Concise Summary简洁概述
Yudkowsky draws a deep parallel between libertarianism and Science: both are pragmatic systems that distrust beautiful-sounding theories and the people who implement them, and both try to harness individual flaws — selfishness, stubbornness — to produce good outcomes at the group level. Libertarianism routes selfish agents through voluntary transactions to produce positive-sum growth. Science routes stubborn theorists through falsifiable experiments to produce a steady stream of knowledge. The upshot: Science is not a mere approximation to Bayesian ideal rationality. It is built on the assumption that you are too biased and self-deceiving to apply probability theory correctly — so it demands a definitive experiment instead. Subordinating Science to personal Bayesian reasoning is therefore a much riskier step than it sounds.
Yudkowsky 在自由意志主义与科学之间发现了一个深刻的类比：两者都是实用主义系统，不信任听起来美妙的理论及实施这些理论的人，并试图驾驭个体缺陷——自私、固执——来在群体层面产生好的结果。自由意志主义通过自愿交易引导自私的行为者，产出正和增长；科学通过可证伪实验引导固执的理论家，产出稳定的知识流。结论是：科学并非贝叶斯理想理性的简单近似。它建立在一个假设之上：你有太多偏见和自欺，无法正确运用概率论——因此它要求一个决定性的实验来替代。将科学从属于个人的贝叶斯推理，因此是一个比听起来更危险的步骤。
Infographic信息图
Beautiful theories are dangerous
美好的理论是危险的
History shows that when someone gains enough power to enact their lovely theory, the result is Revolutionary France or Soviet Russia — not the predicted paradise.
历史表明，当某人获得足够权力来实施其美好理论时，结果是大革命时的法国或苏联——而非预言中的天堂。
Harness flaws, don't eliminate them
驾驭缺陷，而非消除它
Libertarianism doesn't demand unselfish people; it channels selfishness through voluntary exchange. Science doesn't demand unbiased scientists; it channels stubbornness through falsifiable experiments.
自由意志主义不要求人们无私；它通过自愿交换来疏导自私。科学不要求科学家客观无偏；它通过可证伪实验来驾驭固执。
Experiments beat rationality
实验胜过理性推断
Science doesn't trust your ability to do Bayesian reasoning correctly — it wants a definitive experiment because your rationality is likely compromised by motivated reasoning.
科学不信任你正确进行贝叶斯推理的能力——它要求一个决定性实验，因为你的理性很可能已被动机性推理所腐蚀。
Overriding Science with Bayes is risky
以贝叶斯推翻科学是危险的
Subordinating Science's verdict to personal Bayesian judgment — as with Many-Worlds — is not trivial. Science is built on the premise that you're too biased to reason correctly without experimental constraints.
将科学的裁决从属于个人的贝叶斯判断——如在多世界诠释上——并非小事。科学建立在这样的前提上：没有实验约束，你的推理偏差太大。
Detailed Summary详细概述
The Scott Aaronson Springboard
Yudkowsky opens with a comparison borrowed from Scott Aaronson: both Many-Worlds interpretation and libertarianism are cases of "bullet-swallowing" — accepting conclusions most people find uncomfortable because they follow logically from premises almost everyone accepts. This launches the essay's central claim: libertarianism and Science share a deeper structural kinship.
Three Structural Parallels
Yudkowsky identifies three features both systems share:
- Pragmatic distrust of reasonable-sounding arguments. Both start from the empirical observation that clever-sounding theories, when implemented, tend to go badly wrong.
- Building systems more trustworthy than the people in them. Rather than selecting wise leaders or virtuous scientists, both try to construct institutional arrangements that produce good outcomes despite human flaws.
- Harnessing human flaws to power the system. They don't try to eliminate selfishness or stubbornness — they channel those tendencies productively.
The Libertarian Case
The core libertarian argument, as Yudkowsky presents it, is historically motivated: theories that promised societal flourishing via top-down regulation have a poor track record. Revolutionary France and Soviet Russia are the canonical examples of what happens when someone with enough power tries to implement a lovely theory. So the libertarian solution is to constrain power rather than trust theory-makers: require voluntary transactions, restrain violence, enforce contracts. Individual selfishness then powers a globally productive system. Yudkowsky is careful to note this is an argument about graceful degradation, not perfection — libertarianism degrades less awkwardly than alternatives in real life.
The Scientific Revolution as Distrust
The Scientific Revolution, Yudkowsky argues, was not merely "trust us, not Aristotle" — that would have faded like the French Revolution. Its lasting power came from a stranger philosophy: trust no one, not even ourselves. The key move was insisting that we must talk to Nature and actually listen. But implementation required dealing with stubborn scientists who won't accept experimental verdicts. Rather than trying to train impartiality into them, Science harnesses their stubbornness: "Make a new experimental prediction, and do the experiment. If you're right, and it's replicated, you win." This creates incentives to actually risk falsification — only by accepting possible defeat can you win. Replication requirements further incentivize honesty.
And so the stubbornness of individual scientists is harnessed to produce a steady stream of knowledge at the group level. The System is somewhat more trustworthy than its parts.
The Critical Upshot
Both systems have caveats. Libertarianism secretly relies on people being sufficiently prosocial (you tip at restaurants you'll never visit again). Science relies on scientists not committing sins so egregious they can't rationalize away. Both degrade when these conditions fail — but gracefully.
The final, sharpest point: Science is not reconcilable with just using Bayesian reasoning. Science doesn't trust your rationality. It doesn't rely on your ability to use probability theory correctly. It wants a definitive experiment. The seemingly reasonable argument that Bayes explains why Science works — and therefore that personal Bayesian judgment should be able to override Science's verdicts — is not a trivial step. Science is built around the assumption that you are too stupid and self-deceiving to use Solomonoff induction correctly. That assumption deserves to give you pause.
斯科特·阿伦森的跳板
Yudkowsky 以斯科特·阿伦森的一个类比开场：多世界诠释与自由意志主义都是「吞子弹」的案例——接受大多数人觉得令人不适的结论，因为它们从几乎每个人都接受的前提中逻辑地推出。这引出了文章的核心主张：自由意志主义与科学共享更深的结构亲缘性。
三个结构平行点
Yudkowsky 识别出两个系统共享的三个特征：
- 对听起来合理的论证保持实用主义的不信任。 两者都从经验观察出发：听起来聪明的理论，在实施时往往会出严重差错。
- 建立比其中的人更可信赖的系统。 不是挑选智慧的领导者或有德行的科学家，而是构建尽管有人类缺陷也能产出好结果的制度安排。
- 驾驭人类缺陷来驱动系统。 它们不试图消除自私或固执——而是将这些倾向引向有生产力的方向。
自由意志主义的论证
Yudkowsky 所呈现的自由意志主义核心论证，是有历史动机的：承诺通过自上而下的管制来繁荣社会的理论，有着糟糕的记录。大革命时的法国和苏联，是拥有足够权力者试图实施美好理论时会发生什么的经典案例。所以自由意志主义的解决方案是约束权力，而非信任理论制定者：要求自愿交易，限制暴力，执行合同。个体自私随后驱动一个全球生产性的系统。Yudkowsky 特别指出，这是一个关于优雅退化而非完美的论证——在现实生活中，自由意志主义比各种替代方案退化得更不尴尬。
作为不信任的科学革命
科学革命，Yudkowsky 认为，并不仅仅是「信任我们，不要信任亚里士多德」——那样的话，它早就像法国大革命一样转瞬即逝了。它的持久力来自一种更奇特的哲学：不信任任何人，包括我们自己。 关键举措是坚持我们必须与自然交谈，并真正倾听它的回答。但实施这一点需要应对固执的科学家——他们不愿接受实验的裁决。与其试图训练他们变得中立，科学驾驭了他们的固执：「做出新的实验预测，再做实验。如果你是对的，且实验被重复验证，你就赢了。」这创造了真正冒证伪风险的激励——只有接受可能失败，才能赢得胜利。重复验证要求进一步激励诚实。
于是，个别科学家的固执在群体层面被驾驭，产出了稳定的知识流。这个系统在某种程度上比其组成部分更可信赖。
关键推论
两个系统都有附带条件。自由意志主义暗地里依赖人们有足够的亲社会性（你在一家再也不会光顾的餐厅留了小费）。科学依赖科学家不犯大到无法合理化的错误。当这些条件失败时，两者都会退化——但退化得很优雅。
最后也是最尖锐的一点：科学不能与单纯使用贝叶斯推理调和。科学不信任你的理性。它不依赖你正确使用概率论的能力。它要求一个决定性实验。贝叶斯定理解释了科学为何有效——因此个人的贝叶斯判断应该能够推翻科学裁决——这个看似合理的论证，不是一个微不足道的步骤。科学建立在这样的假设之上：你太愚蠢、太自欺，无法正确使用索洛莫诺夫归纳。这个假设应该让你停下来好好想想。
FAQ常见问答
What is the libertarianism–Science analogy actually claiming?自由意志主义与科学的类比究竟在主张什么？
Yudkowsky is not endorsing libertarianism as a policy. He is using it as a structural comparison: both systems are pragmatic responses to the unreliability of smart-sounding human reasoning. Both distrust beautiful theories, both accept human flaws as given, and both try to build institutions that harness those flaws productively rather than assuming they can be argued or trained away.
Yudkowsky 并非在政策上背书自由意志主义。他用它作为结构性比较：两种系统都是对聪明人推理不可靠性的实用主义回应。两者都不信任美好理论，都把人类缺陷视为既定事实，并都试图建立能生产性地驾驭这些缺陷的制度，而非假设可以通过论证或训练来消除它们。
Why does Yudkowsky say Science "doesn't trust your rationality"?为什么 Yudkowsky 说科学「不信任你的理性」？
Science is built on the premise that individual humans — including smart, well-intentioned scientists — are too prone to motivated reasoning and self-deception to be trusted to simply reason their way to truth. That is precisely why Science demands external, falsifiable experiments rather than relying on anyone's probability-theoretic calculations, however sophisticated.
科学建立在这样的前提上：个体人类——包括聪明、善意的科学家——太容易受动机性推理和自我欺骗的影响，无法被信任仅凭推理找到真相。这正是科学要求外部可证伪实验，而非依赖任何人的概率论计算（无论多么复杂）的原因。
Why is it risky to override Science with Bayesian reasoning?为什么用贝叶斯推理凌驾于科学之上是危险的？
The essay's implicit warning, raised through the Many-Worlds example, is that invoking personal Bayesian judgment to override a scientific consensus is exactly the kind of move that feels like bullet-swallowing rationality but is structurally identical to the lovely-theory overconfidence that Science was designed to guard against. Science is built around assuming you might be the one who is too biased to reason correctly.
文章通过多世界诠释的例子隐含地警告：援引个人贝叶斯判断来凌驾于科学共识之上，这恰恰是那种感觉像是「吞子弹」式理性的举动，但结构上与科学被设计来防范的「美好理论过度自信」完全相同。科学建立在假设你可能正是那个偏见太大而无法正确推理的人之上。
What does "graceful degradation" mean for these systems?「优雅退化」对于这些系统意味着什么？
Neither libertarianism nor Science works perfectly in the real world — libertarianism requires some prosocial behavior; Science requires scientists who don't commit unreplicable fraud. But when conditions fall short, both systems fail less catastrophically than alternatives. The argument is comparative: not "this is perfect" but "this degrades less awkwardly than any known alternative."
自由意志主义和科学在现实世界中都不能完美运作——自由意志主义需要一定的亲社会行为；科学需要不造假的科学家。但当条件不足时，两个系统都比替代方案失败得不那么灾难性。这个论证是比较性的：不是「这是完美的」，而是「这比任何已知的替代方案退化得更不尴尬」。
Does the essay say Science and Bayesianism are incompatible?这篇文章是说科学与贝叶斯主义不相容吗？
Not exactly. Yudkowsky acknowledges there is a "reasonable-sounding argument" that Bayes's Theorem is the hidden structure explaining why Science works. But he insists that subordinating Science to Bayesian judgment — letting personal probabilistic reasoning override experimental verdicts — is "not a trivial step." It requires trusting your own rationality in a way that Science was designed to render unnecessary.
并非完全如此。Yudkowsky 承认有一个「听起来合理的论证」：贝叶斯定理是解释科学为何有效的隐藏结构。但他坚持认为，将科学从属于贝叶斯判断——让个人概率推理凌驾于实验裁决之上——「并非微不足道的步骤」。它要求信任自己的理性，而这恰恰是科学被设计来使之不必要的。
How does the Many-Worlds example fit into the essay?多世界诠释的例子在文章中起什么作用？
Yudkowsky has previously argued that Bayesian reasoning accepts Many-Worlds while Science (the social process, focused on experimental tests) rejects it. He uses this as his test case for the tension: it is precisely the kind of situation where someone might invoke Bayesian bullet-swallowing to override Science. The essay's point is that doing so requires a high level of confidence in your own rationality — which is exactly what Science is designed to make you distrust.
Yudkowsky 此前论证过，贝叶斯推理接受多世界诠释，而科学（这个聚焦于实验检验的社会过程）则拒绝它。他用这个例子作为这种张力的测试案例：这正是那种人们可能援引贝叶斯式「吞子弹」来凌驾于科学之上的情境。文章的要点是：这样做需要对自己理性能力的高度信任——而这恰恰是科学被设计来让你不信任的那种东西。
In-depth Analysis · Pros & Cons深入解读 · 优缺点
This short essay does something structurally ambitious: it reframes both Science and libertarianism as solutions to a common problem — the untrustworthiness of individual human rationality — and uses that reframing to argue against the seemingly obvious move of subordinating Science to Bayesian reasoning.
这篇短文在结构上做了一件雄心勃勃的事：它将科学和自由意志主义重新定义为同一个问题的解决方案——个体人类理性的不可信赖性——并用这一重新定义来反对那个看似显然的举动：将科学从属于贝叶斯推理。
- A genuinely illuminating structural analogy真正具有启发性的结构类比The libertarianism–Science parallel is not decorative. Identifying that both systems share pragmatic distrust, flaw-harnessing, and graceful degradation gives the reader a new lens for understanding what Science actually is for.自由意志主义与科学的平行并非装饰性的。识别出两个系统都共享实用主义不信任、驾驭缺陷和优雅退化，为读者提供了理解科学究竟为何而存在的全新视角。
- The warning against Bayesian overconfidence is the real payload对贝叶斯过度自信的警告是真正的核心The essay's most important move is its final one: pointing out that Science was designed around the assumption that you are too biased to reason correctly. This is a genuine and often-neglected check on epistemic overconfidence among rationalists.文章最重要的举动是最后那一步：指出科学被设计时基于这样的假设——你有太多偏见，无法正确推理。这是对理性主义者认知过度自信的真正且常被忽视的制衡。
- Graceful degradation as a realistic criterion优雅退化作为现实主义标准Framing both systems in terms of how they fail, rather than how they succeed in ideal conditions, is honest and sophisticated. It avoids the trap of defending institutions only in their idealized form.从失败方式而非理想条件下如何成功来框定两个系统，是诚实而复杂的。这避免了只为制度的理想化形式辩护的陷阱。
- Historical grounding of the libertarian argument自由意志主义论证的历史基础Presenting libertarianism as a historically-motivated distrust of top-down theorizing — not an ideological commitment to perfection — makes the analogy with Science more defensible and less politically loaded.将自由意志主义呈现为对自上而下理论化的历史动机不信任——而非对完美的意识形态承诺——使与科学的类比更可辩护，政治色彩也更少。
- The analogy obscures important disanalogies类比掩盖了重要的不类比之处Libertarianism and Science differ in who bears the costs of being wrong. A failed scientific theory costs reputation; a failed libertarian policy can cost millions of lives. The essay treats both as symmetric "graceful degradation" cases, which papers over this asymmetry.自由意志主义与科学在谁承担犯错代价这一点上有所不同。失败的科学理论损失声誉；失败的自由意志主义政策可能损失数百万条生命。文章将两者都视为对称的「优雅退化」案例，这掩盖了这种不对称性。
- The argument against Bayesian override proves too much反对贝叶斯推翻的论证证明得过多If Science's distrust of individual rationality means we should never let personal probabilistic reasoning override experimental consensus, then the argument applies equally to rejecting well-established scientific consensus using Bayesian reasoning — a conclusion Yudkowsky himself doesn't hold, as his defense of Many-Worlds shows.如果科学对个体理性的不信任意味着我们永远不应该让个人概率推理凌驾于实验共识之上，那么这个论证同样适用于拒绝用贝叶斯推理挑战已确立的科学共识——而这是 Yudkowsky 本人并不持有的结论，他对多世界诠释的辩护就是证明。
- Libertarianism's historical track record is contested自由意志主义的历史记录存在争议The essay asserts that top-down regulation reliably produces Revolutionary France or Soviet Russia, but market failures, monopolies, and externalities are also historically well-documented. The empirical basis of the libertarian argument is far more contested than the paragraph implies.文章断言自上而下的管制可靠地产出大革命时的法国或苏联，但市场失灵、垄断和外部性在历史上也有充分记录。自由意志主义论证的经验基础远比那段话所暗示的更具争议。
- No account of when personal Bayesian reasoning IS warranted未说明何时个人贝叶斯推理才是正当的The essay warns against overriding Science with personal probabilistic reasoning but gives no criteria for when such override is legitimate versus hubristic. This leaves the reader without guidance for the hard cases — which are the ones that actually matter.文章警告不要用个人概率推理凌驾于科学之上，但没有给出什么时候这种凌驾是正当的而非傲慢的标准。这让读者在困难案例面前没有指导——而那些案例才是真正重要的。
The essay is at its strongest as a corrective to Bayesian overconfidence: the reminder that Science is built on distrust of individual rationality — including the reader's — is genuinely important and often ignored in rationalist circles. It is weakest in the libertarianism parallel, which is used more as rhetorical scaffolding than rigorous argument, and in its failure to give any positive account of when personal probabilistic reasoning legitimately can and should override scientific consensus.
这篇文章在纠正贝叶斯过度自信方面最为有力：提醒我们科学建立在对个体理性——包括读者自身——的不信任之上，这一点真正重要，且在理性主义圈子里常被忽视。它最薄弱的地方在于自由意志主义的平行，这更多作为修辞脚手架而非严谨论证被使用；以及它未能给出任何积极说明，说明个人概率推理何时可以且应当正当地凌驾于科学共识之上。
Original Text原文
Scott Aaronson suggests that Many-Worlds and libertarianism are similar in that they are both cases of bullet-swallowing, rather than bullet-dodging:
Libertarianism and MWI are both are grand philosophical theories that start from premises that almost all educated people accept (quantum mechanics in the one case, Econ 101 in the other), and claim to reach conclusions that most educated people reject, or are at least puzzled by (the existence of parallel universes / the desirability of eliminating fire departments).
Now there's an analogy that would never have occurred to me.
I've previously argued that Science rejects Many-Worlds but Bayes accepts it. (Here, "Science" is capitalized because we are talking about the idealized form of Science, not just the actual social process of science.)
It furthermore seems to me that there is a deep analogy between (small-'l') libertarianism and Science:
- Both are based on a pragmatic distrust of reasonable-sounding arguments.
- Both try to build systems that are more trustworthy than the people in them.
- Both accept that people are flawed, and try to harness their flaws to power the system.
The core argument for libertarianism is historically motivated distrust of lovely theories of "How much better society would be, if we just made a rule that said XYZ." If that sort of trick actually worked, then more regulations would correlate to higher economic growth as society moved from local to global optima. But when some person or interest group gets enough power to start doing everything they think is a good idea, history says that what actually happens is Revolutionary France or Soviet Russia.
The plans that in lovely theory should have made everyone happy ever after, don't have the results predicted by reasonable-sounding arguments. And power corrupts, and attracts the corrupt.
So you regulate as little as possible, because you can't trust the lovely theories and you can't trust the people who implement them.
You don't shake your finger at people for being selfish. You try to build an efficient system of production out of selfish participants, by requiring transactions to be voluntary. So people are forced to play positive-sum games, because that's how they get the other party to sign the contract. With violence restrained and contracts enforced, individual selfishness can power a globally productive system.
Of course none of this works quite so well in practice as in theory, and I'm not going to go into market failures, commons problems, etc. The core argument for libertarianism is not that libertarianism would work in a perfect world, but that it degrades gracefully into real life. Or rather, degrades less awkwardly than any other known economic principle. (People who see Libertarianism as the perfect solution for perfect people, strike me as kinda missing the point of the "pragmatic distrust" thing.)
Science first came to know itself as a rebellion against trusting the word of Aristotle. If the people of that revolution had merely said, "Let us trust ourselves, not Aristotle!" they would have flashed and faded like the French Revolution.
But the Scientific Revolution lasted because—like the American Revolution—the architects propounded a stranger philosophy: "Let us trust no one! Not even ourselves!"
In the beginning came the idea that we can't just toss out Aristotle's armchair reasoning and replace it with different armchair reasoning. We need to talk to Nature, and actually listen to what It says in reply. This, itself, was a stroke of genius.
But then came the challenge of implementation. People are stubborn, and may not want to accept the verdict of experiment. Shall we shake a disapproving finger at them, and say "Naughty"?
No; we assume and accept that each individual scientist may be crazily attached to their personal theories. Nor do we assume that anyone can be trained out of this tendency—we don't try to choose Eminent Judges who are supposed to be impartial.
Instead, we try to harness the individual scientist's stubborn desire to prove their personal theory, by saying: "Make a new experimental prediction, and do the experiment. If you're right, and the experiment is replicated, you win." So long as scientists believe this is true, they have a motive to do experiments that can falsify their own theories. Only by accepting the possibility of defeat is it possible to win. And any great claim will require replication; this gives scientists a motive to be honest, on pain of great embarrassment.
And so the stubbornness of individual scientists is harnessed to produce a steady stream of knowledge at the group level. The System is somewhat more trustworthy than its parts.
Libertarianism secretly relies on most individuals being prosocial enough to tip at a restaurant they won't ever visit again. An economy of genuinely selfish human-level agents would implode. Similarly, Science relies on most scientists not committing sins so egregious that they can't rationalize them away.
To the extent that scientists believe they can promote their theories by playing academic politics—or game the statistical methods to potentially win without a chance of losing—or to the extent that nobody bothers to replicate claims—science degrades in effectiveness. But it degrades gracefully, as such things go.
The part where the successful predictions belong to the theory and theorists who originally made them, and cannot just be stolen by a theory that comes along later—without a novel experimental prediction—is an important feature of this social process.
The final upshot is that Science is not easily reconciled with probability theory. If you do a probability-theoretic calculation correctly, you're going to get the rational answer. Science doesn't trust your rationality, and it doesn't rely on your ability to use probability theory as the arbiter of truth. It wants you to set up a definitive experiment.
Regarding Science as a mere approximation to some probability-theoretic ideal of rationality... would certainly seem to be rational. There seems to be an extremely reasonable-sounding argument that Bayes's Theorem is the hidden structure that explains why Science works. But to subordinate Science to the grand schema of Bayesianism, and let Bayesianism come in and override Science's verdict when that seems appropriate, is not a trivial step!
Science is built around the assumption that you're too stupid and self-deceiving to just use Solomonoff induction. After all, if it was that simple, we wouldn't need a social process of science... right?
So, are you going to believe in faster-than-light quantum "collapse" fairies after all? Or do you think you're smarter than that?
斯科特·阿伦森认为，多世界诠释和自由意志主义有相似之处：两者都是「吞子弹」而非「躲子弹」的案例：
自由意志主义和多世界诠释都是宏大的哲学理论，它们从几乎所有受过教育的人都接受的前提出发（前者是量子力学，后者是经济学入门），并声称达到大多数受过教育的人所拒绝、或至少感到困惑的结论（平行宇宙的存在 / 废除消防队的可取性）。
呵，这个类比真是我永远想不到的。
我此前论证过，科学拒绝多世界诠释，但贝叶斯接受它。（这里「科学」大写，因为我们谈的是理想化形式的科学，而非仅仅是科学的实际社会过程。）
此外，在我看来，（小写的）自由意志主义和科学之间存在一个深刻的类比：
- 两者都基于对听起来合理的论证的实用主义不信任。
- 两者都试图建立比其中的人更可信赖的系统。
- 两者都接受人是有缺陷的，并试图驾驭这些缺陷来驱动系统。
自由意志主义的核心论证在历史上有其动机，即对「如果我们只是制定一条规定 XYZ，社会会变得多么美好」这类美妙理论的不信任。如果这类把戏真的奏效，那么随着社会从局部最优走向全局最优，更多法规应该与更高经济增长相关。但当某个人或利益集团获得足够权力，开始做所有他们认为是好主意的事情时，历史告诉我们实际发生的是大革命时的法国或苏联俄国。
那些在美好理论中本应让所有人从此幸福的计划，并没有产生听起来合理的论证所预测的结果。而权力会腐蚀人，并吸引腐败者。
所以你尽量减少管制，因为你无法信任美好的理论，也无法信任实施它们的人。
你不会因为人们自私而摇手指责怪他们。你试图用自私的参与者建立一个高效的生产系统，方法是要求交易必须是自愿的。这样一来，人们被迫进行正和博弈，因为这才是让对方签合同的方式。在暴力受到约束、合同得到执行的情况下，个体的自私可以驱动一个全球性的高效系统。
当然，这在实践中效果没有理论上那么好，我不打算深入讨论市场失灵、公地问题等。自由意志主义的核心论证不是说自由意志主义在完美世界里会奏效，而是说它能优雅地退化到现实生活中。或者说，比任何其他已知经济原则退化得都不那么尴尬。（那些把自由意志主义视为完美人群的完美解决方案的人，在我看来，有点错过了「实用主义不信任」这件事的重点。）
科学最初认识自身，是作为对信任亚里士多德权威的反叛。如果那场革命的人们只是说：「让我们信任我们自己，不要信任亚里士多德！」他们早就像法国大革命一样昙花一现了。
但科学革命得以延续，是因为——像美国革命一样——其设计者提出了一种更奇特的哲学：「让我们不信任任何人！甚至不信任我们自己！」
一开始，这个想法是：我们不能只是扔掉亚里士多德的扶手椅推理，然后用不同的扶手椅推理取而代之。我们需要与自然交谈，并真正倾听它的回答。这本身就是一个天才之举。
但随后来了实施的挑战。人是固执的，可能不愿接受实验的裁决。我们该摇动一根令人不满的手指责怪他们，说「不乖」吗？
不；我们假设并接受每个科学家可能都会疯狂地执着于自己的个人理论。我们也不假设任何人可以被训练脱离这种倾向——我们不试图选出应该保持公正的尊贵评判者。
相反，我们试图驾驭科学家证明其个人理论的固执欲望，方法是说：「做出新的实验预测，然后做实验。如果你是对的，且实验被重复验证，你就赢了。」只要科学家相信这是真的，他们就有动机做能证伪自己理论的实验。只有接受失败的可能，才能赢得胜利。而任何重大主张都需要重复验证；这给了科学家保持诚实的动机，否则将面临巨大的尴尬。
于是，个别科学家的固执在群体层面被驾驭，产出了稳定的知识流。这个系统在某种程度上比其组成部分更可信赖。
自由意志主义在暗地里依赖大多数个体有足够的亲社会性，会在一家他们再也不会光顾的餐厅留小费。一个由真正自私的人类级别行为者组成的经济体将会崩溃。同样，科学依赖大多数科学家不犯大到无法合理化的罪过。
在科学家相信他们可以通过玩弄学术政治来推广自己理论的程度上——或者操纵统计方法，潜在地赢而没有失败的可能——或者在没有人费心重复验证主张的程度上——科学的有效性就会降低。但就此类事情而言，它退化得很优雅。
成功的预测属于最初做出它们的理论和理论家，不能被一个后来出现的理论仅仅——没有新的实验预测——就窃取——这是这个社会过程的一个重要特征。
最终的结论是，科学不容易与概率论调和。如果你正确地进行概率论计算，你将得到理性的答案。科学不信任你的理性，它不依赖你使用概率论作为真理仲裁者的能力。它要求你设立一个决定性实验。
将科学视为某种概率论理性理想的近似……当然看起来是理性的。似乎有一个听起来极其合理的论证，即贝叶斯定理是解释科学为何有效的隐藏结构。但将科学从属于贝叶斯主义的宏大图式，让贝叶斯主义在看起来合适的时候进来推翻科学的裁决，这并不是一个微不足道的步骤！
科学建立在这样的假设之上：你太愚蠢、太自欺，无法直接使用索洛莫诺夫归纳。毕竟，如果那么简单，我们就不需要科学的社会过程了……对吧？
那么，你究竟是要相信比光速更快的量子「坍缩」仙子？还是你认为自己比那更聪明？