d/acc: one year later

Special thanks to Liraz Siri, Janine Leger and Balvi volunteers for feedback and review

特别感谢 Liraz Siri、Janine Leger 和 Balvi 志愿者的反馈和审查

About a year ago, I wrote an article on techno-optimism , describing my general enthusiasm for technology and the massive benefits that it can bring, as well as my caution around a few specific concerns, largely centered around superintelligent AI, and the risk that it may bring about either doom, or irreversible human disempowerment, if the technology is built in the wrong ways.

大约一年前，我写了一篇关于技术乐观主义的文章，描述了我对技术的总体热情及其所能带来的巨大好处，以及我对一些具体问题的谨慎态度，这些问题主要集中在超级智能人工智能上，以及如果技术以错误的方式构建，它可能带来厄运或不可逆转的人类剥夺权力的风险。

One of the core ideas in my post was the philosophy of : decentralized and democratic, differential defensive acceleration .

我的帖子的核心思想之一是这样的哲学：去中心化和民主、差异化防御加速。

Accelerate technology, but differentially focus on technologies improve our ability to defend, rather than our ability to cause harm, and in technologies that distribute power rather than concentrating it in the hands of a singular elite that decides what is true, false, good or evil on behalf of everyone.

加速技术发展，但有差别地关注技术提高我们的防御能力，而不是我们造成伤害的能力，以及分配权力而不是将权力集中在代表每个人决定什么是真、假、善或恶的单一精英手中的技术。

Defense like in democratic Switzerland and historically quasi-anarchist Zomia , not like the lords and castles of medieval feudalism.

像民主瑞士和历史上准无政府主义的佐米亚那样的防御，而不是像中世纪封建主义的领主和城堡。

In the year since then, the philosophy and ideas have matured significantly.

从那时起的一年里，哲学和思想已经显着成熟。

I talked about the ideas on 80,000 Hours , and have seen many responses, largely positive and some critical.

我在《80,000小时》中谈到了这些想法，并看到了很多回应，大部分是积极的，也有一些批评的。

The work itself is continuing and bearing fruit: we're seeing progress in verifiable open-source vaccines , growing recognition of the value of healthy indoor air, Community Notes continuing to shine, a breakout year for prediction markets as an info tool , ZK-SNARKs in government ID and social media (and securing Ethereum wallets through account abstraction ), open-source imaging tools with applications in medicine and BCI, and more.

这项工作本身仍在继续并取得成果：我们看到可验证的开源疫苗方面取得了进展，人们对健康室内空气的价值的认识不断提高，社区笔记继续大放异彩，预测市场作为信息工具取得了突破性的一年，政府 ID 和社交媒体中的 ZK-SNARK（并通过帐户抽象保护以太坊钱包），在医学和 BCI 中应用的开源成像工具等等。

In the fall, we had the first significant d/acc event: "d/acc Discovery Day" (d/aDDy) at Devcon , which featured a full day of speakers from all pillars of d/acc (bio, physical, cyber, info defense, plus neurotech).

秋天，我们在 Devcon 上举办了第一个重要的 d/acc 活动：“d/acc 发现日”(d/aDDy)，来自 d/acc 所有支柱（生物、物理、网络、信息防御以及神经技术）的演讲者一整天都在其中。

People who have been working on these technologies for years are increasingly aware of each other's work, and people outside are increasingly aware of the larger story: the same kinds of values that motivated Ethereum and crypto can be applied to the wider world .

多年来致力于这些技术的人们越来越了解彼此的工作，而外部的人们也越来越意识到更大的故事：推动以太坊和加密货币的相同价值观可以应用于更广阔的世界。

It's the year 2042.

这是 2042 年。

You're seeing reports in the media about a new pandemic potentially in your city.

您在媒体上看到有关您所在城市可能出现新的流行病的报道。

You're used to these: people get over-excited about every animal disease mutation, and most of them come to nothing.

你已经习惯了这些：人们对每一种动物疾病突变都感到过度兴奋，而大多数突变都毫无结果。

The previous two actual potential pandemics were detected very early through wastewater monitoring and open-source analysis of social media , and stopped completely in their tracks.

前两次实际的潜在流行病很早就通过废水监测和社交媒体的开源分析被发现，并完全停止了发展。

But this time, prediction markets are showing a 60% chance of at least 10,000 cases, so you're more worried.

但这一次，预测市场显示至少有 10,000 个案例的可能性为 60%，因此您会更加担心。

The sequence for the virus was identified yesterday.

昨天确定了该病毒的序列。

Software updates for your pocket air tester to allow it to detect the new virus (from a single breath, or from 15 minutes of exposure to indoor air in a room) are available already.

您的袖珍空气检测仪的软件更新已经可用，使其能够检测新病毒（通过一次呼吸，或通过暴露在房间内的室内空气 15 分钟）。

Open-source instructions and code for generating a vaccine using equipment that can be found in any modern medical facility worldwide should be available within weeks.

使用全球任何现代医疗设施中都可以找到的设备生成疫苗的开源指令和代码应该会在几周内提供。

Most people are not yet taking any action at all, relying mostly on widespread adoption of air filtering and ventilation to protect them.

大多数人还没有采取任何行动，主要依靠广泛采用的空气过滤和通风来保护自己。

You have an immune condition so you're more cautious: your open-source locally-running personal assistant AI, which handles among other tasks navigation and restaurant and event recommendation, is also taking into account real-time air tester and CO2 data to only recommend the safest venues.

你有免疫状况，所以你会更加谨慎：你的开源本地运行的个人助理人工智能，除了处理导航、餐厅和活动推荐等其他任务外，还会考虑实时空气测试仪和二氧化碳数据，仅推荐最安全的场所。

The data is provided by many thousands of participants and devices using ZK-SNARKs and differential privacy to minimize the risk that the data can be leaked or abused for any other purpose (if you want to contribute data to these datasets, there's other personal assistant AIs that verify formal proofs that these cryptographic gadgets actually work).

这些数据由成千上万的参与者和设备使用 ZK-SNARK 和差分隐私提供，以最大限度地降低数据泄露或滥用于任何其他目的的风险（如果您想向这些数据集贡献数据，还有其他个人助理 AI 可以验证这些加密小工具实际工作的正式证据）。

Two months later, the pandemic disappeared: it seems like 60% of people following the basic protocol of putting on a mask if the air tester beeps and shows the virus present, and staying home if they test positive personally, was enough to push the transmission rate, already heavily reduced due to passive heavy air filtering, to below 1.

两个月后，大流行消失了：似乎有 60% 的人遵循基本方案，即如果空气检测仪发出蜂鸣声并显示病毒存在，就戴上口罩；如果他们个人检测呈阳性，就呆在家里，这足以将由于被动的重型空气过滤而已经大大降低的传播率推至 1 以下。

A disease that simulations show might have been five times worse than Covid twenty years ago turns out to be a non-issue today.

模拟显示，一种疾病可能比二十年前的新冠病毒严重五倍，但今天却不再是问题。

One of the most positive takeaways from the d/acc event at Devcon was the extent to which the d/acc umbrella successfully brought people together from very different fields, and got them to actually be interested in each other's work.

Devcon d/acc 活动最积极的收获之一是 d/acc 成功地将来自不同领域的人们聚集在一起，并使他们真正对彼此的工作感兴趣。

Creating events with "diversity" is easy, but making different people with different backgrounds and interests actually relate to each other is hard.

创建具有“多样性”的活动很容易，但让具有不同背景和兴趣的不同人真正相互联系却很难。

I still have memories of being forced to watch long operas in middle school and high school, and personally finding them boring.

我还记得中学和高中时被迫看长剧，个人觉得很无聊。

I knew that I was "supposed to" appreciate them, because if I did not then I would be an uncultured computer science slob, but I did not connect with the content on a more genuine level.

我知道我“应该”欣赏它们，因为如果我不这样做，我就会成为一个没有文化的计算机科学懒汉，但我并没有在更真实的层面上与内容产生联系。

d/acc day did not feel like that at all: it felt like people actually enjoyed learning about very different kinds of work in different fields.

d/acc 日完全没有这样的感觉：感觉人们实际上很喜欢学习不同领域的不同类型的工作。

If we want to create a brighter alternative to domination, deceleration and doom, we need this kind of broad coalition building.

如果我们想创造一个更光明的替代方案来取代统治、减速和厄运，我们就需要建立这种广泛的联盟。

d/acc seemed to be actually succeeding at it, and that alone shows the value of the idea.

d/acc 似乎实际上在这方面取得了成功，仅此一点就表明了该想法的价值。

The core idea of d/acc is simple: decentralized and democratic differential defensive acceleration .

d/acc的核心思想很简单：去中心化、民主的差异化防御加速。

Build technologies that shift the offense/defense balance toward defense, and do so in a way that does not rely on handing over more power to centralized authorities.

构建将进攻/防御平衡转向防御的技术，并且以不依赖于将更多权力移交给中央当局的方式实现这一点。

There is an inherent tie between these two sides: any kind of decentralized, democratic or liberal political structure thrives best when defense is easy, and suffers the most challenge when defense is hard - in those cases, the far more likely outcome is some period of war of all against all, and eventually an equilibrium of rule by the strongest.

这两方面之间存在着内在的联系：任何一种去中心化的、民主的或自由的政治结构在防御容易时会发展得最好，而在防御困难时会遭受最大的挑战——在这种情况下，更有可能的结果是一段时期的所有人对所有人的战争，并最终达到最强者统治的均衡。

One way to understand the importance of trying to be decentralized, defensive and acceleration-minded at the same time, is to contrast it with the philosophy that you get when you give up each of the three.

理解同时尝试去中心化、防御性和加速思维的重要性的一种方法是将其与放弃三者中的每一个时所获得的哲学进行对比。

Decentralized acceleration, but don't care about the "differential defensive" part .

分散加速，但不关心“差动防守”部分。

Basically, be an e/acc , but decentralized.

基本上，是一个 e/acc ，但是分散的。

There are plenty of people who take this approach, some who label themselves d/acc but helpfully describe their focus as "OFFENSE", but also plenty of others who are excited about "decentralized AI" and similar topics in a more moderate way, but in my view put insufficient attention on the "defensive" aspect.

有很多人采取这种方法，有些人给自己贴上 d/acc 的标签，但将他们的关注点描述为“进攻”，但也有很多人以更温和的方式对“去中心化人工智能”和类似主题感到兴奋，但在我看来，对“防御”方面的关注不够。

In my view, this approach may avoid the risk of global human dictatorship by the specific tribe you're worried about, but it doesn't have an answer to the underlying structural problem: in an offense-favoring environment, there's constant ongoing risk of either catastrophe, or someone positioning themselves as a protector and permanently establishing themselves at the top.

在我看来，这种方法可能会避免你所担心的特定部落的全球人类独裁的风险，但它并不能解决根本的结构性问题：在有利于进攻的环境中，持续存在着灾难的风险，或者有人将自己定位为保护者并永久确立自己的最高地位的风险。

In the specific case of AI, it also doesn't have a good answer to the risk of humans as a whole being disempowered compared to AIs.

就人工智能的具体情况而言，与人工智能相比，它也无法很好地解决人类整体被剥夺权力的风险。

Differential defensive acceleration, but don't care about "decentralized and democratic" .

差异化防御加速，却不在乎“去中心化和民主”。

Embracing centralized control for the sake of safety has permanent appeal to a subset of people, and readers are undoubtedly already familiar with many examples, and the downsides of them.

为了安全而采用集中控制对一小部分人具有永久的吸引力，读者无疑已经熟悉了许多例子及其缺点。

Recently, some have worried that extreme centralized control is the only solution to the extremes of future technologies: see this hypothetical scenario where "Everybody is fitted with a ‘freedom tag' – a sequent to the more limited wearable surveillance devices familiar today, such as the ankle tag used in several countries as a prison alternative ...

最近，一些人担心极端的集中控制是解决未来技术极端情况的唯一解决方案：请看这个假设场景，其中“每个人都配备了‘自由标签’——这是当今常见的更有限的可穿戴监控设备的后续产品，例如在几个国家用作监狱替代品的脚踝标签......

encrypted video and audio is continuously uploaded and machine-interpreted in real time".

加密的视频和音频不断上传并实时机器翻译”。

However, centralized control is a spectrum.

然而，集中控制是一个范围。

One milder version of centralized control that's usually overlooked, but is still harmful, is resistance to public scrutiny in biotech (eg.

集中控制的一种较温和的版本通常被忽视，但仍然有害，那就是抵制生物技术领域的公众监督（例如，生物技术领域）。

food , vaccines ), and the closed source norms that allow this resistance to go unchallenged.

食品、疫苗），以及允许这种抵抗不受挑战的闭源规范。

The risk of this approach is, of course, that the center is often itself the source of risk.

当然，这种做法的风险在于，中心本身往往就是风险的来源。

We saw this in Covid, where gain-of-function research funded by multiple major world governments may have been the source of the pandemic, centralized epistemology led to the WHO not acknowledging for years that Covid is airborne, and coercive social distancing and vaccine mandates led to political backlash that may reverberate for decades.

我们在新冠疫情中看到了这一点，由世界多个主要政府资助的功能获得性研究可能是大流行的根源，集中化的认识论导致世界卫生组织多年来不承认新冠病毒是通过空气传播的，强制性的社会疏远和疫苗强制要求导致了政治反弹，可能会持续数十年。

A similar situation may well happen around any risks to do with AI, or other risky technologies.

与人工智能或其他风险技术有关的任何风险很可能会发生类似的情况。

A decentralized approach would better address risks from the center itself.

去中心化的方法可以更好地解决来自中心本身的风险。

Decentralized defense, but don't care about acceleration - basically, attempting to slow down technological progress, or economic degrowth .

分散防御，但不关心加速——基本上，试图减缓技术进步或经济衰退。

The challenge with this strategy is twofold.

这一策略面临双重挑战。

First, on balance technology and economic growth have been massively good for humanity, and any delay to it imposes costs that are hard to overstate .

首先，总的来说，技术和经济增长给人类带来了巨大的好处，任何延迟都会带来难以估量的成本。

Second, in a non-totalitarian world, not advancing is unstable: whoever "cheats" the most and finds plausibly-deniable ways to advance anyway will get ahead.

其次，在一个非极权主义的世界里，不进步是不稳定的：无论谁“作弊”最多，并找到看似合理的可否认的进步方式，谁就会取得进步。

Decelerationist strategies can work to some extent in some contexts: European food being healthier than American food is one example, the success of nuclear non-proliferation so far is another.

在某些情况下，减速主义策略可以在一定程度上发挥作用：欧洲食品比美国食品更健康就是一个例子，迄今为止核不扩散的成功是另一个例子。

But they cannot work forever.

但他们不可能永远工作。

Another way to think about d/acc is to go back to a frame from the Pirate Party movements in Europe in the late 00s: empowerment .

思考 d/acc 的另一种方式是回到 00 年代末欧洲海盗党运动的一个框架：赋权。

The goal is to build a world where we preserve human agency, achieving both the negative freedom of avoiding active interference (whether from other people acting as private citizens, or from governments, or from superintelligent bots) with our ability to shape our own destinies, and the positive freedom of ensuring that we have the knowledge and resources to.

我们的目标是建立一个保留人类能动性的世界，既实现避免主动干预（无论是来自作为普通公民的其他人、政府还是超级智能机器人）的消极自由，也使我们有能力塑造自己的命运，并确保我们拥有足够的知识和资源来实现积极的自由。

This echoes a centuries-long classical liberal tradition, which also includes Stewart Brand's focus on " access to tools " and John Stuart Mill's emphasis on education alongside liberty as key components of human progress - and perhaps, one might add, Buckminster Fuller's desire to see the process of global solving be participatory and widely distributed .

这呼应了长达几个世纪的古典自由主义传统，其中还包括斯图尔特·布兰德（Stewart Brand）对“获得工具的机会”的关注，以及约翰·斯图尔特·密尔（John Stuart Mill）对教育与自由的强调，将其作为人类进步的关键组成部分——也许，有人可能会补充说，巴克明斯特·富勒希望看到全球解决问题的过程具有参与性和广泛分布性。

We can see d/acc as a way of achieving these same goals given the technological landscape of the 21ˢᵗ century.

考虑到 21 世纪的技术格局，我们可以将 d/acc 视为实现这些相同目标的一种方式。

In my post last year, d/acc specifically focused on the defensive technologies: physical defense, bio defense, cyber defense and info defense.

在我去年的帖子中，d/acc 特别关注了防御技术：物理防御、生物防御、网络防御和信息防御。

However, decentralized defense is not enough to make the world great: you also need a forward-thinking positive vision for what humanity can use its newfound decentralization and safety to accomplish.

然而，去中心化的防御不足以让世界变得伟大：您还需要对人类可以利用其新发现的去中心化和安全性来实现的目标有一个前瞻性的积极愿景。

Last year's post did contain a positive vision, in two places:

去年的帖子确实在两个地方包含了积极的愿景：

But these two points felt disconnected from the d/acc argument: "here are some ideas for creating a more democratic and defense-favoring world at the base layer, and by the way here are some unrelated ideas for how we might do superintelligence".

但这两点感觉与 d/acc 的论点脱节：“这里有一些在基础层创建一个更加民主和有利于国防的世界的想法，顺便说一句，这里有一些关于我们如何实现超级智能的不相关的想法”。

However, I think in reality there are some very important connections between what labelled above as "defensive" and "progressive" d/acc technology.

然而，我认为实际上上面标记的“防御”和“进步”d/acc 技术之间存在一些非常重要的联系。

Let's expand the d/acc chart from last year's post, by adding this axis (also, let's relabel it " survive vs thrive ") to the chart and seeing what comes out:

让我们扩展去年帖子中的 d/acc 图表，通过添加此轴（另外，让我们将其重新标记为“生存 vs 繁荣”）到图表中并查看结果：

There is a consistent pattern, across all domains, that the science, ideas and tools that can help us "survive" in one domain are closely related to the science, ideas and tools that can help us "thrive" .

在所有领域都有一个一致的模式，即可以帮助我们在一个领域“生存”的科学、思想和工具与可以帮助我们“繁荣”的科学、思想和工具密切相关。

Some examples:

一些例子：

In addition to this, there are also important cross-dependencies between subject areas:

除此之外，学科领域之间还存在重要的交叉依赖性：

Different people have very different AI timelines.

不同的人有非常不同的人工智能时间表。

Chart from Zuzalu in Montenegro, 2023.

2023 年黑山祖扎鲁的图表。

The argument against my post last year that I found most compelling was a critique from the AI safety community.

我发现去年针对我的帖子最引人注目的论点是来自人工智能安全社区的批评。

The argument goes: "sure, if we have half a century until we get strong AI, we can concentrate our energies and build all of these good things.

他们的论点是：“当然，如果我们有半个世纪的时间才能获得强大的人工智能，我们就可以集中精力打造所有这些美好的事物。

But actually it's looking likely we have three year timelines until AGI, and another three years until superintelligence.

但实际上，我们可能还有三年的时间才能实现通用人工智能，还有三年的时间才能实现超级智能。

And so if we don't want the world to be destroyed or otherwise fall into an irreversible trap, we can't just accelerate the good, we also have to slow down the bad, and this means passing powerful regulations that may make powerful people upset".

因此，如果我们不想让世界被毁灭或陷入不可逆转的陷阱，我们就不能只是加速好事，我们还必须减缓坏事，这意味着要通过强有力的法规，这可能会让有权势的人感到不安”。

In my post last year, I did indeed not call for any specific strategy to "slow down the bad", beyond vague appeals to not build risky forms of superintelligence.

在我去年的帖子中，除了含糊地呼吁不要建立危险形式的超级智能之外，我确实没有呼吁采取任何具体策略来“减缓坏事的发展”。

And so here, it's worth addressing the question directly: if we are living in the least convenient world , where AI risk is high and timelines are potentially five years away, what regulation would I support?

因此，在这里，有必要直接解决这个问题：如果我们生活在最不方便的世界，人工智能风险很高，而且时间表可能还需要五年，我会支持什么监管？

Last year, the main proposed AI regulation was the SB-1047 bill in California .

去年，主要拟议的人工智能法规是加利福尼亚州的 SB-1047 法案。

SB-1047 required developers of the most powerful models (those that take over $100M to train, or over $10M in the case of fine-tunes) to take some safety-testing measures before releasing.

SB-1047 要求最强大的模型（那些花费超过 1 亿美元进行训练的模型，或者在微调的情况下花费超过 1000 万美元）的开发人员在发布之前采取一些安全测试措施。

In addition, it imposed liability on developers of AI models if they take insufficient care.

此外，如果人工智能模型的开发者不够谨慎，它也会让他们承担责任。

Many detractors argued that the bill was " a threat to open source "; I disagreed, because the cost thresholds meant that it affected only the most powerful models: even LLama3 was probably under the threshold .

许多批评者认为该法案“对开源构成威胁”；我不同意，因为成本阈值意味着它只影响最强大的模型：甚至 LLama3 也可能低于阈值。

Looking back, however, I think there was a larger issue with the bill: like most regulation, it was overfitted to the present-day situation .

然而，回顾过去，我认为该法案存在一个更大的问题：像大多数法规一样，它过于适应当今的情况。

The focus on training cost is proving fragile in the face of new technology already: the recent state-of-the-art quality Deepseek v3 model was trained at a cost of only $6 million , and in new models like o1 costs are shifting from training to inference more generally.

面对新技术，对训练成本的关注已经被证明是脆弱的：最近最先进的 Deepseek v3 模型的训练成本仅为 600 万美元，而在像 o1 这样的新模型中，成本正从训练转向更普遍的推理。

Second, the most likely actors who would actually be responsible for an AI superintelligence doom scenario are realistically militaries.

其次，最有可能真正对人工智能超级智能末日场景负责的参与者实际上是军队。

As we have seen in the last half-century of biosecurity (and beyond), militaries are willing to do scary things, and they can easily make mistakes.

正如我们在过去半个世纪（及以后）的生物安全领域所看到的那样，军队愿意做可怕的事情，而且他们很容易犯错误。

AI military use is advancing rapidly today (see Ukraine , Gaza ).

如今，人工智能的军事应用正在迅速发展（参见乌克兰、加沙）。

And any safety regulation that a government passes, by default would exempt their own military, and corporations that cooperate closely with the military.

默认情况下，政府通过的任何安全法规都会豁免本国军队以及与军队密切合作的公司。

That said, these arguments are not reasons to throw up our hands and do nothing.

也就是说，这些论点并不是我们举手不做的理由。

Rather, we can use them as a guide, and try to come up with rules that would trigger these concerns the least.

相反，我们可以使用它们作为指导，并尝试制定出最少引发这些担忧的规则。

If someone acts in some way that causes legally actionable damage, they could be sued .

如果某人的行为造成了可依法起诉的损害，他们可能会被起诉。

This does not solve the problem of risks from militaries and other "above-the-law" actors, but it is a very general-purpose approach that avoids overfit, and is often supported by libertarian-leaning economists for this exact reason.

这并不能解决来自军队和其他“高于法律”行为者的风险问题，但它是一种非常通用的方法，可以避免过度拟合，并且经常因为这个原因而受到倾向自由主义的经济学家的支持。

The primary targets for liability that have been considered so far are:

迄今为止已考虑的主要责任目标是：

Putting liability on users feels most incentive-compatible.

让用户承担责任感觉最符合激励措施。

While the link between how a model is developed and how it ends up being used is often unclear , the user decides exactly how the AI is used.

虽然模型的开发方式和最终使用方式之间的联系通常不清楚，但用户可以准确决定人工智能的使用方式。

Liability on users creates a strong pressure to do AI in what I consider the right way: focus on building mecha suits for the human mind, not on creating new forms of self-sustaining intelligent life .

对用户的责任产生了强大的压力，要求我们以我认为正确的方式进行人工智能：专注于为人类思维打造机甲套装，而不是创造新形式的自我维持的智能生命。

The former responds regularly to user intent, and so would not cause catastrophic actions unless the user wanted them to.

前者定期响应用户意图，因此不会导致灾难性的操作，除非用户希望这样做。

The latter would have the greatest risk of going off and creating a classic "AI going rogue" scenario.

后者将面临最大的风险，导致出现经典的“人工智能失控”场景。

Another benefit of putting liability as close to end usage as possible is that it minimizes the risk that liability will lead to people taking actions that are harmful in other ways (eg.

将责任尽可能接近最终用途的另一个好处是，它可以最大限度地减少责任导致人们采取其他方式有害的行为的风险（例如，

closed source, KYC and surveillance, state/business collusion to clandestinely restrict users as with eg.

闭源、KYC 和监视、国家/企业串通秘密限制用户，例如。

debanking, locking out large regions of the world).

去银行化，封锁世界大部分地区）。

There is a classic argument against putting liability solely on users: users may be regular individuals without too much money, or even anonymous, leaving no one that could actually pay for a catastrophic harm.

有一个反对仅将责任归咎于用户的经典论点：用户可能是没有太多钱的普通人，甚至是匿名的，因此没有人能够真正为灾难性的伤害买单。

This argument can be overstated: even if some users are too small to be held liable, the average customer of an AI developer is not, and so AI developers would still be incentivized to build products that can give their users assurance that they won't face high liability risk.

这个论点可能被夸大了：即使一些用户太小而无法承担责任，但人工智能开发人员的普通客户却不是，因此人工智能开发人员仍然会被激励去构建可以让用户保证他们不会面临高责任风险的产品。

That said, it is still a valid argument, and needs to be addressed.

也就是说，这仍然是一个有效的论点，需要解决。

You need to incentivize someone in the pipeline who has the resources to take the appropriate level of care to do so, and deployers and developers are both easily available targets who still have a lot of influence over how safe or unsafe a model is.

您需要激励管道中拥有资源的人员采取适当的谨慎措施来做到这一点，而部署人员和开发人员都是容易获得的目标，他们对模型的安全或不安全程度仍然有很大的影响力。

Deployer liability seems reasonable.

部署者的责任似乎是合理的。

A commonly cited concern is that it would not work for open-source models, but this seems manageable, especially since there is a high chance that the most powerful models will be closed source anyway (and if they turn out to be open, then while deployer liability does not end up very useful, it also does not cause much harm).

人们普遍担心的是，它不适用于开源模型，但这似乎是可以管理的，特别是因为最强大的模型很可能是闭源的（如果它们最终是开放的，那么虽然部署者责任最终不会非常有用，但也不会造成太大伤害）。

Developer liability has the same concern (though with open source models there is some speed-bump of needing to fine-tune a model to cause it to do some originally disallowed thing), but the same counterargument applies.

开发人员责任也有同样的担忧（尽管对于开源模型，需要微调模型以使其执行一些最初不允许的事情），但同样的反驳也适用。

As a general principle, putting a "tax" on control , and essentially saying "you can build things you don't control , or you can build things you do control, but if you build things you do control, then 20% of the control has to be used for our purposes", seems like a reasonable position for legal systems to have.

作为一般原则，对控制权征收“税”，本质上是说“你可以建造你无法控制的东西，或者你可以建造你可以控制的东西，但如果你建造你可以控制的东西，那么20%的控制权必须用于我们的目的”，这似乎是法律体系的合理立场。

One idea that seems under-explored is putting liability on other actors in the pipeline, who are more guaranteed to be well-resourced.

一个似乎尚未得到充分探讨的想法是将责任归咎于管道中的其他参与者，这些参与者更能保证拥有充足的资源。

One idea that is very d/acc friendly is to put liability on owners or operators of any equipment that an AI takes over (eg.

一个对 d/acc 非常友好的想法是让人工智能接管的任何设备的所有者或操作员承担责任（例如，

by hacking) in the process of executing some catastrophically harmful action.

通过黑客攻击）在执行一些灾难性的有害操作的过程中。

This would create a very broad incentive to do the hard work to make the world's (especially computing and bio) infrastructure as secure as possible.

这将产生非常广泛的激励，促使人们努力工作，使世界（尤其是计算和生物）基础设施尽可能安全。

If I was convinced that we need something more "muscular" than liability rules, this is what I would go for.

如果我确信我们需要比责任规则更“强有力”的东西，那么我就会这么做。

The goal would be to have the capability to reduce worldwide available compute by ~90-99% for 1-2 years at a critical period, to buy more time for humanity to prepare.

目标是能够在关键时期将全球可用计算量减少约 90-99%，持续 1-2 年，为人类争取更多时间做好准备。

The value of 1-2 years should not be overstated: a year of "wartime mode" can easily be worth a hundred years of work under conditions of complacency.

1-2年的价值不应被夸大：一年的“战时模式”很容易抵得上自满的情况下一百年的工作。

Ways to implement a " pause " have been explored, including concrete proposals like requiring registration and verifying location of hardware.

已经探索了实现“暂停”的方法，包括要求注册和验证硬件位置等具体建议。

A more advanced approach is to use clever cryptographic trickery: for example, industrial-scale (but not consumer) AI hardware that gets produced could be equipped with a trusted hardware chip that only allows it to continue running if it gets 3/3 signatures once a week from major international bodies, including at least one non-military-affiliated.

一种更先进的方法是使用巧妙的加密技巧：例如，生产的工业规模（但不是消费级）人工智能硬件可以配备可信硬件芯片，只有在每周一次从主要国际机构（包括至少一个非军事附属机构）获得 3/3 签名的情况下才允许其继续运行。

The signatures would be device-independent (if desired, we could even require a zero-knowledge proof that they were published on a blockchain), so it would be all-or-nothing: there would be no practical way to authorize one device to keep running without authorizing all other devices.

签名将是独立于设备的（如果需要，我们甚至可以要求零知识证明它们是在区块链上发布的），因此这将是全有或全无：没有实用的方法可以在不授权所有其他设备的情况下授权一台设备继续运行。

This feels like it "checks the boxes" in terms of maximizing benefits and minimizing risks:

这感觉就像它在利益最大化和风险最小化方面“勾选了复选框”：

Hardware regulation is being strongly considered already, though generally through the frame of export controls , which inherently have a "we trust our side, but not the other side" philosophy.

人们已经在大力考虑硬件监管，尽管通常是通过出口管制框架，其本质上具有“我们信任我们一方，但不信任另一方”的理念。

Leopold Aschenbrenner has famously advocated that the US should race to get a decisive advantage and then essentially force China to sign a protocol limiting how many boxes they are allowed to run.

利奥波德·阿申布伦纳 (Leopold Aschenbrenner) 曾主张美国应该竞相获得决定性优势，然后从本质上迫使中国签署一项限制其运行盒子数量的协议。

To me, this approach seems risky, and could combine the flaws of multipolar races and centralization.

对我来说，这种方法似乎有风险，并且可能结合多极种族和集权的缺陷。

If we have to limit people, it seems better to limit everyone on an equal footing, and do the hard work of actually trying to cooperate to organize that instead of one party seeking to dominate everyone else.

如果我们必须限制人，那么似乎更好的是在平等的基础上限制每个人，并努力实际尝试合作来组织这一点，而不是一个政党试图主宰其他所有人。

Both of these strategies (liability and the hardware pause button) have holes in them, and it's clear that they are only temporary stopgaps: if something becomes possible to do on a supercomputer at time T, it will likely be possible on a laptop at time T + 5 years.

这两种策略（责任和硬件暂停按钮）都有漏洞，很明显它们只是暂时的权宜之计：如果某件事在时间 T 可以在超级计算机上完成，那么在时间 T + 5 年在笔记本电脑上也可能实现。

And so we need something more stable to buy time for .

所以我们需要一些更稳定的东西来争取时间。

Many d/acc technologies are relevant here.

许多 d/acc 技术都与此相关。

We can look at the role of d/acc tech as follows: if AI takes over the world, how would it do so?

我们可以这样看待 d/acc 技术的作用：如果人工智能接管世界，它将如何做到这一点？

As briefly mentioned above, liability rules are a naturally d/acc-friendly style of regulation , because they can very efficiently motivate all parts of the world to adopt these defenses and take them seriously.

正如上面简要提到的，责任规则自然是一种对 d/acc 友好的监管方式，因为它们可以非常有效地激励世界各地采用这些防御措施并认真对待它们。

Taiwan has been experimenting with liability for false advertising recently, which can be viewed as one example of using liability to encourage info defense.

台湾最近一直在尝试对虚假广告承担责任，这可以被视为利用责任来鼓励信息防御的一个例子。

We should not be too enthusiastic about putting liability everywhere, and remember the benefits of plain old freedom in enabling the little guy to participate in innovation without fear of lawsuits, but where we do want a stronger push to be secure, liability can be quite flexible and effective.

我们不应该太热衷于将责任无处不在，并记住普通自由的好处，使小人物能够参与创新而不必担心诉讼，但在我们确实希望更有力地推动安全的地方，责任可以相当灵活和有效。

Much of d/acc goes far beyond typical blockchain topics: biosecurity, BCI and collaborative discourse tools seem far away from things that a crypto person normally talks about.

d/acc 的大部分内容远远超出了典型的区块链主题：生物安全、BCI 和协作讨论工具似乎与加密货币人士通常谈论的事情相去甚远。

However, I think there are some important ties between crypto and d/acc, particularly:

然而，我认为加密货币和 d/acc 之间存在一些重要的联系，特别是：

Cursive, an app that uses fully homomorphic encryption (FHE) to allow users to identify areas of common interest with other users, while preserving privacy.

Cursive，一款使用完全同态加密 (FHE) 的应用程序，允许用户识别与其他用户共同感兴趣的领域，同时保护隐私。

This was used at Edge City, one of the many offshoots of Zuzalu , in Chiang Mai.

这是在 Edge City 使用的，它是清迈 Zuzalu 的众多分支之一。

In addition to these direct intersections, there is also another crucial shared point of interest: funding mechanisms .

除了这些直接交集之外，还有另一个重要的共同利益点：融资机制。

One of my ongoing interests is coming up with better mechanisms to fund public goods : projects that are valuable to very large groups of people, but that do not have a naturally accessible business model.

我持续的兴趣之一是提出更好的机制来资助公共产品：对大量人群有价值的项目，但没有自然可及的商业模式。

My past work on this includes my contributions to quadratic funding and its use in Gitcoin Grants , retro PGF , and more recently deep funding .

我过去在这方面的工作包括我对二次融资的贡献及其在 Gitcoin Grants、复古 PGF 和最近的深度融资中的使用。

Many people are skeptical of public goods as a concept.

许多人对公共物品这个概念持怀疑态度。

The skepticism generally comes from two sources:

怀疑论通常有两个来源：

These are important critiques, and good critiques.

这些都是重要的批评，也是好的批评。

However, I would argue that strong decentralized public goods funding is essential to a d/acc vision, because a key d/acc goal (minimizing central points of control) inherently frustrates many traditional business models.

然而，我认为强大的去中心化公共产品资金对于 d/acc 愿景至关重要，因为关键的 d/acc 目标（最小化中心控制点）本质上会挫败许多传统的商业模式。

It is possible to build successful businesses on open source - several Balvi grantees are doing so - but in some situations it is hard enough that important projects needs extra ongoing support.

在开源基础上建立成功的企业是可能的——一些 Balvi 受资助者正在这样做——但在某些情况下，重要的项目需要额外的持续支持是非常困难的。

Hence, we have to do the hard thing, and figure out how to do public goods funding in a way that addresses both of the above critiques .

因此，我们必须做困难的事情，找出如何以解决上述批评的方式进行公共物品资助。

The solution to the first problem is basically credible neutrality and decentralization .

第一个问题的解决基本上是可信中立和去中心化。

Central planning is problematic because it gives control to elites who might turn abusive, and because it often overfits to the present-day situation and becomes less and less effective over time.

中央计划是有问题的，因为它把控制权交给了可能会滥用职权的精英，而且它常常过于适应当前的情况，并且随着时间的推移变得越来越无效。

Quadratic funding and similar mechanisms were precisely about funding public goods in a way that is as credibly neutral and (architecturally and politically) decentralized as possible.

二次方融资和类似机制正是以一种尽可能可信的中立和（结构上和政治上）去中心化的方式为公共产品提供资金。

The second problem is more challenging.

第二个问题更具挑战性。

With quadratic funding, a common critique has been that it quickly becomes a popularity contest , requiring project funders to spend a lot of effort publicly campaigning.

对于二次方资助，一个常见的批评是它很快就变成了一场人气竞赛，要求项目资助者花费大量精力进行公开宣传。

Furthermore, projects that are "in front of people's eyeballs" (eg.

此外，“在人们眼前”的项目（例如

end-user applications) get funded, but projects that are more in the background (the archetypal " dependency maintained by a guy in Nebraska ") don't get any funding at all.

最终用户应用程序）得到资助，但更多在后台的项目（典型的“由内布拉斯加州的一个人维护的依赖”）根本没有得到任何资助。

Optimism retro funding relies on a smaller number of expert badge holders; here, popularity contest effects are diminished, but social effects of having close personal ties with the badge holders are magnified.

乐观复古资金依赖于少数专家徽章持有者；在这里，人气竞赛的影响减弱了，但与徽章持有者建立密切个人关系的社会影响却被放大了。

Deep funding is my own latest effort to solve this problem.

深度资助是我自己为解决这个问题所做的最新努力。

Deep funding has two primary innovations:

深度融资有两个主要创新点：

But deep funding is only the latest example; there have been other public goods funding mechanism ideas before, and there will be many more in the future.

但深度融资只是最新的例子。之前已经有过其他公共产品融资机制的想法，未来还会有更多。

allo.expert does a good job of cataloguing them.

allo.expert 很好地对它们进行了分类。

The underlying goal is to create a societal gadget that can fund public goods with a level of accuracy, fairness and open entry that at least approximates the way that markets fund private goods .

根本目标是创建一个社会工具，能够以一定程度的准确性、公平性和开放性为公共产品提供资金，至少类似于市场为私人产品提供资金的方式。

It does not have to be perfect; after all, markets are far from perfect themselves.

它不必是完美的；毕竟，市场本身远非完美。

But it should be effective enough that developers working on top-quality open-source projects that benefit everyone can afford to keep doing so without feeling the need to make unacceptable compromises.

但它应该足够有效，让致力于造福每个人的高质量开源项目的开发人员能够继续这样做，而不需要做出不可接受的妥协。

Today, the leading projects in most d/acc subject areas: vaccines, BCI , "borderline BCI" like wrist myoelectrics and eye tracking , anti-aging medicines, hardware, etc, are proprietary.

如今，大多数 d/acc 学科领域的领先项目：疫苗、BCI、“边缘 BCI”（如腕部肌电和眼动追踪）、抗衰老药物、硬件等，都是专有的。

This has big downsides in terms of securing public trust, as we have seen in many of the above areas already .

正如我们在上述许多领域已经看到的那样，这在确保公众信任方面存在很大的缺点。

It also shifts attention toward competitive dynamics ("OUR TEAM must win this critical industry!"), and away from the larger competition of making sure these technologies come fast enough to be there to protect us in a world of superintelligent AI.

它还将注意力转移到竞争动态上（“我们的团队必须赢得这个关键行业！”），而不是确保这些技术足够快地发展以在超级智能人工智能世界中保护我们的更大竞争。

For these reasons, robust public goods funding can be a strong booster of openness and freedom.

出于这些原因，强劲的公共产品资金可以成为开放和自由的有力推动者。

This is another way in which the crypto community can help d/acc: by putting serious effort into exploring these funding mechanisms and making them work well within its own context, preparing them for much wider adoption for open-source science and technology more generally.

这是加密社区可以帮助 d/acc 的另一种方式：通过认真努力探索这些融资机制，并使它们在自己的背景下良好运行，为更广泛地广泛采用开源科学和技术做好准备。

The next few decades bring important challenges.

未来几十年将带来重大挑战。

There are two challenges that have recently been on my mind:

最近我想到了两个挑战：

However, each of these challenges has a silver lining.

然而，每一个挑战都有一线希望。

First, we now have very powerful tools to do our remaining work more quickly :

首先，我们现在拥有非常强大的工具来更快地完成剩余的工作：

Second, now that many principles that we hold dear are no longer occupied by a few particular segments of the old guard, they can be reclaimed by a broad coalition that anyone in the world is welcome to join .

其次，既然我们所珍视的许多原则不再被少数特定的守旧派所占据，它们就可以被一个广泛的联盟重新夺回，欢迎世界上任何人加入。

This is probably the biggest upside of recent political "realignments" around the world, and one that is worth taking advantage of.

这可能是全球近期政治“重组”的最大好处，也是值得利用的。

Crypto has already done an excellent job taking advantage of this and finding global appeal; d/acc can do the same.

加密货币已经利用这一点做得非常出色，并找到了全球吸引力； d/acc 也可以做同样的事情。

Access to tools means that we are able to adapt and improve our biologies and our environments, and the "defense" part of d/acc means that we are able to do this without infringing on others' freedom to do the same.

获得工具意味着我们能够适应和改善我们的生物学和环境，而 d/acc 的“防御”部分意味着我们能够在不侵犯他人做同样事情的自由的情况下做到这一点。

Liberal pluralist principles mean we can have a lot of diversity on how this is done, and our commitment to common humanity goals means it should get done.

自由多元主义原则意味着我们在实现这一目标的方式上可以有很大的多样性，而我们对共同人类目标的承诺意味着它应该完成。

We, humans, continue to be the brightest star.

我们人类仍然是最闪亮的星星。

The task ahead of us, of building an even brighter 21ˢᵗ century that preserves human survival and freedom and agency as we head toward the stars, is a challenging one.

我们面前的任务是建设一个更加光明的 21 世纪，在我们迈向星空的过程中保护人类的生存、自由和能动性，这是一项艰巨的任务。

But I am confident that we are up to it.

但我相信我们能够做到。