第一章：那杯被思维“拿”到的可乐 | Chapter 1: The Cola “Grabbed” by Thought

2019年10月的明尼苏达州，阳光透过康复中心的落地窗，落在金属质感的机械臂上，也落在18岁少年埃里克·怀特（Eric Whitaker）愣住的脸上。三年前的那场潜水事故让他的颈椎骨折，脖子以下的肢体完全失去知觉，医生曾断言他这辈子都不可能再自己拿起一杯饮料——直到此刻，他看着机械臂稳稳夹住桌上那罐冒着冷气的冰可乐，慢慢送到自己唇边。

October 2019, sunlight streamed through the floor-to-ceiling windows of a rehabilitation center in Minnesota, casting light on the metallic robotic arm and the stunned face of 18-year-old Eric Whitaker. Three years prior, a scuba diving accident had shattered his cervical vertebrae, leaving all limbs below his neck completely paralyzed. Doctors had once declared he would never be able to hold a drink on his own again — until this moment, when he watched the robotic arm firmly clamp the ice-cold can of cola on the table, slowly bringing it to his lips.

出事之前，埃里克是校游泳队的队员，最爱在打完训练赛之后灌下一罐冰可乐，气泡冲进喉咙的爽快感是他最怀念的日常。瘫痪之后，所有动作都要靠护工帮忙，连喝口水都要等别人喂到嘴边，他曾经在日记里写：“我觉得我的身体成了一个监狱，连喝可乐的愿望都成了奢望。”2019年夏天，他报名参加了匹兹堡大学的脑机接口临床试验， researchers 把一个只有指甲盖大的电极芯片植入他的运动皮层——那个负责控制肢体运动的大脑区域，芯片上连着100根比头发丝还细的微电极，能直接捕捉神经元发出的电信号。

Before the accident, Eric was a member of his school’s swim team, and his favorite post-practice ritual was chugging an ice-cold can of cola, the fizz rushing down his throat the most refreshing part of his day. After being paralyzed, every small movement required a caregiver’s help, even drinking water had to be fed to his mouth by someone else. He once wrote in his diary: “I feel like my body is a prison, even the simple wish to drink cola feels like a luxury.” In the summer of 2019, he signed up for a brain-computer interface (BCI) clinical trial run by the University of Pittsburgh. Researchers implanted a tiny electrode chip the size of a fingernail into his motor cortex — the brain region responsible for controlling limb movement, with 100 micro-electrodes thinner than a human hair attached to it, capable of directly capturing electrical signals sent out by neurons.

训练了六个星期之后， researchers 问他最想做什么，他脱口而出：“我想自己喝一罐冰可乐。”那天实验现场围了很多医生、记者和志愿者，所有人都盯着那罐可乐，和连接着机械臂的电脑屏幕。埃里克盯着可乐，在心里一遍遍重复：“我要拿到那罐可乐，我要拿到那罐可乐。”几秒钟之后，屏幕上的神经信号曲线突然波动起来，机械臂动了——它慢慢向前伸，稳稳夹住易拉罐，没有碰倒旁边的水杯，然后精准地送到埃里克嘴边。

After six weeks of training, researchers asked him what he wanted to do most, and he blurted out without hesitation: “I want to drink an ice-cold can of cola on my own.” That day, the experimental site was surrounded by doctors, journalists and volunteers, all staring at the can of cola and the computer screen connected to the robotic arm. Eric stared at the cola, repeating over and over in his mind: “I want to grab that can of cola, I want to grab that can of cola.” A few seconds later, the neural signal curve on the screen suddenly fluctuated, and the robotic arm moved — it slowly stretched forward, firmly clamped the can without knocking over the water cup next to it, then precisely delivered it to Eric’s lips.

当气泡第一次碰到他的舌尖时，这个18岁的少年眼泪一下子就掉了下来。他对周围哭着说：“我三年没喝到冰可乐了，我以为我这辈子都没机会了。”这段视频被发到网上之后，几天之内播放量就破了亿，无数网友在评论区留言：“这就是科技的意义吧，帮那些被困在身体里的人，拿到他们想要的东西。”

When the fizz touched his tongue for the first time, the 18-year-old boy burst into tears immediately. He cried to the people around him: “I haven’t had ice cola for three years, I thought I’d never have the chance again.” When this video was posted online, it got over 100 million views in just a few days, and countless netizens left comments under the post: “This is the meaning of technology, right? Helping people trapped in their own bodies get what they want.”

你可能会问：埃里克到底是怎么做到的？他脑子里想“拿可乐”，怎么就能让几米外的机械臂动起来？答案就是本章要讲的核心：脑机接口（Brain-Computer Interface, BCI）。

You might be wondering: How exactly did Eric do it? He thought “grab the cola” in his head, so how did that make the robotic arm a few meters away move? The answer is the core topic of this chapter: Brain-Computer Interface (BCI).

什么是脑机接口？你的大脑和机器的“翻译官”

要理解脑机接口，我们首先要认识我们的大脑：你的 skull 里装着约1400克重的“超级计算机”，也就是大脑，它里面有约860亿个叫做神经元（neuron）的“微型信号兵”。你产生的每一个想法、做的每一个动作，本质都是这些神经元在“发电报”——它们通过微弱的电信号互相传递信息，比如你想抬手，大脑运动皮层的神经元就会集体发出“抬手”的电信号，沿着脊髓传到你的手臂肌肉，你的手就抬起来了。

To understand BCI, we first need to get to know your brain: inside your skull sits a “supercomputer” weighing about 1,400 grams, which is your brain. It contains about 86 billion “micro signal soldiers” called neurons. Every thought you have and every action you take is essentially these neurons “sending telegrams” — they pass information to each other via weak electrical signals. For example, when you want to raise your hand, neurons in the brain’s motor cortex will collectively send out an “raise hand” electrical signal, which travels along the spinal cord to your arm muscles, and your hand lifts up.

但如果你的脊髓受伤了，或者运动神经元坏死，这些“电报”就发不出去了，你的大脑想抬手，信号却“堵”在大脑里，没法变成实际的动作——这就是埃里克的困境，也是全球超过5000万瘫痪病人的共同困境。而脑机接口，就是专门帮你把这些“堵”在大脑里的信号“接”出来，翻译给机器听的“翻译官”。

But if your spinal cord is injured or your motor neurons die, these “telegrams” can’t be sent out. Your brain wants to raise your hand, but the signal gets “stuck” in the brain and can’t turn into actual movement — this is Eric’s dilemma, and the common dilemma of more than 50 million paralyzed people around the world. BCI is a “translator” specially designed to “pick up” these “stuck” signals from your brain and translate them for machines to understand.

你可以把脑机接口的工作过程想象成你和外国人交流：你的大脑是“说话的人”，机器是“听不懂中文的外国人”，脑机接口就是中间的“翻译器”。你不需要学机器的语言，也不需要机器学你的语言，只要你想，翻译器就会把你的想法变成机器能懂的指令，机器执行完之后，还能把结果反馈给你，形成一个完整的闭环。比如埃里克想“拿可乐”，大脑发出电信号，脑机接口把信号翻译成“机械臂向右移动10厘米，夹住易拉罐”的指令，机械臂执行完之后，视觉信号传回埃里克的大脑，他知道自己成功了，就形成了完整的反馈。

You can think of how BCI works as if you are communicating with a foreigner: your brain is the “speaker”, the machine is a “foreigner who doesn’t understand Chinese”, and BCI is the “translator” in between. You don’t need to learn the machine’s language, nor does the machine need to learn yours. As long as you think, the translator will turn your thoughts into instructions the machine can understand, and after the machine executes the task, it can send the result back to you, forming a complete closed loop. For example, when Eric thought “grab the cola”, his brain sent out electrical signals, BCI translated the signals into the instruction “move the robotic arm 10 centimeters to the right, clamp the can”, after the robotic arm executed the task, the visual signal was sent back to Eric’s brain, and he knew he succeeded, forming a complete feedback loop.

现在的脑机接口主要分成两大类，就像两种不同的“翻译官”，各有优缺点： Current BCIs are mainly divided into two categories, just like two different “translators”, each with its own pros and cons:

1. 硬核翻译官：侵入式脑机接口（Invasive BCI）

这种脑机接口需要做一次微创脑外科手术，把电极阵列直接植入你的大脑皮层，和神经元“挨家挨户”对话，所以信号特别清楚，能读懂非常精细的想法。埃里克用的就是这种，植入他大脑的犹他阵列（Utah Array）只有指甲盖大，上面有100根微电极，能精准捕捉他运动皮层的每一个电信号，甚至能让他控制机械臂捏起一颗花生、转动一个瓶盖。 This type of BCI requires a minimally invasive neurosurgery to implant an electrode array directly into your cerebral cortex, “talking to each neuron door to door”, so the signal is very clear, and it can understand very fine-grained thoughts. Eric used this type. The Utah Array implanted in his brain is only the size of a fingernail, with 100 micro-electrodes on it, which can accurately capture every electrical signal from his motor cortex, even allowing him to control the robotic arm to pinch a peanut or twist a bottle cap. > 📝 小科学家笔记（纸书侧边栏） | 📝 Little Scientist’s Notebook (Print Book Sidebar) > 侵入式脑机接口的优缺点 | Pros and Cons of Invasive BCI > ✅ 优点：信号精度高，能读取精细运动、视觉甚至语言信号，适合重度残疾人恢复功能 > ✅ Pros: High signal accuracy, can read fine motor, visual and even language signals, suitable for severely disabled people to restore function > ❌ 缺点：需要开刀，有感染、排异风险，电极长期植入可能会被大脑的胶质细胞包裹，导致信号衰减，一般只能用5-10年 > ❌ Cons: Requires surgery, with risks of infection and rejection, long-term implanted electrodes may be wrapped by glial cells in the brain, leading to signal attenuation, and can generally only be used for 5-10 years

2. 温和翻译官：非侵入式脑机接口（Non-invasive BCI）

这种脑机接口完全不用开刀，你只需要戴上一个像毛线帽一样的脑电帽（EEG cap），上面贴满柔软的电极，贴在头皮上就能捕捉大脑表面的电信号，安全无创，价格也便宜，几百块就能买到。不过因为信号是从头皮上“隔空”听来的，有点模糊，只能读懂比较粗的想法，比如“我想让光标向左移”“我想让机械臂抬起”，适合普通人日常用。比如现在很多学生用的专注力头环，就是非侵入式脑机接口，能监测你上课时的专注度，走神的时候会轻轻震动提醒你。 Th