---
author: TED
date: '2026-09-25'
guest: ''
layout: post.njk
source: https://www.youtube.com/watch?v=txf8CrI5z-k
speaker: TED
tags:
  - music-history
  - gender-gap
  - cultural-erasure
  - ai-bias
title: 消失的另一半音乐史：在算法时代抢救人类被抹去的声音
summary: 音乐历史中近90%的现存作品由男性创作，这种历史性的声音缺失正在被生成式AI系统放大。通过发掘与记录上万名被遗忘的女性作曲家作品，我们必须在算法定型未来文化之前补全人类的声音档案。
insight: ''
draft: true
series: ''
category: culture-history
area: society-thinking
project: []
people: []
companies_orgs: []
products_models: []
media_books: []
status: evergreen
---
### 历史失语的裂隙：从六千到一万八千位被遗忘的女性作曲家

音乐的传承历来伴随着权力的筛选与沉默。开场的这首旋律已有近九百年的历史，它由中世纪法国南部的女吟游诗人兼作曲家**迪亚伯爵夫人**（Comtessa de Dia）所作，是整个南法吟游传统中唯一一首由女性创作且谱曲完整幸存下来的旋律。面对这段跨越千年的声音，一个根本性的问题油然而生：究竟是谁在决定哪些声音能够穿越历史留存下来？

作为一名受过严苛学院训练的专业音乐家，我在此前的人生中对此毫无察觉。直到数年前，我在伦敦的一处**二手书市**偶然翻到一本百科全书，书中赫然记录了音乐史上六千名女性作曲家的名字。这带来了巨大的认知震荡——在那个时刻，我在脑海中能拼凑出的女性作曲家名字甚至不超过五个。随着我逐一检索她们的生平并聆听那些动人心魄的作品，我彻底被这些被湮没的创造力所震撼，并开始不知疲倦地向身边的每一个人分享。

这种热忱首先波及了我的家庭。2018年，我的丈夫——一位普通的**铁路工程师**，在日复一日的早餐时光里被迫听了一整年女性作曲家的作品，甚至戏谑地成为了“全世界最懂女性作曲家的铁路工程师”。然而，正如工程师对现实行动的执着，他向我提出了一个关键质问：“你每天在餐桌前放给我也改变不了任何现状，你到底打不打算为此做点实事？”这句话促使我建立了初步的名录并公布到互联网上。在短短数月内，世界各地的女性音乐家纷至沓来，纷纷申请将自己的名字与作品载入其中。如今，最初的网络名单已发展为一个正式的**基金会**，我们已经在音乐长河中发掘并记录了超过一万八千名女性作曲家的名字。

<details>
<summary>Original English</summary>

[singing]

This song is almost 900 years old. It was written by Comtessa de Dia, a poet and composer from medieval southern France. It is the only surviving melody written by a woman from that entire tradition.

So, let me start with a question. Who decides which voices survive?

I never really thought about that until a few years ago when I was walking on a second-hand book market in London and I found an encyclopedia that listed 6,000 women composers in the history of music. I could not believe it because at that time I think I could name five.

But as soon as I started looking into their names and listening to this incredible music, I kid you not, I would not shut up. Well, I haven't shut up. I'm here and I just sharing this music with everybody all the time. In fact, please allow me.

[music]

I mean, do you even believe that? Did hear all the classical strings, the African-American dancers mixed with the romantic music?

Well, welcome to my husband's life in 2018. Now, my husband, he's a railway engineer, which means he's now the railway engineer who knows the most about women composers in the whole world. I know it's not a very competitive field, but still.

So, one day he said to me, "Look, it's very nice that you're playing this music by these women at breakfast for like, I don't know, a year now, but you do understand that nothing is going to happen if you just tell me. I'm an engineer, so are you going to do something or what?" And I was like, "Fair enough."

So, I did. I made a list of these incredible women, I put it online, and within months, women from all over the world, they start writing to me. And they said, "Thank you so much for what you're doing. I'm a composer. Could you please add me to your list?" And I tell you this list continues to grow. It is now a foundation, and we are now uncovering over 18,000 women composers in the history of music.

</details>

### 数据偏见与静默回响：AI 时代的文化重构与声音留存

一个终身接受专业演奏与学术训练的音乐家，竟然对成千上万杰出女性创作者的存在一无所知，这绝非个人视角的偶然盲区，而是系统性结构失衡的结果。深入行业核心数据后，真相昭然若揭：在全人类跨流派聆听的所有音乐作品中，近 90% 都出自男性之手。我们不仅错失了浩如烟海的人类灵感，更为严峻的是，当今的**生成式人工智能**（Generative AI: 依据海量既有数据进行模式学习与内容生成的算法模型）正在直接吞噬并继承这种历史性的沉默。

以单一流媒体平台为例，每天有超过五万首 AI 生成的音轨被注入数字洪流之中。算法系统的输出边界完全受制于其训练集的输入质量；而眼下正在喂养这些庞大模型的，是一份缺失了半数人类创造力的残缺档案。诚然，我坚信音乐的创作权应当牢牢掌握在具有血肉和灵魂的真正音乐家手中，任何未经授权或未给予合理报酬便利用艺术家心血进行模型训练的行径在伦理上都是错误的。但现实是不可逆的，技术浪潮已然降临，我们必须直面算法重构人类记忆的规则。

为了防止历史上的沉默在机器时代被无限倍增，当务之急是主动介入底层数据的构建——全面**录制缺失的声音**。这不仅涵盖那些被历史尘封的古典先驱，也包括当下尚未获得足够代表性的现代创作者。唯有通过高保真的现代录音，这些作品才能被数字化索引，转化为可检索（Searchable）、可教学（Teachable）的结构化资产，真正进入主流广播电台、流媒体曲库与课堂教学，进而成为滋养下一代智能系统演进的正向养分。

<details>
<summary>Original English</summary>

Now, this story makes absolutely no sense. I'm a professional musician. How is it possible that I spent my life training and performing without hearing about thousands of women who wrote beautiful music?

Well, when I started doing my own research and I came across the data, then it became clear because it turns out that nearly 90% of the music we hear across genres is written by men. That's a very big gap. Can you all imagine all the amazing music we've been all missing all this time?

But, also that AI is now learning from that same silence. On one platform only, more than 50,000 AI-generated tracks are added every single day. The systems can only learn from the data available to them. But, right now that data is missing half of humanity's creativity.

Now, I believe musicians should be writing music, not machines. There are millions of talented people who can do that. Writing music fills our soul as well as humans. And to generate music from artists' work without their consent or proper remuneration is wrong.

But, the technology is here to stay. So, let's make sure all voices are heard, right? So, how do we do that? It's actually quite easy. We must record the missing voices. The historical ones, but also the current ones who are not represented. Because our recording makes this music searchable, teachable, and part of the systems shaping what the world will hear next. It brings this music to classrooms, to radio stations, to streaming platforms, so everybody in this room can play and share it.

</details>

### 泥土下的摇篮曲：塑造未来记忆的最终抉择

基于这一迫切的文化现实，我向整个科技界与商业生态发出明确呼吁：所有依托 AI 音乐模型构建产品并获取利润的科技公司与利益相关方，都应当拿出部分资本反哺人类的声音遗产，共同投资补全人类音乐目录的缺失版图。今天我们主动选择记录什么、让什么被世人听见，直接决定了未来的智能系统将替人类保留怎样的历史记忆。究竟是谁在决定哪些声音能够存续？在智能演进的拐点上，答案不是虚无的命运，而是此时此刻我们每一个人的具体行动。

在人类对抗遗忘的历史中，个体的微光始终具有穿透黑暗的力量。二战期间，犹太女诗人兼歌手**伊尔瑟·韦伯**（Ilse Weber）与家人一同被囚禁在捷克的**泰雷津集中营**（Theresienstadt Ghetto）。在绝望的炼狱中，她没有创作宏大的交响曲，而是拿起乐器为集中营里的老人和孤弱儿童编写温暖的歌曲，用旋律维系人性最后的体温。不幸的是，她最终与年幼的儿子一同在奥斯威辛集中营的毒气室中罹难。

但在走上绝路之前，她的丈夫冒死将她的乐谱手稿深埋在集中营工具棚的泥土之下。正是凭借这深埋地底的纸页，韦伯的创作奇迹般地幸免于难。其中最为人所铭记的，是一首名为《摇篮曲》（Wiegalalied / Wiegala）的作品。人工智能正在以前所未有的速度学习和吞吐人类文明的全部造物；在我们教导机器描摹未来的时刻，请务必教导它倾听并铭记属于我们全体人类的完整回响。

<details>
<summary>Original English</summary>

So, here's my invitation. What if every tech company making AI music and anyone profiting from it would invest in making our musical catalog more complete? Can you imagine what gets heard? It is chosen, and those choices will shape what the future will remember.

So, who decides which voices survive? Now, more than ever, we do.

And I want leave you all with one woman's story. Ilse Weber was a poet, a singer, who was imprisoned in the ghetto of Theresienstadt during the Second World War with her family. As she was there, she didn't write symphonies, but she wrote songs to entertain the elderly and the children. That's the power of music. And that was the gift of one woman.

She and her son sadly died in Auschwitz. But her husband buried her manuscripts under a tool shed. And because of that, a few of her songs survived. This one is a lullaby. It's called Wiegalalied.

[screaming]

Artificial intelligence is learning. Let's teach it to remember the sounds of all of us. Thank you.

[applause]

Woo! [cheering]

</details>