夜雨聆风Ten years ago, as the Al revolution was gathering pace, arts and humanities students were told that, if they wanted to make themselves employable, they should "learn to code".
十年前，随着人工智能革命的加速推进，人文艺术专业的学生被告知，如果他们想让自己具备就业能力，就应该“学习编程”。
That may have been bad advice. These days, it is programmers who are nervous about Al taking their jobs. They might consider learning to philosophise.
那可能是个糟糕的建议。如今，担心人工智能抢走自己工作的是程序员。他们或许可以考虑学习哲学。
Earlier this year the Federal Reserve Bank of New York published figures showing that American philosophy graduates are more likely to have jobs than their peers who studied computer science.
今年早些时候，纽约联邦储备银行公布的数据显示，美国哲学专业毕业生比学习计算机科学的同龄人更有可能找到工作。
In 2024, the most recent year for which numbers are available, 7% of those who had studied computer science were unemployed, against just 5.1% of philosophers.
在有数据可查的最近一年2024年，计算机科学专业毕业生的失业率为7%,而哲学专业毕业生的失业率仅为5.1%。
Many are being snapped up by Al firms themselves. Students get job offers before they have graduated, says Luciano Floridi, a philosopher at Yale University.
许多人才正被人工智能公司抢购一空。耶鲁大学的哲学家卢恰诺·弗洛里迪说，学生们在毕业前就收到了工作邀请。
Academics are moving, too. Dr Floridi describes the scale of departures from philosophy departments as a "haemorrhaging".
学术界也在发生变动。弗洛里迪博士将哲学系人员的离职规模形容为“大出血”。
Some of the lessons that philosophy can offer Al researchers are ancient.
哲学能为人工智能研究人员提供的一些经验教训是古老的。
The Socratic method-as described by Plato, an ancient Greek philosopher-uses feigned ignorance and sequential questioning to clarify meanings, spot contradictions and reveal ramifications.
正如古希腊哲学家柏拉图所描述的，苏格拉底教学法运用佯装无知和连续提问来澄清含义、 发现矛盾并揭示后果。
Many current Al systems tend towards sycophancy.
许多当前的人工智能系统往往趋于讨好逢迎。
Models trained in the Socratic method, says Jorg Noller, an expert on philosophy and Al at Ludwig Maximilian University of Munich, are less keen on people-pleasing and more willing to pursue the truth.
路德维希-马克西米利安-慕尼黑大学哲学与人工智能专家约尔格·诺勒表示，采用苏格拉底教学法训练的模型不太热衷于取悦他人，而更愿意追求真理。
Then there is the idea of "Socratic ignorance". In the "Apology", Plato has Socrates claim that his wisdom consists mostly of being aware of how much he does not know.
然后就是“苏格拉底式无知”这个概念。在《苏格拉底的申辩》中，柏拉图笔下的苏格拉底声称，他的智慧主要在于意识到自己有多么无知。
Implanting that humility into a model can help limit overconfidence, a common flaw that Dr Noller describes as "AI immaturity".
将这种谦逊植入模型有助于限制过度自信，这是诺勒博士描述为“人工智能不成熟”的一个常见缺陷。
lason Gabriel, a senior philosopher at Google DeepMind, an Al lab based in London, attributes an industry-wide decline in hallucinations to such efforts.
伦敦人工智能实验室谷歌DeepMind的资深哲学家伊森·加布里埃尔将整个行业幻觉现象的减少归功于这些努力。
More broadly, he says, philosophy lessons are "a powerful mechanism" for improving long Al reasoning processes known as "chains of thought".
他说，更广泛地讲，哲学课程是提升被称为“思维链”的长期人工智能推理过程的“强大机制”。
Philosophical training can also affect a model's outlook in more specific ways.
哲学训练也能以更具体的方式影响一个模型的观点。
Feed an Al legal assistant the writings of John Locke, says Thomas Powers, a philosopher of technology at the University of Delaware, and it will favour robust property rights as an underpinning of political liberty.
特拉华大学的技术哲学家托马斯·鲍尔斯表示， 给人工智能法律助手输入约翰·洛克的著作，它会支持将强有力的财产权作为政治自由的基础。
And if you don't like those principles, the model-makers have others.
如果你不喜欢那些原则，模型构建者还有其他的原则。
The "Granite" series of models from IBM, an American computing giant, come with dials that let business customers better align outputs with their own corporate philosophies.
美国计算机巨头IBM的“Granite"系列模型配备了刻度盘，能让企业客户更好地使输出结果符合其自身的企业理。
Francesca Rossi, IBM's head of responsible Al, says these can let users choose where to strike the balance between philosophical trade-offs, such individual agency versus social harmony.
IBM负责任人工智能负责人弗朗西斯卡·罗西表示，这些技术可以让用户选择在哲学层面的权衡中找到平衡点，比如个人能动性与社会和谐之间的平衡。
Philosophy can help with safety, too. Researchers have documented all sorts of ominous behaviour in Al models, including attempts to evade oversight and even blackmail their users.
哲学也有助于保障安全。研究人员已经记录了人工智能模型中各种不祥的行为，包括试图逃避监管，甚至勒索其用户。
One way model-makers try to discourage this sort of misbehaviour is called AI constitutionalism.
模型制作人员试图阻止这类不当行为的一种方法被称为人工智能立宪主义。
This involves building a model around a scaffolding of rules and principles culled from philosophical writings with legal or moral authority.
这涉及围绕从具有法律或道德权威性的哲学著作中挑选出来的规则和原则框架构建一个模型。
Anthropic, an Al lab based in San Francisco, is one proponent.
总部位于旧金山的人工智能实验室Anthropic就是支持者之一。
Constitutions for its Claude models have incorporated material from sources as diverse as Immanuel Kant, Apple's terms of service and the Universal Declaration of Human Rights.
其Claude模型的宪法纳入了来自不同来源的内容，包括伊曼努尔·康德的思想、苹果公司的服务条款以及《世界人权宣言》。
