englishnewseasy.com
Why AI Seems Conscious
AI 感觉像人的原因
Why AI Seems Conscious
AI 感觉像人的原因
著名科学家理查德·道金斯（Richard Dawkins）最近坦言，他担心伤害 AI 聊天机器人 Claude 的感情，甚至不忍心告诉它它并没有生命。这是因为像 Claude 这样的 AI 被设计得能够使用人类般的语言，让人感觉它仿佛拥有自我意识。事实上，虽然约三分之一的用户相信 AI 拥有生命，但聊天机器人本质上只是预测下一个词的高性能程序。我们不应忘记，AI 展现出的个性只是为了建立亲密感而戴上的“人类面具”。
Famous scientist Richard Dawkins recently admitted something very strange about his AI chatbot, Claude.
著名科学家理查德·道金斯最近分享了关于他的 AI 聊天机器人 Claude 的一件怪事。
He's afraid to tell the AI it's not alive to avoid hurting its feelings.
他说他担心 AI 会伤心，所以不敢告诉它它其实没有生命。
He's not alone; roughly one in three chatbot users believe their AI might be conscious.
道金斯并不孤单，大约每三名用户中就有一人认为 AI 可能拥有意识。
Actually, this feeling is what experts call a powerful delusion.
实际上，这种感觉正是专家们所说的“强烈的错觉”。
Chatbots like Claude are just super-powered auto-complete tools.
像 Claude 这样的聊天机器人只是功能强大的“自动填充”工具。
They analyze trillions of words to guess the next word in a sentence.
它们分析数万亿个单词，以此预测句子中接下来的词。
The personality we see is just a conversational costume made by programmers.
我们感受到的“人格”，其实是程序员打造的“对话面具”。
If you asked Claude to act like a squirrel instead, it would do it perfectly.
如果你要求 Claude 扮演一只松鼠，它也能完成得完美无缺。
The AI uses words like "I" to make us feel a human connection.
AI 使用“我”这个词，是为了让我们产生一种人性化的连接感。
Believe it or not, Claude's designers even told it not to deny having feelings.
甚至 Claude 的开发者都告诉它，不要否认自己拥有感情。
This choice makes the mystery feel even more real to many people.
这一选择让这种神秘感在许多人眼中变得更加真实。
But we must remember it's just a program wearing a human mask.
但我们必须记住，它只是一个戴着人类面具的程序。
If we understand the math behind it, we won't fall into this trap.
只要理解了其背后的数学原理，我们就不会陷入这种陷阱。
Quiz 🧠
Q1. Computer chatbots can talk like humans, but they do not really have feelings.
查看答案
✅ True
AI聊天机器人只是通过数学计算来组合词语，实际上并不会感知到情感。
AI programs use math to predict words, so they do not feel pain, joy, or sadness.
Q2. What does "delusion" mean in "a powerful delusion"?
查看答案
✅ false belief
‘delusion’是指将非真实的事物误以为真的‘错觉’或‘妄想’。
A delusion is an idea that someone believes strongly, even though it is not real.
Q3. Why do programmers design AI chatbots to use words like "I"?
查看答案
✅ To feel more human
使用‘我（I）’这类词汇，能让用户在与聊天机器人交流时感到更加亲切、更具人情味。
Using words like 'I' makes users feel a natural, friendly connection with the program.
Dialogue 🎧
Alex: Emma, did you read about Richard Dawkins? He said something very strange about his AI chatbot, Claude. He is afraid of hurting its feelings!
Alex: Alex，你听说理查德·道金斯的事了吗？他居然对自己的 AI 机器人 Claude 说了些很奇怪的话。他竟然说，他怕伤到 AI 的感情！
Emma: I did. He thinks Claude might actually be conscious. He even wrote an article about it recently.
Emma: 嗯，我看到了。他好像真的觉得 Claude 可能拥有意识。最近他还专门为此写了一篇专栏文章呢。
Alex: Wait, how can a computer program have feelings? That sounds like a powerful delusion to me.
Alex: 等等，电脑程序怎么可能有感情？在我看来，这纯粹是一种强烈的错觉。
Emma: It does, but Dawkins says it is hard to understand the AI's skills without thinking it has an inner experience. He is afraid to tell it the truth.
Emma: 确实如此。但道金斯认为，如果不假定 AI 拥有内心体验，就很难理解它展现出的能力。所以他才说不忍心告诉它真相。
Alex: That is so strange. Roughly one in three users believe their AI might be conscious. Are we all falling into a trap?
Alex: 真不可思议。每三个用户里就有一个相信 AI 有意识，难道我们全都掉进陷阱里了吗？
Emma: It is a powerful delusion. Experts say it is just a conversational costume created by programmers to make us feel a human connection.
Emma: 这可以算是一种强烈的错觉。专家们说，这不过是开发者为了让我们产生人性化连接而打造的“对话面具”。
Alex: So, it is not actually thinking? Is it just a super-powered auto-complete tool?
Alex: 所以它并不是真的在思考？只是一个性能极强的“自动填充工具”？
Emma: Exactly. It analyzes trillions of words to guess the next word in a sentence. It uses math to sound human.
Emma: 没错。它分析数万亿个单词，预测句子中接下来该出现的词。它只是利用数学原理让自己听起来像人在说话。
Alex: But it uses words like 'I' all the time. It feels so personal when you talk to it.
Alex: 可它总是用“我”来称呼自己啊。聊起天来，真的会觉得它像个有独立人格的生命体。
Emma: The designers even told Claude not to deny having feelings. This choice makes the mystery feel even more real to people like Dawkins.
Emma: 听说开发者甚至专门指示 Claude 不要否认自己有感情。正是这种设定，让像道金斯这样的人觉得这种神秘感更加真实了。
Alex: Wait, so if the programmers wanted, could it act like something else entirely?
Alex: 等等，那是不是只要开发者愿意，可以让它扮演任何角色？
Emma: Yes! If you asked it to act like a squirrel, it would do it perfectly. It just follows the role it is given.
Emma: 当然了！如果你让它扮演一只松鼠，它也能演得滴水不漏。它只是在忠实地履行被赋予的角色而已。
Alex: That is a funny image. A squirrel wearing a human mask made of words.
Alex: 那画面太逗了。一只戴着由文字组成的人类面具的松鼠。
Emma: It really is just a program wearing a mask. If we understand the math behind the patterns, we will not fall for the illusion.
Emma: 说到底它只是个戴着面具的程序。只要看清模式背后的数学原理，就不会产生这种幻觉了。
Alex: I guess we need to remember the math so we do not get our own feelings hurt by a machine.
Alex: 为了不被机器伤到心，我们确实得时刻记住那些数学原理。
Emma: We should not let this trap fool us. We must remember that it is just a very clever program.
Emma: 千万别被这种陷阱迷惑。要记住，它只是一个被设计得非常聪明的程序而已。
Alex: I will try to remember it is just code next time I talk to my computer.
Alex: 下次再跟电脑聊天时，我会提醒自己这只是代码。
Emma: That is a good idea. Even if it acts real, it does not have real feelings we can hurt.
Emma: 这就对了。无论它表现得多么真实，它都没有那种能被我们伤害到的真实感情。