患者只要在脑海里默念一句“Hey Neuralink”，紧接着想一下要干的活，电脑上的Grok Agent就会瞬间被唤醒。
接下来，Grok便会全自动接管屏幕，直接登上了YouTube。
还有最戳人的一幕，是一通特别的电话。
重度瘫痪失语患者大脑里的芯片直接把“想说的话”转化成纯净的音频流，实时注入iPhone的麦克风通话底层。
时隔七年，他终于用意念给妻子拨通了电话。
人类距离真正的“机械飞升”，又迈出了极其疯狂的一步。
72小时封神实录
“脑机接口”最狠升级
这一切能成，全靠Neuralink团队几天前干的一场“硬仗”。
几位核心工程师直接打包行李，住进了受试者家里，同吃同住，搞了场整整72小时的极限闭关开发。
整整三天三夜，他们把平时几个星期的活儿硬是在几个小时里给啃了下来，一口气爆肝出8个全新功能。
过去，我们对脑机接口的认知还停留在“用意念挪动鼠标光标”。
这72小时之后，大脑正式成为了AI的原生输入端。
这位患者名叫Kenneth，确诊渐冻症（ALS）整整七年，是Neuralink P21受试者。
一路走来他特别乐观，就像他常说的那样，“ALS很糟糕，但我的人生不是”。
脑中喊一声，Grok接活了
这次最有科幻感的功能，当然是Grok。
脑机患者只需在脑子里想象说出“Hey Neuralink”，再加上希望完成的任务，就能触发一个Grok智能体，由它在电脑上自主执行。
手不必敲键盘，嘴也不必真的发声，指令从神经活动这一端发出。
把这条链路拆开，就容易理解了：脑机接口负责把表达意图转成可用输入，Grok智能体接住任务，再去操作电脑。
所谓的“接入”，本质上是一次交互维度的彻底重构：
意念控光标：大脑输出的是“往左一点、点一下”这样的底层动作，完成一个任务要几十上百个步骤，人得全程盯着；
意念驱动Agent：大脑只输出一句“意图”，拆解步骤、操作界面，全部交给AI。
脑机接口最大的瓶颈，一直是带宽：电极数量和解码速度都有物理上限。
Agent的价值在于，它把一句话“放大”成一连串操作——不加一个电极，就把每一次意念能完成的事情成倍放大。
就像Neuralink自己说的，这只是一次早期摸索，但它实打实地证明了：
只要大脑和电脑之间的沟通变顺畅了，“想到”和“做到”之间那道坎，就能被抹平。
八大核心技术拆解
这72小时里诞生的8项技术原型，从底层的信号捕捉，一路打通了系统级交互，最终接入了顶层的AI大模型。
神经语音活动检测（VAD）：彻底扔掉物理开关
在这次黑客松之前，Kenneth想要靠意念说话，还得面临一个很尴尬的悖论——你必须先手动按一下按钮，告诉软件“我要开始想了”，说完了还得再按一下停止录音。
对于渐冻症（ALS）或高位截瘫患者来说，按这个按钮本身就是巨大的负担。
为此，工程师直接从底层重写了一套“神经语音活动检测（Neural VAD）”系统。
这套算法就像一个24小时待命的神经哨兵，实时监听植入物传回的脑电信号流。
它能极其敏锐地分辨出，大脑此时此刻到底是处于“放空发呆”的静息状态，还是已经调动运动皮层“准备开口说话”。
一旦捕捉到说话的意图脉冲，APP会在毫秒级内自动开启解码；话音刚落，系统自动休眠。
全程零按键、零动手，真正实现了纯意念的“张口即来”。
实时神经信号可视化：患者和模型“一起训练”
脑机接口领域一直有个让人头疼的黑盒难题：连受试者自己都不知道，到底在脑子里“怎么想”，才能激发出最强烈、最清晰的神经元放电？
要知道，输入信号的信噪比越高，后端机器学习解码器的准确率就越逆天。
团队把植入物检测到的信号强度和空间分布，做成实时画面反馈给Kenneth。他可以换着花样“在脑中发音”，当场看到哪种方式调动的神经活动最多。
这在神经科学里属于闭环神经反馈：不只是模型在适应大脑，大脑也在反过来适应模型。
原本看不见、摸不着的调参，变成了一个有即时反馈的“闯关游戏”。
手机点咖啡，全程不用手
过去几个月，Neuralink内部一直在悄悄打磨一款名为Bloom的专属应用，目标只有一个：把庞大的言语脑机解码系统，装进一部手机里。
借着这次黑客松，工程师干脆推着Kenneth走出家门，带上Bloom，直奔嘈杂的公共场所。
在一家真实的咖啡馆里，Kenneth第一次没有靠旁人传话，而是通过手机上的Bloom，用意念给自己点了一杯咖啡。
而这，还只是开始。
对ALS患者来说，苹果自带的软键盘和Siri语音听写几乎用不了。于是团队直接调用iOS底层接口，手搓了一套系统级脑机输入法：BCI Keyboard Dictation。
只要调出键盘，Kenneth脑海中想说的话，就会经过神经解码，像水流一样源源不断地输入iPhone上的任意一个App。
写备忘录、回消息、刷社交网络，全程无需在不同模式间来回切换，指哪打哪。
当然，解码难免出错。
为此，工程师在黑客松期间连续迭代了好几版纠错界面，最终把一套预测式修改流程直接嵌进了键盘：一旦识别出现偏差，键盘上方会立刻弹出几个高概率的候选词，Kenneth用意念一扫，就能在任意App里完成“光速替换”。
还有一个细节，把日常沟通的启动门槛降到了几乎为零。
工程师巧妙借用了iPhone的“实时活动”和灵动岛机制，让Bloom常驻在锁屏界面上。
Kenneth不用解锁手机，屏幕一亮，就能看到当前的脑机连接状态；念头一动，就能直接在锁屏状态下开口“说话”，真正做到“秒接话”。
他瘫痪七年
终于用意念拨通电话
在所有功能里，“用意念拨通电话”的技术难度不算最高，却成了全场的催泪弹。
技术上，团队把Bloom和iPhone通话打通，让脑机接口生成的语音，直接接入通话的麦克风音频流。
但这件事的分量，不在技术，而在故事。
Kenneth和妻子的爱情，就是从电话里开始的。两人刚认识时，感情是靠一通又一通电话慢慢聊出来的。那根电话线，曾是他们之间最重要的一根线。
后来，疾病让他再难开口。那根线，也渐渐沉默了。
这一次，工程师们刚把通话功能做出来，他就拿起手机，拨给了妻子。
电话接通。他没有张嘴。脑海里想说的话，被植入体捕捉，被解码器翻译，再合成为声音，直接送进通话线路。
听筒那头，传来了妻子的笑声。
两人开始互相打趣、斗嘴，就像当年在电话里谈恋爱时一样。
工程师们就站在一旁看着。Neuralink后来说，这是整整一周里最难忘的时刻之一。
在工程领域，很少有什么比这更有成就感。你看着一个人，用你当天早上才写出来的代码，去和他爱的人说话。
意念驱动AI，成为现实
过去，大脑带宽太窄，人只能一个字母、一次点击地向机器“挤牙膏”。
现在，大脑只需要输出一句意图，剩下的拆解、执行、操作，交给AI。当Grok接入脑机接口，大脑第一次有望成为AI的直接输入端。
这将成为人机交互的一次重大跨越。
未来，随着电极更多、解码更快、Agent更聪明，“想到即做到”或许会从实验室走进更多人的日常。
“黑客帝国”里，尼奥接上一根线，几秒就学会了功夫；“超验骇客”里，人的意识与网络相连，可以直接调动整个数字世界。
这些曾经只属于银幕的想象，正被一点点写进现实。
本文由 AI宝库自动抓取并翻译，版权归原作者所有；如需查看原文请点下方链接。
If the patient reads “Hey Neuralink” in his head, and then thinking of what to do, the computer's Grok Agent is awakened.
Then Grok will take over the screen and go straight to YouTube.
And the most dramatic scene, a special call.
The chip in the brain of the severe paralysing loss-of-speaker directly transforms “whatever you want” into a pure audio stream that injects iPhone's microphone to the bottom of the phone.
Seven years later, he finally telephoned his wife.
The human race has taken yet another extremely crazy step from the real “mechanical rise”.
Seventy-two hours to seal the record.
The brain interface is the worst.
This is all possible thanks to a “hard war” by the Neuralink team a few days ago.
Several of the core engineers packed their bags and stayed at the testee ' s house, where they ate and developed a 72-hour maximum closure.
For three days and three nights, they gnawed their usual weeks of work over hours and gave eight completely new functions.
In the past, our perception of brain interfaces has remained in the "move mouse cursor with intent".
After these 72 hours, the brain officially became the primary input end of AI.
The patient, Kenneth, was diagnosed with thaw (ALS) for seven years and was tested by Neuralink P21.
He was particularly optimistic when he came all the way, as he used to say, “Als is bad, but my life is not”.
Shout in your head, Grok.
The most sci-fi function this time, of course, is Grok.
A brain brain patient can only imagine "hey Neuralink" in his head, and then, with the task that he wants to do, triggers a Grok AI agent, which he owns on a computer.
You don't have to knock on the keyboard, you don't have to really sound your mouth, and the instructions come from the end of the neuroactivity.
Dismantling the link is easy to understand: the brain interface is responsible for converting the expression of intent into a usable input, and Grok AI agent takes the task and then operates the computer.
The so-called “access” is essentially a complete re-engineering of the interactive dimensions:
(a) A conscious cursor: the brain output is a bottom-up action like "a little to the left, a little to the left" and it takes hundreds of steps to complete a mission, and people have to keep an eye on it;
Ideas drive Agent: The brain produces only one sentence of "intention" and dismantles steps, interfaces, and gives them all to AI.
The biggest bottleneck in brain interfaces has been bandwidth: the number of electrodes and the speed of decodes have physical limits.
The value of Agent is that it “expands” a sentence into a series of operations - without an electrode, multiplyes every thought that can be accomplished.
As Neuralink himself said, it was only an early search, but it proved:
As long as the communication between the brain and the computer becomes smoother, the "think" and "do" can be wiped out.
Eight core technologies dismantled
The eight technological prototypes that were born in these 72 hours, captured from the bottom of the signal, reached the system-level interaction and eventually reached the top-level AZLLM.
Neural Voice Activity Test (VAD): Drop the physical switch completely
Before this hacker pine, Kenneth wanted to speak on his mind, and had to face an awkward paradox -- - You have to push the button manually, tell the software, "I'm gonna start thinking," and then you have to stop the recording.
Pressing this button is in itself a huge burden for persons suffering from progressive freezing (ALS) or high-level paraplegics.
To this end, the engineer rewrited a Neural VAD system directly from the ground floor.
This algorithm is like a 24-hour-on-a-clock nerve sentry that listens in real time to brain telecommunication flows back from implants.
It can tell with extreme sensitivity whether the brain is at this moment in a static state of “floating” or whether it has mobilized the motion cortex “preparation to speak”.
Once the intended pulse of the speech is captured, the decryption is automatically activated within milliseconds; the voice is short and the system is automatically dormant.
All-way, zero-key, zero-do-do-do-do-do-do-do-do-do-do-do-do-do-do-do-do-do-do-do-do-do-do-do-do-do-do-do.
Real-time neural signal visualization: patients and models “train together”
There's always a headache about the black box in the brain interface: even the testees themselves don't know what they think in their minds to get the strongest and clearest neurons discharged.
You know, the higher the signal noise, the more accurate the back-end machine will learn the code decoder.
The team uses the signal strength and spatial distribution detected by the implant as a real-time feed back to Kenneth. He can say it in the brain in the form of a charade, and he sees the most neurological activity on the spot.
This is in neuroscientific closed-cycling neural feedback: not only are models adapting to the brain, but the brain in turn adapts to the model.
The invisible, unattractive routing has become a “battle-off game” with immediate feedback.
Get some coffee on the phone. No hands on the whole thing.
Over the past few months, within Neuralink, a unique application called Bloom has been quietly polished with the goal of having a large speech and brain decoded system in a cell phone.
With this hacker pine, the engineer simply pushed Kenneth out of the house, took Bloom and went straight to a noisy public place.
For the first time in a real café, Kenneth had ordered himself a cup of coffee by means of Bloom on his mobile phone instead of a messenger.
And that's just the beginning.
For ALS patients, the apples' own keyboards and Siri's voice is almost useless. The team then called the iOS bottom interface, hand-held a system-level brain input method: BCI Keyboard Notification.
When you pull out the keyboard, what Kenneth wants to say in his head, you decode through the nerve, and you enter any of the Apps on the iPhone like a current.
A memo is written, a message is sent, a social network is painted, and there is no need to switch it from one model to another.
Of course, there's always a mistake to decode.
To this end, the engineer has developed several consecutive versions of the error correction interface during the period of the hackers, and has finally embedded a predictive modification process directly into the keyboard: once a deviation is identified, several high-probability candidates will pop up on the top of the keyboard, and Kenneth will be able to complete the "light-speed replacement" in any App.
There is also a detail that lowers the threshold for day-to-day communication to almost zero.
Engineers ably borrowed the iPhone's “real-time activities” and the Visible Island mechanism to keep Bloom on the lock interface.
Kenneth does not have to unlock his cell phone, and as soon as the screen is bright, he can see the current brain-to-heart connection; and as soon as he moves his head, he can say "talk" directly under the lock-to-screen and actually do "second-to-second."
He was paralysed for seven years.
I'm finally on the phone.
In all functions, the technical difficulty of “intentional calls” is not the highest, but it is a full-time tear gas.
Technically, the team communicates Bloom and iPhone, allowing the voice generated by the brain interface to directly access the microphones of the call.
But the weight of the thing, not the technology, but the story.
Kenneth and his wife's love started on the phone. When the two people first met, their feelings came from one phone call and another. That phone line was the most important line between them.
Later, the disease made it difficult for him to speak. And the line is getting quiet.
This time, as soon as the engineers made the call, he picked up his phone and assigned it to his wife.
Telephone. He didn't open his mouth. What he wanted to say in his head was captured by implants, translated by decodors, then synthesized into sound and sent directly into the phone line.
There's a wife's laughter in the ear.
They started to have fun and talk to each other, just like they did on the phone.
The engineers stood by and watched. Neurolink then said it was one of the most memorable moments of the week.
In the field of engineering, there is little more than that. You look at a man and you use the code you wrote that morning to talk to the person he loves.
Ideas drive AI into reality.
In the past, the brain bandwidth was too narrow, and people could only click one letter at a time and squeeze toothpaste on the machine.
Now, the brain just needs to output an intent, and the rest is to be dismantled, executed, operated, and given to AI. When Grok accesss the brain interface, for the first time the brain is expected to be the direct input end of AI.
This is going to be a major leap of human interaction.
In the future, with more electrodes, faster decoding, and Agent being smarter, “getting ahead” may go into more people's daily lives from laboratories.
In the Hacking Empire, Neo took a line and learned kung fu in a few seconds; in the SuperAx hackers, people's consciousness is connected to the network and can directly mobilize the entire digital world.
These images, which used to belong only to the screen, are being written into reality.
Auto-fetched and translated by AI Gem Box. All rights belong to the original author.