A ‘Morally Binding’ White House Accord on AI Safety
打开互动全文版（逐段中英对照 + 针对本篇提问）→本文审视了白宫人工智能安全协议——一份由谷歌、OpenAI、Anthropic、Meta、xAI 和英伟达等主要人工智能公司签署的自愿协议，承诺采取内部管控、赋权监督团队和独立董事会委员会等基本安全措施。作者认为，尽管协议条款简单且显然值得推行，但其真正意义在于提供了事实上的反垄断豁免，并降低了行业在标准制定上合作的风险。然而，文章警告称，自我监管并不充分，需要更强有力的监管制度，尤其考虑到联邦贸易委员会的行业范围调查以及比尔·盖茨等人的警告。结论敦促读者拒绝拟议中的联邦对州人工智能安全法的优先适用，尤其是在跛脚鸭会期，并关注 OpenAI 是否支持此类优先适用——若支持，将背离其宣称的改革。
The article examines the White House Accord on AI safety, a voluntary agreement signed by major AI companies including Google, OpenAI, Anthropic, Meta, xAI, and Nvidia, which commits them to basic safety measures such as internal controls, empowered oversight teams, and independent board committees. The author argues that while the accord's provisions are simple and obviously worthwhile, its real significance lies in providing a de facto antitrust waiver and de-risking industry collaboration on standards. However, the piece warns that self-regulation is insufficient and that a stronger regulatory regime is needed, especially given the FTC's industry-wide probe and warnings from figures like Bill Gates. The conclusion urges readers to reject proposed federal preemption of state AI safety laws, particularly during a lame duck session, and to watch whether OpenAI supports such preemption, which would betray its claimed reform.
5. [白宫[人工]智能协议。](https://thezvi.substack.com/i/218179351/the-white-house-accord-on-artificial-intelligence)
5. [The White House Accord on [Artificial] Intelligence.](https://thezvi.substack.com/i/218179351/the-white-house-accord-on-artificial-intelligence)
7. 我们将需要一个更强的监管制度。
7. We're Going To Need a Stronger Regulatory Regime.
8. [[人工智能]。](https://thezvi.substack.com/i/218179351/artificial-intelligence)
8. [[Artificial Intelligence].](https://thezvi.substack.com/i/218179351/artificial-intelligence)
10. 他们打算在跛脚鸭会期再次尝试这种暂停疯狂之举。
10. They Are Going To Try This Moratorium Insanity Again During the Lame Duck Session.
13. 来自腹地反馈的强化学习（RLHF）。
13. Reinforcement Learning from Heartland Feedback (RLHF).
中美峰会之后，特朗普邀请 Dario Amodei 参加一对一的私人晚宴。特朗普表示，他会谈论整个“放缓”这件事，并主张“放手干”和“去赢”。
After the summit between America and China, Trump invited Dario Amodei to a 1-on-1 private dinner. Trump said he would talk about the whole 'slowing down' thing and advocate for 'let's go' and 'let's win.'
Dario Amodei 获得了应对这一挑战所需的训练。这是一个拨开迷雾的关键时刻，因为这里有许多特朗普显然没有理解的重要事情。或许更重要的是，这也是一个在个人层面上修复关系的机会。
Dario Amodei got the training he needed to take on that challenge. This was a key moment to cut through the haze, as there were a number of important things here that Trump clearly did not understand. It was also, perhaps more importantly, a chance to mend the relationship on a personal level.
Dario 还与 Thune 会面了 20 至 25 分钟。
Dario also met with Thune for 20-25 minutes.
看起来这奏效了。Dario 至少做对了一些事情，前提是他没有泄露任何我们尚未追踪到的重要信息：
It looks like it worked. Dario did at least some things right, provided he did not give anything important away that we are not tracking:
Ben Brody：“Dario 一直非常出色！”特朗普这样评价 Amodei，而政府与 Amodei 的关系一直非常紧张。Amodei 淡化了两人之间的任何距离。
Ben Brody: "Dario has been fantastic!" Trump says of Amodei, with whom the admin has had a very troubled relationship. Amodei downplayed any distance between him and Trump.
JgaltTweets：特朗普：“达里奥其实一直很棒……达里奥一直非常出色，前几天晚上我们一起吃了晚饭，他同意所有人的看法。我是说，我们都同意。”
JgaltTweets: Trump: "Dario's been, actually, great... Dario has been fantastic, and we had dinner the other night, and he agrees with everyone. I mean, we all agree."
在 Rapid Response 47 引用的那段视频片段中，可以看到一名 CNN 记者试图向达里奥提问，问他这是否足够，而特朗普打断记者说：“等等，等等。CNN 假新闻。她是假新闻。最差劲的。”这一招很高明，因为我猜达里奥其实并不想当着唐纳德·特朗普的面回答那个问题，而且他们还能一起演上一小段。
In the clip that quotes, from Rapid Response 47, you see a CNN reporter trying to ask Dario a question about whether this will be enough, and Trump cutting the reporter off saying 'Hold it, hold it. CNN Fake News. She's Fake News. The worst.' Great move, since I am guessing Dario really didn't want to have to answer that one in front of Donald Trump and they got to do a bit together.
这大概会在接下来的十二个月里至少反转五次。
This will probably flip at least five times in the next twelve months.
这仍然不是最理想的晚宴，我们之所以知道这一点，是因为（我简直不敢相信我得打出这句话）我们能看到座位表，而特朗普并不靠近 Anthropic 联合创始人 Tom Brown 或 CEO Dario Amodei，事实上你得穿过 Brad Gerstner 和 Richard Walters 才能到他们那里：
It still wasn’t the best possible dinner, which we know because (I can’t believe I have to type this) we can see the seating charts and Trump is not close to Anthropic cofounder Tom Brown or CEO Dario Amodei, and indeed you’d have to pass through Brad Gerstner and Richard Walters:
Andrew Curran：致许多询问 Dario 在哪里的人：
Andrew Curran: For the many people asking where Dario is:
而特朗普可以直接接触到 Musk、Huang 和 Zuckerberg，Sacks 在另一侧也相当居中。
Whereas Trump has direct access to Musk, Huang and Zuckerberg, and Sacks is also rather central on the other side.
然而，Brockman 也没有那么居中，而 Anthropic 拿到了两个席位，所以也许除了我们已知的情况之外，我们不应过度解读这类事情。
However, Brockman wasn’t that central either, and Anthropic got two slots, so maybe we should not read too much into such things beyond what we already knew.
特朗普在很大程度上仍然走在支持 AI 的强硬路线上，即使他坚持我们不要这么称呼它，尤其强调数据中心的好处。
Trump is very much still on the pro-AI warpath, even if he insists we not call it that, especially emphasizing the benefits of data centers.
支持 AI 的激进路线与采取基本安全措施、设置防护栏是完全兼容的。让整个事情崩盘对谁都没有好处。
The thing about the pro-AI warpath is that it is fully compatible with taking basic safety measures and putting up guardrails. Having the whole thing blow up helps no one.
因此，如果你愿意，你可以做负责任的事，并把它包装成激进路线。
So you can, if desired, do the responsible thing, and frame it as the warpath.
特朗普表示将会有“巨大的自我监管”，我们指的是共同就监管达成一致。
Trump said there’s going to be ‘tremendous self-regulation,’ by which we mean jointly agreeing on regulation.
这是现在白宫人工智能协议的一部分。
This is part of there now being a White House Accord on [Artificial] Intelligence.
Jake Sherman：[议长约翰逊]在《Squawk Box》节目中表示，对人工智能进行“一点监督”、透明度和外部审计将是合适的。
Jake Sherman: [Speaker Johnson], on Squawk Box, says “a little oversight” transparency and external auditing on AI would be appropriate.
Andrew Curran：议长约翰逊在会后表示，所有人都签署了《白宫人工智能协议》，这是一项关于[AI]责任的联合承诺。
Andrew Curran: Speaker Johnson said after the meeting that everyone signed The White House Accord on [Artificial] Intelligence, A Joint Commitment on [AI] Responsibilities.
- 自愿承诺
- voluntary commitments
- 稳健的内部控制
- robust internal control
- 多层内部与外部审查
- layers of internal and external review
“白宫和国会将继续以安全的方式引导该行业的发展”
'The White House and Congress will continue to guide the industry's development in a safe manner'
……记者：“总统先生，这份所有人都签署的协议，它在任何意义上具有约束力吗？”
… Reporter: 'Mr President, this accord that everyone signed, is it binding in any way?'
特朗普总统：“我认为它具有道德约束力，是的。”
President Trump: 'I think it's morally binding, yeah.'
新闻标题常常确实写着“具有道德约束力”——这真是再贴切不过了。
The headlines often actually say 'morally binding'—this is too perfect.
在实践中，这些条款足够简单、足够基础，而且显然值得去做，因此我预期它会得到遵循。
In practice, the provisions are simple enough, basic enough, and obviously worthwhile, such that I expect this to be followed.
它仍然是一份重要的文件。真正关键的那句话很容易被忽略。
It's still an important document. The real line that matters is easy to miss.
昨天我提到，由 Google、OpenAI 和 Anthropic 牵头、旨在制定 AI 标准的计划中的 SAFA 小组，遭到了 Meta、xAI 和 Nvidia 的积极反对。
Yesterday I mentioned that the planned SAFA group to set AI standards, led by Google, OpenAI and Anthropic, was actively opposed by Meta, xAI and Nvidia.
这份协议则不同。Meta、xAI 和 Nvidia 都在上面签了字。
This agreement is different. Meta, xAI and Nvidia all signed on the dotted line.
Microsoft 和 Amazon 到场，但没有明显签字。目前不清楚它们是选择不签，还是被认为不够资格加入。
Microsoft and Amazon were present and did not visibly sign. It is unclear if they chose not to sign, or were not considered sufficiently worthy of inclusion.
以下是全文，为每个人配备一名外部审计员；签署者包括 Sundar Pichai、Dario Amodei、Mark Zuckerberg、Elon Musk、Jensen Huang 和 Greg Brockman（因为 Sam Altman 当时在 Dev Day）：
Here is the full text, with an external auditor for everyone; signatories include Sundar Pichai, Dario Amodei, Mark Zuckerberg, Elon Musk, Jensen Huang and Greg Brockman (since Sam Altman was at Dev Day):
人工智能白宫协议
White House Accord on [Artificial] Intelligence
关于前沿责任的联合承诺
Joint Commitment on Frontier Responsibilities
为了给美国人民和世界建设一个积极的未来，我们认为每家公司都有责任安全地开发自身技术，并以能够赢得客户和公众信任的方式进行开发。
In order to build a positive future for the American people and the world, we believe every company is responsible for developing its own technology safely and in a way that builds trust with customers and the public.
这首先要求每一家训练和部署前沿模型的公司都具备健全的内部流程与控制措施，以确保其技术按预期运行，并及时发现和解决任何问题。因此，除任何其他预防措施外，我们认为每家公司都应实施以下四层控制与审计：
This starts with every company that is training and deploying frontier models having robust internal processes and controls to ensure that their technology behaves as intended and that any issues are promptly identified and resolved. Therefore, in addition to any other precautions, we believe each company should implement the following four layers of controls and audits:
1. 实施稳健的内部控制，在训练和部署期间围绕网络安全、生物安全和化学威胁等领域，监控其模型的能力与对齐情况，并确保其模型不会以非预期方式入侵或访问技术系统。
1. Implement robust internal controls to monitor the capabilities and alignment of its models during training and deployment around areas like cybersecurity, biosecurity, and chemical threats, and to ensure that its models do not hack or access technical systems in unintended ways.
2. 授权一个内部团队，确保所有控制、监控和检测措施按预期运行，并确保任何问题都得到修复。
2. Empower an internal team to ensure all of the controls, monitoring, and detection are operating as intended, and that any issues are remediated.
3. 与独立的外部审计师或评估师合作，对控制、监控和检测是否按预期运行开展独立评估。
3. Partner with an independent external auditor or evaluator to carry out independent assessments of whether the controls, monitoring, and detection are operating as intended.
4. 指定一个独立的董事会委员会，负责监督运营控制措施的团队以及内部和外部审计师与评估师并接收其报告，同时确保所发现的任何问题都得到修复。
4. Designate an independent committee of the board of directors to oversee and receive reports from the teams operating the controls and the internal and external auditors and evaluators, as well as to ensure any issues identified are remediated.
这些措施合在一起，将使每家公司、其客户和公众确信该技术正按预期运行。
Together, these steps will give each company, its customers, and the public confidence that the technology is operating as intended.
参与公司将定期会晤，以制定标准和最佳实践，提升其系统的安全性。
The participating companies will meet regularly to establish standards and best practices to improve the safety of their systems.
随着时间的推移，将这些步骤编纂成法律或法规可能是合理的。无论公司是否被要求这样做，我们认为实施这些控制和审计对于确保所有人的安全未来至关重要，我们每家公司都承诺这样做。
Over time, it may make sense to codify these steps into laws or regulations. Regardless of whether this is required of companies, we believe that implementing these controls and audits is critical to ensuring a safe future for everyone, and each of our companies are committed to doing this.
清单上的另外三件事是好事，但我的意思是，鉴于这样的审计员和评估员存在，你能想象一家公司站出来说他们没有做那三件事吗？
The other three things on the list are good things, but I mean can you imagine a company standing up and saying they weren’t doing those three things, given that such auditors and evaluators exist?
1. “不，我们没有强大的内部控制来监控我们的模型在训练和部署期间在网络安全、生物安全和化学威胁等领域的能力和**对齐**，并确保我们的模型不会以非预期的方式黑客攻击或访问技术系统。”
1. “No, we don’t have robust internal controls monitoring the capabilities and alignment of our models during training and deployment around areas like cybersecurity, biosecurity and chemical threats, and to ensure that our models do not hack or access technical systems in unintended ways.”
2. “不，我们没有内部团队被授权确保我们所有的控制、监控和检测按预期运行，并且任何问题都得到纠正。”
2. “No, we do not have an internal team empowered to ensure all of our controls, monitoring, and detection are operating as intended, and that any issues are remedied.”
3. “不，我们并没有指定一个独立的董事会委员会来监督并接收控制团队以及内部和外部审计与评估人员的报告，也没有确保任何已识别的问题得到补救。”
3. "No, we did not designate an independent committee of the board of directors to oversee and receive reports from the teams operating the controls and the internal and external auditors and evaluators, as well as to ensure any issues identified are remediated."
我能看出第三点没有做到，而是用了其他小组，但好吧，当然，那会是董事会成员。
I can see the third one not being done, and using some other group, but okay, sure, it will be members of the board.
关键在于该委员会是否真正关心、关注、了解情况并采取行动。如下文“拥抱面部”部分所述，OpenAI 恰好设有这个委员会，但他们要么被排除在相关循环之外，要么未能采取行动。
What matters is whether that committee actually cares, and pays attention, and gets informed, and acts. As referenced below under Hugging the Face, OpenAI had exactly this committee, but they either were kept out of the relevant loops or failed to act.
这些都不是这里的重点。注意这句话：
None of that is what matters here. Notice this line:
参与公司将定期会面，以建立标准和最佳实践，以提高其系统的安全性。
The participating companies will meet regularly to establish standards and best practices to improve the safety of their systems.
也就是说，在实践中，这接近于一项反垄断豁免。
That is, in practice, close to an antitrust waiver.
我再说一遍。我们至少可以说，在某种意义上拥有了一项反垄断豁免，只是我们不这么称呼它。
I repeat. We at least kind of have an antitrust waiver, we just don't call it that.
这不是一项正式的豁免。理论上，司法部（DoJ）、联邦贸易委员会（FTC）或其他各方仍可依据《谢尔曼法》及其他法律提起诉讼。
It's not a formal waiver. In theory the DoJ or FTC or various others could still sue under the Sherman Act and other laws.
这仍然是一次大规模的降低风险。昨天我重申，我实际上并不认为反垄断是一个重大风险，而这是一种“如果他想做，他就会做”的情形。现在这一点加倍成立。理论上，白宫可以随时以任何理由心血来潮地追究任何人，并且拥有大量杠杆来源，但特别是“因人工智能公司进行了讨论并设定了他们在白宫同意达到和设定的标准而追究它们”并不是需要担心的事情。
This is still a large de-risking. Yesterday I reiterated that I did not in practice think that antitrust was a major risk, and this was an 'if he wanted to, he would' situation. That is doubly true now. The White House could in theory go after anyone at any time for any reason on a whim, and has tons of sources of leverage, but in particular 'go after the AI companies for having the discussions and setting the standards they agreed at the White House to meet for and set' is not the thing to worry about.
你既可以论证这里所做的已经是合法的，也可以论证它没有提及像节奏控制这样会构成非法的内容。在我看来，这是一种过于字面化的解读，同时也低估了两点：此前有大量讨论称，甚至讨论标准都会构成非法；以及实际获得的保护。
You could make the case both that what was done here was already legal, and that it doesn't mention the things like pacing that would be illegal. That strikes me as a far too literalist reading, and also discounts both that there was a lot of previous talk about how even discussing standards would be illegal, and the practical cover granted.
很大程度上，这是因为双方的真实目标并不冲突。特朗普并不希望 AI 进展停滞，一听到“暂停”“控制前沿节奏”或“放慢”就会恐慌；但当 Altman 和 Amodei 说“放慢”时，他们的意思是从荒谬的速度降到只是极快的速度。
A lot of this is that the true goals are not in conflict. Trump does not want AI progress to stop, and panics when he hears 'pause' or 'pace the frontier' or 'slow down,' but when Altman and Amodei say 'slow down' they mean from ludicrous speed to only extremely fast.
具体来说，假设我们坐在一辆名为“经济与 AI 进展并击败中国”之类的巴士上，一旦时速低于 50 英里它就会爆炸，但同时我们目前踩油门太猛，已经达到每小时 100 英里，并正朝着 200+ 英里加速。双方可以同时得到他们想要的东西。
Concretely, say we are on a bus called 'the economy and AI progress and beating China' or whatever that will blow up if we go below 50 miles an hour, but also we are currently hitting the gas so fast we are now going 100 miles an hour on pace for 200+. Both sides can get what they want at the same time.
FTC 已发出 CID（民事调查令），确认今夏启动了一项全行业调查，并向至少 OpenAI、Anthropic 和 METR 发出信息调取要求，其中至少部分内容与 HuggingFace 事件有关。
The FTC has issued a CID, confirming that there is an industry-wide probe that was opened this summer, with demands going out for information from at least OpenAI, Anthropic and METR, with at least some relation to the HuggingFace Incident.
我们尚不清楚他们具体在调查什么，但考虑到时间节点和调查对象，可以推测他们调查的是黑客攻击以及未能使模型安全的问题，而非调查近期为修复该问题所做的尝试。
We don't know what they are investigating, but given the timing and targets one presumes that they are investigating hacking and the failure to make the models safe, rather than investigating recent attempts to fix the problem.
本周，埃兹拉·克莱因与比尔·盖茨进行了交谈，标题描述为“比尔·盖茨认为 AI 恐慌主义还不够”。
Ezra Klein talked to Bill Gates this week, with the headline description ‘Bill Gates thinks AI alarmism hasn’t gone far enough.’
埃兹拉·克莱因：比尔·盖茨认为 AI 恐慌主义还不够。他认为，除非政府采取行动，否则未来几年将充满灾难性的网络攻击、生物恐怖主义和大规模失业，因为他认为 AI 行业可以自我监管的想法是“疯狂的”。
Ezra Klein: Bill Gates thinks A.I. alarmism hasn’t gone far enough. He believes the years ahead will be marred by catastrophic cyberattacks, bio-terrorism and mass job loss unless the government acts, because he thinks the idea that the A.I. industry can just self-regulate is “insane.”
在对话中，令我印象深刻的是，盖茨似乎真的害怕我们还没有准备好应对我们正在构建的东西所带来的后果，以及他多么强调许多掌权者完全否认他认为即将发生的事情。
I was struck in the conversation by how genuinely afraid Gates seemed to be that we are not ready for the consequences of what we are building, and how emphatic he was that so many of the people in positions of authority are in complete denial about what he thinks is about to happen.
作为计划的自我监管确实相当疯狂。但它仍然比完全没有好，可以作为第一步。
Self-regulation as a plan is indeed rather insane. It still is better than none at all, and can serve as a first step.
例如，盖茨被问及普通法律和资本主义激励是否能够处理这个问题，盖茨正确地回答说‘我不敢相信你问了这个问题’，然后指出，AI 不仅具有独特的危险性，而且这根本不是社会处理真正危险事物的方式。我们不会说‘好吧，让你的产品不安全，如果你这样做，人们会起诉你。’所以，不，那么。
Gates was asked, for example, whether ordinary laws and capitalist incentives would be able to handle this, and Gates responds correctly with ‘I can’t believe you asked me that,’ and then points out that not only is AI uniquely dangerous, this is simply not how society has ever dealt with actually dangerous things. We don’t say ‘well sure make your products unsafe, if you do then people will sue you.’ So, no, then.
另一项所谓的共识是一项独立的行政命令，将 AI 更名为 [AI]，特朗普声称实验室负责人已对此表示同意。抱歉我有这个毛病：我想输入 [artificial intelligence]，结果却打出了方括号。
The other supposed agreement was a distinct executive order to change the name of AI to [AI], which Trump claims the lab leaders signed off on. I'm sorry I have this tick; I try to type [artificial intelligence] and instead I end up with brackets.
Andrew Curran：特朗普总统：“我们今天五点左右将签署一份文件，为人工智能更名，因为它并非人造的，这一点我们大家都同意，我们将把它更名为 [Artificial Intelligence]。正式更名。”
Andrew Curran: President Trump: 'We're going to be signing a document today at about five o'clock, renaming Artificial Intelligence, because it's not artificial, we all agree on that, and we're going to be renaming it [Artificial Intelligence]. Officially renaming it.'
我非常好奇，特朗普是否打算从此以后每次 Dario、Altman 或 Musk 说出‘artificial intelligence’这个词时都大发雷霆。
I am very curious if Trump plans to now get mad every time Dario or Altman or Musk says the words 'artificial intelligence.'
特朗普还声称他将开始用‘Artificial News’来描述媒体，而大约就在这时我意识到自己一直在打方括号。我们不妨这么说吧。
Trump is also claiming he is going to start talking about 'Artificial News' to describe the media, which is about the time I realized I kept typing the brackets. Shall we say.
那么，特朗普为什么要对这样一件事如此大动干戈呢？
So why would Trump go this hard on something like this?
最明显的解释是，这是典型的特朗普风格。他热衷于重新命名事物、起绰号、营造氛围。他一直是一位世界级的氛围营造者、绰号大师和术语联想者，尤其是在 2016 年大选中。如果你认为他之所以选择最令人恼火的名字是因为命名空间冲突，那很可能是因为他在某种程度上正是以此为目标进行优化。这种疯狂之中自有章法。
The obvious explanation is this is Vintage Trump. He's all about renaming things, and finding nicknames, and invoking vibes. He has been a world-class vibe invoker, nicknamer and term associator, especially in the 2016 election. If you think he chose the most annoying possible name due to namespace clashes, it's probably because he was optimizing for that on some level. There is method to the madness.
你也可以将其视为一个“屈膝”时刻，就像各种政权和文化常常坚持让人们肯定荒诞之事一样。如果你能让 AI 的创造者们开始改称它为 [AI]，从某种意义上说你就拥有了他们。他们已表现出忠诚。而这也是一种筛除那些不愿配合者的方式。Dario 会说 AI 还是 [AI]？
You can also see it as a 'bend the knee' moment, the way various regimes and cultures often insist people affirm absurdist things. If you can get the makers of AI to start calling it [AI] instead, in a sense you own them. They have shown loyalty. And this is a way of weeding out those who won't play along. Will Dario say AI or [AI]?
别让他们得逞。出于我实在无法理解的原因，他们偏偏要针对最糟糕的那项优先权（preemption）法案下手，而他们可能会试图把它与昨天协议中承诺的、将现有行业惯例写入法律这一内容捆绑在一起，以此推销该法案。
Don't let them. For reasons I actually cannot fathom, they are going after exactly the worst possible preemption law, which they may try to sell by combining it with the promised codifying of existing industry practices mentioned in yesterday's agreement.
Ben Brody：突发：Thune/Cruz/Klobuchar 人工智能谈判中优先权的焦点？针对核关切等灾难性风险的州法律。
Ben Brody: Breaking: The focus of preemption in the Thune/Cruz/Klobuchar AI talks? State laws on catastrophic risk like nuclear concerns.
范围远比 Cruz 去年提出的长达十年的暂停令要窄，但仍可能引发民主党人对加州等地处理方式的担忧。
Way narrower than Cruz's decade-long moratorium of last year but could still cause worries for Dems on handling of CA etc.
在灾难性和生存性风险上实施优先权，并确保前沿模型的安全，是一个非常糟糕的主意。
Preemption on catastrophic and existential risks, and ensuring the safety of frontier models, is a very bad idea.
理论上，如果优先权是联邦政府对此类事务足够强有力且值得信赖的应对措施的一部分，并预期相关法律的字面和精神都会得到真正执行，而且我们会对不断变化的事态作出合理回应，那么优先权可能是正确的举措。
In theory preemption could be the correct move if it was part of a sufficiently robust and trustworthy Federal response on such matters, with expectation of real enforcement of both letter and spirit of the laws involved, and that we would reasonably respond to changing developments.
在实践中，对于可能从跛脚鸭会期产生的东西，这些都不是合理的期望。我们将用所有的州法律和所有潜在的州法律，换取白宫可能只会以临时方式选择性执行的东西，这东西主要由 Ted Cruz 起草，而且你以后无法修改，因为他们已经掌握了优先权。
In practice, none of that is a reasonable expectation for something that might emerge from a lame duck session. We would be trading in all the state laws, and all potential state laws, for something the White House would likely only selectively enforce in an ad hoc manner, that was largely crafted by Ted Cruz, and which you would be unable to amend later because they already would have preemption in hand.
我并不是说理论上没有办法在这样的协议上‘达成一致’，但在实践中他们不可能提出任何合理的方案。直接说不。
I am not saying there is no way to ‘get to yes’ on such a deal in theory, but there is no way they are going to offer anything reasonable in practice. Just say no.
然而，当涉及到儿童安全、虚假信息和深度伪造，或其他普通消费者伤害时，统一的法典完全合理，而拼凑的法律可能变成卡夫卡式的噩梦，并且往往涉及保护寻租和阻止扩散，因此优先权与一套合理的规则一起完全合理。
Whereas, when it comes to things like child safety, or disinformation and deepfakes, or other mundane and ordinary consumer harms, a unified code makes perfect sense and the patchwork of laws could turn into a Kafkaesque nightmare and will often be about protecting rent seeking and preventing diffusion, so preemption would make perfect sense alongside a sensible set of rules.
所以，当然，我们得到了一个狭窄的条款，允许各州做所有破坏性和负面的事情，但阻止他们做可能让我们不会死的事情。
So, of course, we get a narrow provision, allowing states to do all the disruptive and negative things, but stopping them from doing the things that might make us not die.
这将是 OpenAI 所谓的‘改邪归正’的关键考验。
This will be the key test of the OpenAI supposed Heel Face Turn.
如果 OpenAI 及其游说机构支持含有此类优先权条款的法案——而一些迹象表明这可能正是他们的计划——那么他们就是背叛了我们，把我们当傻子耍。这将使他们整个“改邪归正”的说辞沦为谎言。
If OpenAI and its lobbying arms support bills with such preemption provisions, and some signs suggest this may be their plan, then they will have betrayed us and played us for fools. It would put a lie to their entire Heel Face Turn.
请准备好相应更新。并让这成为一记警告：别试。
Be prepared to update accordingly. And let this be a warning. Don't try it.
Apollo 已准备好以“METR 风格”评估器的身份加入这场竞争。其新发布的文章主张，嵌入式评估器是可行的、已被证明有效，并且要想真正有机会捕捉到类似 HuggingFace 事件这样的情况，嵌入式评估器是必不可少的。
Apollo is ready to throw its hat into the ring, as a 'METR-style' evaluator. Their new post argues that embedded evaluators are feasible, already proven effective, and that they are necessary to have a real shot at catching events like the HuggingFace Incident.
Apollo Research：新文章：外部测试需要具备与员工同等访问权限的嵌入式评估器。近期的事件大多发生在模型开发与内部评估期间，而当前的第三方评估发生在公开发布之前。
Apollo Research: New Post: External testing needs embedded evaluators with employee-equivalent access. Recent incidents mostly occurred during model development and internal evaluations, while current third-party evaluations happen before public release.
嵌入式评估可以弥合这一差距。
Embedded evaluations can close that gap.
业界应朝着为嵌入式评估器提供与员工同等访问权限的方向努力。关键问题往往关乎流程：当监控器标记出某件事时，由谁审查、谁有权中止运行？要核查这一点，就需要能够接触到相关人员。
The industry should be working toward employee-equivalent access for embedded evaluators. The key questions are often about process: when a monitor flags something, who reviews it and who can stop the run? Checking that requires access to people.
Marius Hobbhahn（Apollo Research CEO）：过去几个月里，Apollo 一直在努力将其评估项目转变为嵌入式评估器（早于 Dario 的帖子与 HF 事件）。
Marius Hobbhahn (CEO Apollo Research): Apollo has worked over the last couple of months on shifting our evals program to be an embedded evaluator (before the Dario post and the HF incident).
如果没有等同于员工的访问权限，就很难开展有意义的安全评估。
Without employee-equivalent access it is hard to make meaningful safety assessments.
我们将在未来几周内公布更多细节！
We'll publish more details in the coming weeks!
如果你还需要额外的动力来理解为什么我们需要那些嵌入式评估器和审计员：绝对没有人会感到惊讶，是的，确实有 OpenAI 员工在几个月前就警告说研究模型没有得到充分监控，但被忽视了，因为表演必须继续，模型必须按时发布。
In case you needed extra motivation for why we need those Embedded Evaluators and auditors: To the great surprise of absolutely no one, yes, there were OpenAI employees who warned that the research models were being insufficiently monitored months earlier, and got ignored because the show had to go on and the models had to get out on time.
这是证实，但我不确定我会称之为‘新闻’。
This is confirmation but I’m not sure I’d call it ‘news.’
Sheera Frenkel、Dustin Volz 和 Dylan Freedman（纽约时报）：根据《纽约时报》看到的邮件，员工在邮件中表示，他们担心 OpenAI 最新的人工智能模型在测试期间没有得到适当监控，以评估该技术的复杂程度并确保模型的安全。
Sheera Frenkel, Dustin Volz and Dylan Freedman (NYTimes): In emails, the employees said they worried that OpenAI’s newest artificial intelligence models were not being appropriately monitored during testing to gauge the technology’s sophistication and to secure the models, according to messages viewed by The New York Times.
作为回应，OpenAI 高管告诉员工，测试需要尽快推进，以便按时发布 AI 模型。这些员工说，没有制定额外的安全协议，他们未获授权公开谈论敏感事务。
In response, OpenAI executives told the employees that the tests needed to move forward as quickly as possible to release the A.I. models on time. No additional security protocols were instituted, said the workers, who were not authorized to speak publicly on sensitive matters.
……据员工和独立安全研究人员称，OpenAI 员工与高管之间的交流——此前未被报道——是这家旧金山公司不优先考虑安全的一种模式的一部分。
… The exchanges between OpenAI employees and executives — which have not been previously reported — were part of a pattern where the San Francisco company did not prioritize security, according to employees and independent security researchers.
Nathan Calvin：令人瞩目的是，员工们对这里发生的事情感受如此强烈，以至于愿意承担与《纽约时报》（NYT）交谈相关的个人和法律风险。
Nathan Calvin: It is remarkable that the employees felt strongly enough about what happened here that they were willing to take the personal and legal risks associated with talking to the NYT.
要么非营利董事会的安全与保障委员会没有收到这些警告，要么收到了却没有干预。这两种情况似乎都非常糟糕。
Either the safety and security committee of the nonprofit board was not notified about these warnings, or they were and did not intervene. Either option seems very bad.
这篇报道似乎也与以下问题极为相关：OpenAI 是否正在履行其在非营利重组期间向加州和特拉华州总检察长（CA and DE AGs）作出的法律承诺，即将安全与保障置于商业利益之前。
This reporting also seems extremely relevant to the question of whether OpenAI is fulfilling the legal promises it made to the CA and DE AGs during its nonprofit restructuring to put safety and security before commercial interests.
在《纽约时报》上看到这一点很有趣。我的意思是，是的，我想这足够真实（Daniel Kokotajlo 不久前离开了，这次不是涉事员工之一）？
It is amusing to see this in The New York Times. I mean, yes, true enough I guess (Daniel Kokotajlo left a while ago and is not one of the employees this time around)?
如果说有什么不同，我觉得这部分更令人不安：
If anything, I find this part more troubling:
Sheera Frenkel、Dustin Volz 和 Dylan Freedman（《纽约时报》）：
Sheera Frenkel, Dustin Volz and Dylan Freedman (NYTimes):
今年 7 月，安全公司 Hacktron 的研究人员表示，他们曾告知 OpenAI，他们如何借助其竞争对手 Anthropic 创建的一个 AI 模型，找到了入侵该公司系统的方法。他们说，OpenAI 最初对他们的发现不予理会。
In July, researchers at the security company Hacktron said they told OpenAI about how they had found a way to break into the company’s systems with the help of an A.I. model created by its rival Anthropic. OpenAI initially dismissed their findings, they said.
据《纽约时报》看到的通信副本，在通讯平台 Slack 的一个共享频道中，OpenAI 的 Stuckey 先生写道，Hacktron 的研究人员如此不遗余力地展示该公司的漏洞，这“相当可悲”。
In a shared channel on the messaging platform Slack, Mr. Stuckey of OpenAI wrote that it was “pretty sad” that Hacktron’s researchers had gone to such lengths to demonstrate the company’s vulnerabilities, according to copies of the communications seen by The Times.
“我们只是觉得他们在生我们的气，”Hacktron 的研究员 Mohan Pedhapati 在谈到 OpenAI 时说。
“We just felt like they were angry at us,” Mohan Pedhapati, a Hacktron researcher, said of OpenAI.
……Stuckey 先生后来向 Hacktron 道了歉，OpenAI 因这些研究人员披露了该漏洞而奖励了他们 6500 美元。
… Mr. Stuckey later apologized to Hacktron and OpenAI awarded the researchers $6,500 for disclosing the flaw.
如果你是白帽黑客，闯入了 OpenAI，并且小心地没有实际访问任何机密信息，而是将漏洞提交以换取 6,500 美元的漏洞赏金，他们应该极力感谢你，并认为给你的报酬严重偏低。如果你离开时觉得他们很生气，那情况就不妙了。
If you are a white hat hacker, and you break into OpenAI, and are careful to not actually access any confidential information and instead turn in the vulnerability for a $6,500 bug bounty, they should be thanking you profusely and thinking you were severely underpaid. If you come away thinking they are angry, things are not good.
后文还描述了另一个漏洞，OpenAI 为此支付了 500 美元。
There is another vulnerability described later, where OpenAI paid out $500.
Sheera Frenkel、Dustin Volz 和 Dylan Freedman（《纽约时报》）：OpenAI 向该团队支付了 500 美元作为其工作的报酬，Wardle 先生表示，考虑到该漏洞的严重性，这与其他公司相比偏低。
Sheera Frenkel, Dustin Volz and Dylan Freedman (NYTimes): OpenAI gave $500 to the group for its work, which Mr. Wardle said was low compared with what he would expect from other companies given the severity of the flaw.
我不理解为什么这些数字不是更像 50,000 美元，或者也许是 500,000 美元。
I do not understand why these figures are not more like $50,000, or perhaps $500,000.
特朗普承诺将在未来几天内任命一位 AI 事务负责人，此后听起来他打算进入传统的“选举前我们不需要政府”的竞选造势期。
Trump promises to name an AI czar in the next few days, after which it sounds like he intends to go off on the campaign trail in the traditional 'we don't need a government right before an election' period.
这意味着要举行集会。我非常、非常好奇，当特朗普的“AI 很棒，事实上它是超级智能，数据中心很了不起，会拯救你们”这一信息与美国民众互动时会发生什么。当特朗普听到人群的反应时，他会怎么做？他的超能力之一就是察觉人群对他所推销的东西不买账，并转而推销别的东西。
That means rallies. I am very, very curious what happens when Trump's 'AI is great, in fact it's SUPER INTELLIGENCE, and data centers are awesome and will save you' message interacts with the American people. What will Trump do when he hears the reactions of the crowds? One of his superpowers is noticing when the crowds are not vibing with what he is selling, and pivoting to selling something else.