6:54
马斯克 v. 奥特曼——硅谷最大 AI 恩怨案
Musk v. Altman — Silicon Valley's Biggest AI Feud
This is an HSK 5-6 Chinese listening episode that runs about 7 minutes. The full Mandarin script is shown with tap-for-pinyin and a line-by-line English translation, so you can listen and read at once — comprehensible input in the sense of Stephen Krashen's i+1 theory. It teaches 12 key vocabulary words such as 政府、信任、公平 and walks through 4 grammar patterns, each explained in English with examples. The same news story is retold at 4 difficulty levels — use the level selector above to find the version that is challenging but still understandable for you.
2026 年 5 月 18 号，加州奥克兰。
May 18, 2026, Oakland, California.
一个 12 个人的陪审团，
A 12-person jury,
开了几天会，
after several days of meetings
听了一个 AI 行业最大的官司之后，
and hearing the biggest lawsuit in the AI industry,
只用了不到两个小时，
took less than two hours
得出了一个结论。
to reach a conclusion.
马斯克输了。
Musk lost.
具体地说：
Specifically:
他们认定，
they decided that
马斯克对 OpenAI 和山姆·奥特曼的起诉，
Musk's case against OpenAI and Sam Altman
超过了法律允许的时间。
exceeded the time the law allows.
也就是说：
That is to say:
不是马斯克的理由不对，
it wasn't that Musk's reasoning was wrong,
是他来得太晚了。
he came too late.
这是美国近年来最受关注的科技官司之一。
This is one of the most-watched tech lawsuits of recent years in the US.
我们今天来好好说一说。
Let's go through it properly today.
故事要从 2015 年说起。
The story begins back in 2015.
那一年，
That year,
一群人在旧金山开了一个会。
a group of people held a meeting in San Francisco.
里面有马斯克，
Among them was Musk,
有山姆·奥特曼，
Sam Altman,
还有几个搞 AI 的研究员。
and several AI researchers.
他们都在担心同一件事：
They were all worried about the same thing:
未来 AI 可能会变得非常厉害。
in the future AI could become extremely powerful.
如果只有谷歌、Facebook 这种大公司，
If only big companies like Google and Facebook,
关起门来做 AI，
made AI behind closed doors,
那 AI 可能会被几个公司控制。
then AI might end up controlled by a handful of companies.
这对全人类，
That would be dangerous,
是危险的。
for all of humanity.
所以他们决定：
So they decided:
我们自己做一个 AI 公司。
we'll start our own AI company.
但这个公司，
But this company,
不能像谷歌那样，
must not be like Google,
是为了赚钱。
chasing profit.
它必须是非营利的。
It must be a non-profit.
研究成果，
Research results
要公开。
will be public.
代码，
Code
要开放。
will be open.
这就是 OpenAI。
That was OpenAI.
马斯克在一开始，
At the start,
出了很多钱。
Musk put in a lot of money.
他答应给一亿美元。
He pledged a hundred million dollars.
他在开始的几年，
For the first few years,
一直被认为是 OpenAI 的代表人物。
he was seen as the face of OpenAI.
但是到了 2018 年，
But by 2018,
两个人就吵起来了。
the two of them began to clash.
第一个原因，
The first reason
是钱。
was money.
做 AI 比想象中贵得多。
Making AI was far more expensive than imagined.
非营利模式拿不到足够的钱。
The non-profit model couldn't raise enough.
奥特曼和其他人开始想：
Altman and others began to think:
能不能加一个营利的部分？
could we add a for-profit arm?
还是给所有人用，
Still available to everyone,
但是允许投资人有回报。
but with returns allowed for investors.
马斯克不同意。
Musk disagreed.
马斯克的逻辑是：
Musk's logic was:
我们当初就说好是非营利的。
we said at the start it would be a non-profit.
你们现在变了，
You're changing it now,
就是不守约定。
which means breaking the deal.
第二个原因，
The second reason
更尖锐。
was sharper.
马斯克想自己当 OpenAI 的 CEO。
Musk wanted to be CEO of OpenAI himself.
他想接管。
He wanted to take it over.
但是其他人不同意。
But the others didn't agree.
最后，
In the end,
2018 年 2 月，
in February 2018,
马斯克离开了 OpenAI 董事会。
Musk left OpenAI's board.
也停止了他的捐款。
He also stopped his donations.
OpenAI 那个时候，
OpenAI at that point
其实是危险的。
was actually in danger.
拿不到马斯克的钱，
Without Musk's money,
公司差一点活不下去。
the company almost didn't survive.
但是奥特曼和团队撑住了。
But Altman and the team held on.
他们建立了一个营利的子公司，
They built a for-profit subsidiary,
让微软投资了一百亿美元，
got Microsoft to invest ten billion dollars,
然后做出了 ChatGPT。
then made ChatGPT.
然后做出了 GPT-4。
Then made GPT-4.
然后估值从 200 亿，
Then the valuation went from 20 billion,
涨到了 5000 亿美元。
to 500 billion dollars.
而马斯克呢？
And Musk?
马斯克在外面，
Musk, on the outside,
看着 OpenAI 一步一步变大。
watched OpenAI grow step by step.
看着奥特曼变成 AI 界最有权力的人。
Watched Altman become the most powerful person in AI.
看着 OpenAI 越来越像一个普通的科技公司。
Watched OpenAI become more and more like an ordinary tech company.
他越想越不甘心。
The more he thought, the less he could let it go.
2024 年，
In 2024,
马斯克在加州联邦法院起诉了 OpenAI 和奥特曼。
Musk sued OpenAI and Altman in California federal court.
他的核心说法是两个：
His core claims were two:
第一，
First,
OpenAI 违反了"慈善信托"——
OpenAI violated the "charitable trust" —
也就是说，
meaning,
你们当初是用"慈善"的名义建的，
you originally built this in the name of charity,
钱、人才、关注度，
the money, the talent, the attention,
都是基于这个承诺来的。
were all based on that promise.
现在你们变成了营利公司，
Now you've become a for-profit company,
就是骗了所有当年的捐款人和支持者。
which means you tricked all the donors and supporters from back then.
第二，
Second,
奥特曼和其他高层，
Altman and other leadership
不公平地让自己富了起来。
unfairly enriched themselves.
他们拿了本来应该属于"全人类"的财富。
They took what should have belonged to "all humanity."
这些说法，
These claims
听起来很有力。
sound powerful.
但是法律的世界，
But the legal world
不只看"对错"，
doesn't only look at "right or wrong,"
也看"时间"。
it also looks at "time."
陪审团审了几天。
The jury heard the case for several days.
奥特曼上庭，
Altman took the stand
作了非常关键的证词。
and delivered crucial testimony.
他基本上说：
He basically said:
"马斯克 2018 年就走了。
"Musk left in 2018.
"他走的时候，
"When he left,
"我们差点死掉。
"we almost died.
"是我们自己把它救回来的。
"We saved it ourselves.
"现在他来说我们错了？
"Now he comes saying we're wrong?
"他自己当年都跑了。
"He's the one who ran away.
5 月 18 号，
On May 18,
陪审团 12 个人，
the 12 jurors,
不到两个小时就得出结论：
in less than two hours, reached a conclusion:
马斯克的起诉时间，
the timing of Musk's lawsuit
超过了"诉讼时效"。
exceeded the "statute of limitations."
什么是"诉讼时效"？
What is the "statute of limitations"?
就是任何一个事，
It means for anything,
你要告别人，
if you want to sue someone,
都有时间限制。
there is a time limit.
过了时间，
Once the time is up,
就告不了。
you can't sue.
陪审团认定：
The jury decided:
马斯克其实在 2018 年就知道了 OpenAI 的方向变化。
Musk actually already knew about OpenAI's change of direction in 2018.
那个时候他没告。
At that time he didn't sue.
后来他一直在外面骂，
Later he kept complaining publicly,
但是没去法院。
but didn't go to court.
等到 2024 年才正式起诉，
He waited until 2024 to formally sue,
已经过了法律规定的时间。
which was already past the legal time limit.
法院因此判他败诉。
The court therefore ruled against him.
但是没有讨论：
But did not discuss:
"OpenAI 到底有没有违反约定。
"whether OpenAI actually violated the agreement.
这就是马斯克为什么生气。
That's why Musk is angry.
他说：
He said:
"这只是一个时间问题。
"This is just a timing issue.
"这是技术性问题。
"This is a technicality.
"如果只用技术性问题来挡这种大案，
"If a mere technicality can block a case this big,
"那以后所有的慈善承诺都没意义了。
"then all charitable promises in the future are meaningless.
他宣布要上诉。
He announced he will appeal.
要打到联邦上诉法院。
All the way to the federal appellate court.
奥特曼那边呢？
And Altman?
他很平静。
He was calm.
他说：
He said:
"OpenAI 终于可以专心做我们要做的事了。
"OpenAI can finally focus on doing what we're meant to do.
这个判决，
This ruling
其实告诉我们三件事。
actually tells us three things.
第一件事，
First,
关于 AI 的本质。
about the nature of AI.
OpenAI 从一个非营利的小公司，
OpenAI going from a small non-profit
变成今天一个 5000 亿美元的巨头，
to today's 500-billion-dollar giant,
这个过程，
that process
是一个理想被现实改造的过程。
is one of an ideal being remade by reality.
当年说要"给所有人"，
Back then they said "for everyone,"
今天的 AI，
today's AI
是要月费的，
requires a monthly fee,
是被几个国家控制的，
is controlled by a few countries,
是越来越像水电一样的基础设施。
and is becoming infrastructure, like water or electricity.
理想没有死。
The ideal isn't dead.
但是它被钱重新定义了。
But it has been redefined by money.
第二件事，
Second,
关于硅谷的友谊。
about Silicon Valley friendship.
马斯克和奥特曼，
Musk and Altman,
2015 年是合伙人。
in 2015, were co-founders.
2018 年是不和的同事。
In 2018, estranged colleagues.
2024 年是对手。
In 2024, opponents.
2026 年是法院上的两个人。
In 2026, two people facing off in court.
硅谷的"朋友"，
"Friends" in Silicon Valley
本来就跟普通人的"朋友"，
have never been quite the same
不太一样。
as "friends" for ordinary people.
中间隔着几十亿、几千亿美元，
Between them sit billions and hundreds of billions of dollars,
隔着对未来的不同想法。
and very different visions of the future.
第三件事，
Third,
关于法律。
about the law.
就算你是世界最有钱的人，
Even if you are the richest person in the world,
法律也不会因为你而停下来。
the law will not stop for you.
"诉讼时效"听起来是个小事，
The "statute of limitations" sounds like a small thing,
但它就是这次马斯克输的原因。
but it's exactly why Musk lost this time.
马斯克会上诉。
Musk will appeal.
故事还没有完。
The story isn't over.
但是这一回合，
But this round,
奥特曼赢了。
Altman won.
我们看着这个 AI 行业，
We watch this AI industry
就像看着一部慢慢展开的电视剧。
the way you watch a slowly unfolding television drama.
新的人物会出来，
New characters will appear,
新的官司会出来。
new lawsuits will appear.
但是 ChatGPT 还在那里。
But ChatGPT is still there.
xAI 也还在那里。
xAI is still there.
AI 的故事，
The story of AI
还很长。
still has a long way to go.
好。
Okay.
今天就说到这里。
That's all for today.
我们下个故事见。
See you in the next story.
Government. Used in the closing argument about AI being 'controlled by a few 政府'. Important in any modern AI discussion.
Trust. The unwritten word beneath the lawsuit: donors gave because they 信任 the non-profit framing. Musk's claim is that that 信任 was betrayed.
Fair, fairness. Musk's signature complaint: '这不公平' — and the deeper question of whether the 'statute of limitations' is itself a fair tool here.
Decision. Several pivot points in the story are 决定: to leave, to restructure, to sue, to dismiss. Each one bent the arc.
To accept. Musk refused to 接受 OpenAI's drift, then refused to 接受 the verdict — the refusal is consistent.
Serious, severe. Used implicitly: this is one of the most 严重 tech lawsuits of recent years, given the size of the parties and the stakes.
Non-profit. The founding label of OpenAI in 2015 and the entire basis of Musk's claim. The fight is about how much of that founding identity still binds the company today.
Legal term: the time window within which a lawsuit must be filed. Once it lapses, the case is barred regardless of merit. This is exactly what doomed Musk's claim — not its substance, its timing.
To appeal (a verdict to a higher court). Musk's announced next move: 上诉 to the federal appellate court. The case isn't over until the appeals are exhausted.
Jury — 12 ordinary citizens who decide questions of fact in a US trial. Their less-than-two-hour deliberation here signals confidence in the timing-bar reading.
To violate. Musk's claim: OpenAI 违反了 the founding agreement / 'charitable trust'. The court never ruled on whether this was true — only that he sued too late.
(Company) valuation. The episode tracks OpenAI's 估值 climbing from $20B to $500B as the engine of Musk's grievance: he watched the wealth he helped seed compound into someone else's empire.
* beyond level超纲词
不是 A，而是 B
Sharp contrastive structure. The verdict logic: 不是马斯克的理由不对，是他来得太晚了 — it isn't that Musk was wrong on the merits; it's that he was late. 而是 makes the replacement explicit; 是 alone is the conversational variant used here.
不是马斯克的理由不对，是他来得太晚了。
不是为了赚钱，是为了让 AI 安全。
就算 A，也 B
Concessive structure. The closing principle: 就算你是世界上最有钱的人，法律也不会因为你而停下来 — even if you're the richest person in the world, the law doesn't stop for you. Marks a hard rule that refuses to bend to status.
就算你是世界上最有钱的人，法律也不会因为你而停下来。
就算他上诉，也要等很久。
越 V₁ 越 V₂
Proportional structure. 'Musk 越想越不甘心' — the more he thought, the less he could let it go. One of the most compact ways to render an escalating emotional state.
他越想越不甘心。
越想越生气。
从 A 到 B
Tracks change across a span. The episode stacks them: 2015 是合伙人，2018 是不和的同事，2024 是对手，2026 是法院上的两个人 — a whole arc compressed into time stamps. 从…到 is the formal generalized form.
从一个小的非营利公司，变成了一个非常大的公司。
从 2015 年到 2026 年，他们从朋友变成了对手。