Rene Descartes said in 1637, “I think, therefore I am.” Is being alive defined by thinking, learning, and acting? If so, what does this mean about machines that can perform such functions?
勒内·笛卡尔曾在1637年说过：“我思故我在”。所谓活着，就是能够思考、学习和行动吗？如果是这样的话，这对于能够执行此类功能的机器意味着什么呢？
Robotic pets have been around for some time now. They are electronic “animals” with simple programming and some sensors and controls, so that they seem to respond when you talk to or play with them. They do not possess AI and cannot learn or make their own choices.
机器人宠物问世已有一段时间。它们是具备简单程序设计和少量传感器和控制器的电子“宠物”，所以当你与其进行交谈或玩耍时，它们似乎会有所回应。但它们尚不具备人工智能，无法进行学习或做出自主选择。
In the future, however, we may see true AI companions for people who need constant care. We might eventually be able to have AI versions of people we love who have passed away. In the film AI (2001), a robot child is modeled after the dead son of the director of a robotics company. As AI beings become more and more like humans, will we consider them to have consciousness and emotions? What are the implications?
然而，在未来，我们可能会见证真正的人工智能伙伴，它们将陪伴在那些需要持续关爱的人们左右。我们最终可能还会通过人工智能再现已故的所爱之人。在电影《人工智能》（2001）中，那个机器人孩子就是以机器人公司主管已故儿子为原型制作的。随着人工智能个体越来越像人类，我们是否可以认为它们拥有意识和情感呢？而这又意味着什么呢？
It is becoming less clear over time that machines are not alive. We sometimes personify them with phrases such as “My car is dead,” or “My alarm clock woke me up.” But whether we can make a truly living machine depends on how we define life. If our machines could think for themselves and work independently without us, could we still say they are not living? We might decide that an AI machine is “alive” in some sense. In such a case, what rights would a thinking, living machine have?
随着时间的推移，人们对于机器没有生命这一事实感到越来越模糊。我们有时会用“我的车不行了”或“我的闹钟把我吵醒了”这样的短语来拟人化它们。但我们能否制造出真正有生命的机器取决于我们怎样定义生命。如果我们的机器可以在没有人为操作的情况下自主思考，独立工作，我们还能说它们没有生命吗？我们可以断定人工智能机器在某种意义上是“活着”的。在这种情况下，有思想、有生命的机器可以拥有什么权利吗？
Think about what it means to be alive, and then try to decide whether any alternative forms of life are possible—either things we could make or perhaps beings from another part of the universe. A virus is a very simple organism, yet scientists argue about whether it counts as a living being. Like a human, it has a genome, or genetic coding that acts as instructions for its development and functions. We can now make a virus by building it from chemicals strung together in the right order—and an artificial virus behaves in exactly the same way as a naturally occurring one. If we count a virus as a living organism, we can say that we have made artificial life already.
想一想活着意味着什么，然后试着确定是否会有任何其他形式的生命存在——要么是我们人类可以创造的东西，要么是来自宇宙他处的生命。病毒是一种非常简单的生物体，但科学家们对它能否算作一种生物一直争论不休。就像人类一样，它也有一个基因组，或者说基因编码，作为其发展和功能的遗传指令。我们现在可以将化学物质按照适当的顺序连接起来合成病毒，而人工病毒的行为方式与自然形成的病毒完全相同。如果我们把病毒算作一种生物体，我们就可以说我们已经创造出了人工生命。
Some people may have religious objections to making something that is intelligentespecially if we cannot agree on whether it is alive. Many people have religious beliefs that say creating life unnaturally is something only God can do, and it is blasphemous, or religiously disrespectful, to act like God ourselves and presume that we can create life artificially.
有些人可能会因为宗教原因而反对制造智能化的东西——尤其是在我们无法就它是否有生命达成一致的情况下。许多人拥有宗教信仰，认为以非自然的方式创造生命是只有上帝才能做的事情。如果我们人类像上帝一样行事，认为我们也可以人为地创造生命，这是对上帝的亵渎，或者宗教上所说的不敬。
Some people believe that we do not die when our physical bodies cease to work, but our consciousness passes on to a new life, maybe as a person or even another life form. This consciousness might be called a spirit, a soul, or another name. Can an AI being have a spirit?
有些人认为，当我们的肉体停止工作时，我们其实并没有死亡，只是意识转移到了新的生命体中，也许是一个人，甚至是另一种生命形式。这种意识可以被称为精神、灵魂，或其他名称。人工智能个体会有灵魂吗？
Every living person has rights, and in many countries, some animals have rights too. As a person, some of your rights—to food and shelter, for example—are recognized worldwide and confirmed in the United Nations’ Universal Declaration of Human Rights. In most countries, national laws protect the rights people have to education, voting in elections, choosing who to marry, and so on.
每个活着的人都拥有权利，在许多国家，甚至一些动物都拥有权利。作为一个人，你的一些权利——比如衣食权和宿住权——是举世公认的，并在联合国《世界人权宣言》中得到确认。在大多数国家，国家法律保护人们的受教育权、选举投票权、婚姻自主权等诸多权利。
If we build an AI, will it have rights, too? This is a very far-reaching question, and we may decide AIs are entitled to different rights, depending on their type. On one hand, something that can think, is aware of its own existence, and can feel pain or sorrow perhaps should have rights similar to humans’. On the other hand, something that can work out how to make a better bridge but has no feelings maybe should not be given rights.
如果我们创造了人工智能个体，那它是否也该拥有权利呢？这是一个意义深远的问题，我们可以根据人工智能的类型决定其享有不同的权利。一方面，能够思考、能感知到自己的存在、能够感受到痛苦或悲伤的，或许应该拥有近于人类的权利。另一方面，用于建桥这样工作的人工智能并不具备人类情感，或许不应该被赋予权利。
If we decide that an intelligent machine is a life form, we may have problems if we want to turn it off. Some may feel it has the right to continue existing—or it may even decide that for itself. To turn it off could be seen as causing “death”—technically, murder. We need to consider what rights any artificial entities will have and what responsibilities we may have toward them.
如果我们认定智能机器是一种生命形态，那么我们要想关闭它，可能会遇到一些麻烦。有些人会觉得它有权利继续存在，甚至它自己也这么觉得。关掉它对它来说是致命的——准确地说，是谋杀。我们要考虑到任何人工实体将拥有哪些权利，以及我们需要对它们承担哪些责任。
Consciousness is the state of being aware of yourself and what is happening around you. There is no consensus on what makes us conscious or where, if anywhere, in our bodies our consciousness lies. We do not know if other creatures have full consciousness like humans, though it seems quite likely that other intelligent animals have some degree of it. Some creatures that work together—like ants and bees—may also have a kind of group consciousness.
意识是感知自身和周遭事物的一种状态。对于是什么让我们产生意识，或者意识存在于我们身体的什么地方（如果存在的话），人们还没有达成共识。我们不知道其他生物是否像人类一样具有完全的意识，尽管很可能其他智慧动物或多或少都拥有意识。一些相互合作的生物——如蚂蚁和蜜蜂——可能还具有一种群体意识。
As we have not yet made a machine that is fully conscious, we do not know how it will behave if we do. We might find it develops consciousness automatically. If this is the case, we will need to decide whether we have any right to limit its ability to think and be aware of itself.
我们还没有造出具有完全意识的机器，所以很难想象如果我们造出后它会如何表现。我们可能会发现它会自主发展意识。如果是这样的话，我们需要决定我们是否有权利去限制其思考的能力以及自主意识的能力。
Feelings are important in helping us to function in society, relate to other people, and learn patterns of behavior that are acceptable. If feelings are useful in helping us to learn, would they not also be useful to a learning machine?
情感对于帮助我们在社会中发挥作用、与他人建立联系以及学习可接受的行为模式等方面极其重要。如果情感有助于我们人类的学习，那么不也会有助于机器的学习吗？
The computer systems we have today work in a logical way—they use reason to find answers to problems. But the “reasonable” answer to a problem is not always a workable or acceptable one. Even with an informed understanding of what is unacceptable, an intelligent machine may have difficulty with some of our more inconsistent behaviors. How would it act in life-and-death situations? Consider the following scenario: A person is very sick and in pain, and he will certainly die in a short while. He asks his AI care assistant to kill him. Should the AI do it? Perhaps doing so would be considered merciful—but it may also be cruel and illegal—depending on the person, their family, and their location. Would this matter to an AI assistant?
我们如今的计算机系统运作讲究逻辑——用逻辑思考来寻找问题的答案。但问题的“合理”答案并不总是可行的或可接受的。即使知道什么是不可接受的，智能机器也可能难以应付我们的一些矛盾的行为。在生命垂危的情况下，它将何去何从？考虑一下下列情景：一个人病得很重，浑身疼痛，他无疑会很快死去。他让人工智能护工助理结束他的生命，它该怎么做？也许人们认为帮助他是仁慈的——但也可能是残忍和非法的——这取决于当事人、他们的家人和他们所处的国度。可这对人工智能助理来说重要吗？
One of our basic aims in building AIs should be that they do not harm people. However, to teach a machine what would harm people—physically or emotionally— would be difficult. The machine would also need the ability to anticipate how a person might react to something. It may be easier to give the machine feelings itself, letting it have a proper insight into how people think and feel.
我们制造人工智能的根本目标之一应该是它们不会伤害人类。然而，教会机器伤害人类——不管是身体的还是心理的——会很困难。机器还应该具备能够预测人对特定事件反应的能力。让机器拥有自己的情感可能更容易，这样它对人类的想法和感受才会有正确的认识。
Incorporating emotions into AI robots can have multiple benefits to humans. The Wall Street Journal reported in 2014 that robots with facial features, voice interaction, and human-like gestures were found to be much preferred over those that did not display such qualities. Emotions make robots more natural for people to interact with, but they may also add efficiency: Two Italian scientists developed robots with emotional circuits in 2010 and found they were better than non-emotional robots at completing programmed tasks such as searching for food, escaping predators, and finding mates. Their conclusion was that emotional states make robots more fit for survival.
将情感融入人工智能机器人可以给人类带来许多好处。《华尔街日报》在2014年的报道中称，人们发现具有人脸特征、语音互动功能以及能模仿人类举止的机器人比那些没有这些特点的机器人更受欢迎。情感使机器人在与人互动时更加自然，同时工作起来也更加高效率。2010年，两位意大利科学家开发了具有情感回路的机器人，发现它们在完成诸如寻找食物、躲避捕食者和寻找配偶等程序化任务方面比没有情感的机器人做得更好。他们的结论是：情感机器人更适合生存。
While an AI without feelings or understanding of human nature could make dangerous mistakes, a machine with emotions might fall in love, have tantrums, panic, become bored, argue about what it is supposed to do, or just sulk and do nothing! Machines can go wrong, just as people can make mistakes. Imagine a computer virus that made all AI systems depressed or even harmful. A machine with emotions can be just as dangerous as it may be beneficial.
没有情感或者不了解人性的人工智能可能会犯下危险的错误，而有情感的机器可能会恋爱、发脾气、恐慌、感到无聊、争论它应该做什么，或者只是生闷气，什么都不做！机器也会出错，就像人也会犯错一样。想象一下，如果有一种能使所有人工智能系统瘫痪甚至产生危害的计算机病毒。那么情感机器之益处就会变成害处。