platform naturally needs to take corresponding responsibilities for AI content. "Neither the record company nor the singer will 'cut' against ordinary users, because the user's behavior is of a strong entertainment nature, and even these users are fans of the singer. However, the singer can hold the content platform accountable." Responsible for a certain music product People explained the reason why Douyin vigorously supervises AIGC content.
According to this logic, other content platforms are bound to speed up to catch up with the pace of regulation, and the brutal growth stage of AIGC content has finally come to an end.
02
Watermarking, labeling, AI generation needs to be warned
In this wave of artificial intelligence, the source of human sense of crisis is the rapid evolution of AI, which even has the ability to confuse the real with the fake.
Two months ago, the AI painting, which was criticized for its incomplete hand details, has been iteratively upgraded in a short period of time. Not only can it compete with Taobao models for jobs, but it can even wear the vest of a good-looking blogger. The true face of AI, but fell into it.
During the evolution of AI, the concept of "digital life" has also begun to truly come to fruition. In the past, technology apes used AI to revive deceased relatives and appeared on hot searches. Later, there were conversations between AI Musk and AI Jobs across time and space, and AI Li Bai and AI Lin Daiyu's poetry debate.
Image source: Video account AI.talk
These confusing contents not only have innovations in content form, but also have a sense of freshness to technological changes, and naturally become the "traffic password" of the entire platform.
Kuaiyin has counted the user recognition of AIGC content. According to Li Yan, Kuaiyin mixes the music content of the AI version and the content of the original singing version, and distributes them to users in a unified manner. Judging from the data of the user end, 30-40% of content users have better indicators than the original version.
"With the support of current technology, the quality of AI-generated content is already high enough. At the same time, AI-generated content is still very rare and will bring freshness to users. Therefore, users' feedback on AI-generated content will be high. "
However, in the environment of fissile development of Internet communication, AI-generated content that is difficult to distinguish between true and false makes the information environment more complicated. After technology sinks into the context of the whole people, uncontrollable factors also increase. It can be said that the higher the popularity of AI-generated content, the greater the hidden worries it will cause, and the impact on privacy and content security boundaries is unprecedented. Therefore, the supervision of content platforms has become extremely important.
At present, the solution given by Douyin is to add watermarks and mark them uniformly. Specifically, in terms of content form, the text content needs to be clearly marked with the words "AI generated", and the video content also needs to add a watermark of a specified ratio on the canvas to highlight the AI logo. During the push process, the platform will also add a notification label "this content is generated by AI" in the information flow, video details page and other parts.
Image source: Public account of Douyin Rules Center
The explicit watermark has become the first layer of "gatekeeper" on the dissemination link of AI-generated content. It reminds users to treat the content with caution in the form of a logo, and can avoid the chaos caused by AIGC in the content field to a certain extent.
Regarding the AI logo, Li Yan added an observation. He found that after adding words such as "AI creation" and "AI generation" to the title or text, the user data dropped significantly. For content creators, how to deal with the impact of this watermark on content has become a problem that has to be considered.
For the platform, before AIGC content supervision has not yet formed a clear regulation, this iconic "watermark" ended the brutal development of AIGC content creation and formed a relatively safe content creation environment within the platform. At the same time, this can also be seen as a disclaimer in disguise for the platform, exempting it from the risks of AI-generated content.
03
AIGC speeds up on the right track
Douyin fired the first shot of AIGC content supervision.
On May 9, the short video content platform Douyin released an initiative for artificial intelligence-generated content, in addition to requiring publishers to clearly mark the artificial intelligence-generated content and add a unified official "watermark". Douyin has also strengthened the constraints on virtual people, requiring virtual people to register on the platform and receive an exclusive "ID card". At the same time, virtual human technology users also need real-name authentication.
As an industry that already has a considerable scale, Douyin’s supervision of virtual people is not untimely. The outbreak of virtual humans can be traced back to 2020. The development of AI, cloud computing, 5G and other technologies has unlocked the content value, marketing value and even the value of industrial development of virtual humans.
However, as a new thing, a series of problems inevitably arise in the process of virtual human beings moving towards industrialization. Coupled with the fact that generative AI technology has reached the public level, the governance of virtual humans has become more complicated.
The first is general personality rights such as portraits and voices related to virtual people. With the support of AI drawing tools, the technical threshold is lowered again, and it is not difficult to design a set of virtual images exclusive to ordinary users. However, the convenience of technology has blurred the boundaries of rights, AI technology has become a tool, and user-personalized content available everywhere on the Internet has become a material. Eventually, the chaos of abusing the image of others and data security has become an accompanying problem of technology openness.
At the same time, after the advent of the era when everyone has a virtual human, the entry threshold for content creators will be lowered, which will easily lead to uneven quality of virtual human live broadcasts, videos and other content.
"On the whole, the management and control of Douyin will mainly affect some small companies and ordinary users who are wandering in the gray area." Bi Ming, a virtual human operator, analyzed the owner, "In order to save costs, these small companies and Ordinary users will take shortcuts in technology and regulation before, and even gain a place in the content, which is a common way of playing.”
In Birming’s view, “Real-name authentication has little impact on virtual humans who have formed a mature operating model. On the contrary, it will effectively clean up the development environment of virtual humans and eliminate the harmful factors affecting industrial development. Let virtual people move towards healthy development."
The measure of real-name authentication is actually to use the power of the platform to screen the producers of virtual human beings, to govern the virtual human industry with lower and lower thresholds, and to form a virtuous circle of content ecology.
From the initial enthusiasm under the technological change to the novelty-seeking early adopters, and finally to the new form of content that has become a benign development, AIGC is bound to go through a painful period of development and be constrained by the power of supervision before it can finally move towards a formalized development path.
Babbitt Park is open for cooperation!
Chinese Twitter: https://twitter.com/8BTC_OFFICIAL
English Twitter: https://twitter.com/btcinchina
Discord community: https://discord.gg/defidao
Telegram channel: https://t.me/Mute_8btc
Telegram community: https://t.me/news_8btc
