華人音樂產業風向標「金曲獎」在本月初正式開啟第37屆報名工作，大會特別規定本屆起報名作品允許AI（人工智慧）作為輔助流程工具使用，但禁止作品整體及核心創作部分由AI完全生成。這一新規的確立不僅推動華語音樂產業與前沿科技接軌，更對行業產生深遠影響——華語音樂正式宣告進入與AI共存的全新世代，新篇章已然揭開扉頁。

The Golden Melody Awards, a bellwether for the Chinese music industry, officially opened its 37th year of submissions earlier this month. The awards have a special rule allowing the use of AI as an auxiliary tool in the submission process, but prohibiting the complete generation of works, including the core creative elements, entirely by AI. This new rule not only promotes the integration of the Chinese music industry with cutting-edge technology but also has a profound impact on the industry—officially marking the beginning of a new era of coexistence with AI, and a new chapter has been opened.

生成式AI（Generative Artificial Intelligence）在戲劇及文創行業一直都備受爭議，有一種聲音稱AI在感性創作內容上缺乏章法，缺乏感官靈魂。況且其不斷擠壓真人創作者的生存空間，使得本就不大的行業空間，生存環境愈發艱難。但這並非只是音樂產業的危機，而是在科技世代的大勢之下，與文字、圖像、音頻、視頻、代碼腳本相關的領域都必然受到顛覆性的沖擊。因此，AI已然不再是選擇題，而是音樂產業在新世代實現可持續化發展的必答題。

Generative Artificial Intelligence (AI) has long been controversial in the film and television industry and the arts and culture sector. Some argue that AI lacks the emotional depth and soul of human creation. Furthermore, it continuously squeezes the living space of human creators, making the already limited industry space and working environment even worse. However, this is not just a crisis for the music industry; under the immense trend of the technological age, all fields related to text, images, audio, video, and code/scripts are bound to be disruptively impacted. Therefore, AI is no longer an option, but a crucial question for the music industry to answer in the new era of sustainable development.

需要注意的是，AI帶來的音樂行業變革，並不是追求AI技術的噱頭本身，而是關乎它的商業本質：更低投入成本+更高流程效率+更優落地效果。對於大部分追求商業利益的大陸音樂公司而言，主要瞄準大陸下沉市場（Sinking Market），這類市場人口基數大、消費潛力強，特點是消費者更重視性價比、實用性及情感共鳴。在此前行業不陷入內卷的前提下，這類市場大眾大多數並不會以藝術標準和技術標準作為首要消費考量，更看重「朗朗上口」、「洗腦旋律」、「青春情懷」⋯⋯

It's important to note that the transformation of the music industry brought about by AI isn't about pursuing the hype surrounding AI technology itself, but rather about its commercial essence: lower investment costs + higher process efficiency + better implementation results. For most Chinese mainland music companies pursuing commercial interests, the primary target is the mainland's sinking market. This market has a large population base and strong consumption potential, characterized by consumers who value cost-effectiveness, practicality, and emotional resonance. Before the industry became overly competitive, the majority of consumers in this market didn't prioritize artistic or technical standards; they valued catchy tunes, catchy melodies, and youthful sentimentality...

短影片平台就是主打下沉市場的產物之一。寄生於短影片平台業務的音樂公司，本身並不要求比較高的藝術能力和技術能力，甚至只需要滿足數十秒不同風格的「洗腦」音頻片段即可。在這個商業場景下，如今由AI主導或AI生成的方式，相對低成本就能大量生成一定質量的BGM曲庫，對於商業市場是一個高性價比的優解。甚至，隨著AI的迅猛迭代發展，AI生成的質量也將越來越好，必然會發展到令人不得不刮目相看、甚至驚艷的程度。

Short video platforms are one such product targeting the sinking market. Music companies that rely on these platforms don't necessarily need high artistic or technical skills; they may only need to produce catchy audio clips of different styles lasting a few dozen seconds. In this business context, AI-driven or AI-generated methods can generate a large volume of high-quality background music tracks at a relatively low cost, making it a cost-effective solution for the commercial market. Furthermore, with the rapid iteration and development of AI, the quality of AI-generated content will continue to improve, inevitably reaching a level that is both impressive and astonishing.

拿現實舉例，抖音等大陸短影片平台巨頭，他們的集團組織架構與生態佈局之中，本就佈局投資及自營相關AI科技公司及音樂上下游公司。理論上，只要集團戰略支持與確保AI音樂能夠達到原創的認定標準，能夠確保規避版權紛爭的風險，他們是能夠完成從自產、到自推、到自銷的100%全流程閉環，不僅能夠砍掉一部分對外的巨額版權採買預算，還能夠通過播放與衍生商業用途實現投資與收益可持續的自循環，但我還是覺得大公司不要對行業趕盡殺絕⋯⋯

Taking a real-world example, the organizational structure and ecosystem of major Chinese short-video platforms like Douyin (TikTok) already include investments in and self-operated AI technology companies and companies in the music industry's upstream and downstream sectors. Theoretically, as long as the group's strategy supports and ensures that AI music meets originality standards and avoids copyright disputes, they can achieve a 100% closed-loop process from self-production to self-promotion to self-sales. This would not only cut a significant portion of their huge external copyright acquisition budgets but also achieve a sustainable self-circulation of investment and returns through playback and derivative commercial uses. However, I still think large companies shouldn't try to completely destroy the industry…

對於專注、堅持頂級藝術標準和頂級技術標準的音樂人而言，AI也同樣能夠作為「得力助手」的角色參與其中。只要真人不讓位監製、製作人、內容統籌等核心角色給AI，由真人監製在流程管理中把控藝術標準和技術標準達到行業頂級水平，AI作為流程環節輔助工具使用，使用得當的情況下：「真人+AI」的合作，在效果上非但不會打折扣，甚至讓原本天馬行空、難以實現的想法得以落地，而且比全流程的真人班底更具效率及性價比。

For musicians who are focused on and committed to top-tier artistic and technical standards, AI can also serve as a valuable assistant. As long as human producers don't relinquish their core roles as supervisors, producers, and content coordinators to AI, and human producers maintain top-level control over artistic and technical standards during process management, with AI used as a supplementary tool, the collaboration between "human + AI" can, when used properly, not only maintain its effectiveness but also bring previously unrealistic ideas to fruition, achieving greater efficiency and cost-effectiveness than a fully human-driven team.

雖然人人皆可操作使用生成式AI，但作品想要達到商業級也具有很高門檻。就像在影像領域，非專業的人不會有创意简报（Brief）、分鏡⋯⋯的認知與流程概念，佔據絕大部分的非專業人士都是「傻瓜式」使用，提出一個大概的感覺形容、元素形容與方向形容，再交由AI主導生成影像內容。而對於專業人士而言，會將AI作為輔助工具，拆分流程後，將專業需求交由AI潤色生成创意简报（Brief），期間由真人不斷與AI溝通修正及調整細節，再將创意简报（Brief）及更具像、更細節的要求交由AI進行分鏡創作，再由真人與AI溝通不斷修正及調整細節⋯⋯一步步往下推。同理。

While generative AI is accessible to everyone, achieving commercial-grade quality requires overcoming significant hurdles. For instance, in the professional video field, non-professionals lack the technical understanding and process of creating briefs, storyboards, etc. The vast majority of non-professionals use it in a simplistic way, providing a general description of the feeling, elements, and direction before letting the AI ​​generate the video content. Professionals, on the other hand, use AI as an auxiliary tool. They break down the process, having the AI ​​refine their professional requirements and generate a brief. During this process, human engineers continuously communicate with the AI ​​to revise and adjust details. Then, the AI ​​creates storyboards based on the brief and more detailed requirements, with human engineers again communicating with the AI ​​to refine and adjust details—a process that continues step by step.

不管是少環節使用AI，還是多環節使用AI，AI的使用參與相較傳統模式必然減少成本支出，那麼行業裡就會有部分人開始帶頭卷性價比，難以避免，然後發展成更多人不得不跟著卷性價比。規模化卷性價比的背後，就是推動行業整體預算普遍降級，全真人團隊的成本無法無底線壓縮——在性價比變低以後，在主流商業市場逐漸邊緣化，行業人員進行不斷內部迭代，完成新舊交替——這是客觀的現實趨向與運轉規律。

Whether AI is used in a few or many stages of the workflow, its use will inevitably reduce costs compared to traditional methods. Consequently, some in the industry will inevitably start prioritizing cost-effectiveness, leading to a situation where more people are forced to follow suit. This large-scale focus on cost-effectiveness will result in a general downgrading of industry budgets, while the cost of fully human teams cannot be reduced indefinitely. As cost-effectiveness decreases, these teams will gradually become marginalized in the mainstream business market, leading to continuous internal iteration and personnel turnover – this is an objective trend and operational law.

AI是新世代音樂產業的必答題——說到這個話題，私心來講，其實我還挺期待aMEI張惠妹「休眠已久」的分身「阿密特」在AI新世代再度覺醒，以新單曲的形式與AI對話，帶出有意思的視角與切入點⋯⋯當然不是形式老土的假扮成仿生機器人的設定⋯⋯而是，能夠探討科技與人文的對立矛盾及不得不並存共生的現實，不止是意識和概念⋯⋯而是在項目執行上也通過「真人+AI」的團隊合作形式，高度整合視覺、影像、部分音頻素材⋯⋯從概念到執行，都先鋒且有趣⋯⋯一是具有當下新的時代性，二是為阿密特進行迭代升級，注入新鮮不朽的血液。之所以覺得阿密特最適合，是因為阿密特自從「出生」以來的血脈基因就很適合，特立獨行、自然而然、澎湃自在。（酋長/文）

AI is an essential question for the new generation of the music industry. Speaking of this topic, personally, I'm quite looking forward to aMEI (Chang Hui-mei)'s long-dormant artistic persona, "Amit," awakening again in the new AI era, engaging in dialogue with AI through a new single, and bringing out interesting perspectives and entry points... Of course, it's not the old-fashioned setting of pretending to be a bionic robot... Rather, it's about exploring the contradictions and contradictions between technology and humanity, and the reality of their inevitable coexistence, not just consciousness and concepts... In terms of project execution, it also requires a "real person + AI" team collaboration, highly integrating visual, video, and some audio materials... From concept to execution, it's pioneering and interesting... Firstly, it has contemporary relevance, and secondly, it injects fresh and timeless blood into Amit's iterative upgrade. The reason I think Amit is the most suitable is because Amit's inherent genes since its creation are very suitable—unique, natural, and exuberant. (By 酋長)