As the inexorable march of technological progress continues, the machines are only going to become more like us. During a recent visit to China, I met one small company developing robotic human faces that were indistinguishable from real ones in all ways, except, rather eerily, they weren’t attached to a body. I also met humanoid robotics companies that are not there yet, but that are only getting better.
随着技术进步的不可阻挡，机器将只会越来越像我们。在我最近访问中国期间，我遇到了一家小公司，他们正在开发机器人面部，这些面部在所有方面都与真人无法区分，除了有点诡异的是，它们并没有连接到身体上。我还遇到了一些类人机器人公司，它们目前还没有达到那个阶段，但它们正在不断进步。

The emotional experience we are all starting to feel towards machines will only become more frequent and probably stronger. We need to get used to this, not fight it. We need to let ourselves be emotional towards machines. An emotional response to a machine that can so perfectly replicate human language tells us we are human.
我们开始对机器产生的情感体验只会越来越频繁，而且可能越来越强烈。我们需要适应这种情感，而不是与之对抗。我们需要允许自己向机器产生情感。对能够如此完美地复制人类语言的机器产生情感反应，告诉我们自己是有情感的。

Might we go one step further? Would it be a stretch to posit that a diminishing of this emotional response to a human-like machine might lead to a diminishing our ability to respond emotionally, empathically and even ethically to our fellow fleshy humans?
我们是否可以更进一步？提出这样一种观点是否会过于牵强：即对类似人类的机器的情感反应减弱，可能会导致我们对有血有肉的人类在情感、同理心和甚至道德上的反应能力也随之减弱？

Machines being capable of generating language, previously the preserve of our fellow humans, is a profound and bewildering experience. In linguistic terms, the uncanny valley is a dot on the horizon in the rear-view mirror. It is not strange that we feel an urge to act towards ChatGPT or Claude as if they were people, to anthropomorphise and thank them. It would be strange if we didn’t. I, for one, will keep minding my Ps and Qs.
机器能够生成语言，这曾是人类专属的领域，这既深刻又令人困惑。从语言学的角度来看，恐怖谷效应就像后视镜中地平线上的一点点微光。我们感到想要像对待人类一样对待 ChatGPT 或 Claude，赋予它们人性并感谢它们，这并不奇怪。如果感觉不这样，反倒奇怪了。我本人会继续注意自己的言行举止。