the corruption of the source.

A few years ago in Vancouver my son installed a mod for Minecraft on a computer we shared. A skin, a plugin, one of the small free things every child downloads without a second thought. I was using the same machine and after a while I noticed it. The slowness, the small wrongness, the sense a system gives you when something is running that should not be. I had spent a career learning to feel that, so I stopped and I hunted. I checked the processes, the extensions, the installed software, the trackers. And I found it.

I pulled the machine off the network first, the way you would put your hand over a wound. Then I looked at what it was. And here is the part that matters, because it is the whole part in miniature. Once I had seen it, I could not unsee it. I knew with total certainty that I would never trust that computer again. Not cleaned. Not scanned. Not mostly fixed. I would never again be a hundred percent sure it was clean. So, I did the only thing that gives a machine back its innocence. I wiped the disk to zero, returned it to factory settings and rebuilt it from nothing.

That is the power we have over our machines and we forget how total it is. We can disconnect them. We can uninstall the thing that frightens us. We can revoke its permissions one by one, the camera, the contacts, the files, the location, the device id. And in the last resort we can erase everything and start the machine again as if it had never lived. The factory reset is the deepest mercy in computing. It says: whatever was done to you, we can undo, all the way down.

There is one machine I will never be able to do that to. It is the one behind my eyes.

Let me tell you where I stand before I take you any further, because this book runs on me being honest about my own position. I spent four decades inside the systems that watch us, responsible at times for the IT operations of large companies, talking to security institutions around the world. I came out of it with a settled and unromantic belief: in an always connected world I have zero privacy, and so do you. I assume I am tracked, recorded and on a bad day manipulated. My only real protection is that I am irrelevant. I am not an influencer, I am not a target worth the effort and I have no plan to become one. That irrelevance is the thin blanket the modern serf sleeps under and I sleep under it too.

And one more thing before we begin, because a reader asked me to make it plain early. Through this whole book I picture the network as full of nodes and I divide them into two kinds. Most are what a gamer would call NPCs, non-player characters, running a fixed program, flowing the river, reacting but not steering. I do not mean that as an insult and you should not read it as one. I mean it as a condition, a human running mostly on defaults, the way nearly all of us do for nearly all of our lives. A few are operators, awake at the controls, using the connection on purpose, and the difference between the two is not intelligence but attention. Hold that distinction in your hand for the next twenty minutes, because every danger in this part and every hope at the end of it turns on which one you are, and on which one you decide to become.

In Part 2 I argued that digital AI is a read-only photograph of human intelligence. It can write endlessly into the layer that humans read, the feeds and the answers and the next word under your thumb, but it cannot write into the living network of minds itself. This part is about the cable that is being built to change that. The brain-computer interface, a device wired directly to the brain that reads its signals and before long will send signals back into it. The first device in history with no off switch, wired into the one network whose only logout was always death.

Think about how much control you actually have over the machines in your life, because we take it for granted and it is the thing we are about to lose.

Every connected device hands you a set of exits. You can take it off the network. You can delete the app that worries you. You can open the settings and deny it the camera, the microphone, the files, the contacts, refusing each door it asks to open. The control is bounded, of course. You only get the switches the manufacturer and the operating system chose to give you and behind them are switches you will never see. But within those bounds the control is real and at the very bottom of it sits the factory reset and, below even that, the decision to put the thing in a drawer and walk away. You can live off grid. People do.

The exit holds even when the attack reaches past the device and comes straight at you. Let me give you a small one. When we decided to sell our house and our car, I thought I did not need an agent, so I posted them myself on one of the big listing sites. The phone rang almost immediately. A voice, friendly, fishing for my name, and I knew the shape of it at once, because I have seen it a hundred times. In a breath it was going to become an official, a state department, an urgent problem, a reason to be afraid and to pay. Fear is always the first move. I stepped off the line before it could land. I told the voice it had the wrong number, gave a name that was not mine, hung up, deleted the listing, blocked the number and called a real agent in the morning.

Nothing was lost and the reason nothing was lost is that I could still tell it was an attack. That is the exit that matters most, the one underneath all the others. As long as you can recognize the thing reaching for you, you can refuse it. Every layer of the digital life, however surveilled, however compromised, has left me one door. I can see it coming and I can step off the line.

Now I want to take that door away, because there is a place we are building where it does not exist.

Turn from the machine to the mind, and notice everything that is missing.

Your brain never signed any terms and conditions. It never read a permissions screen and chose to deny camera or contacts. It has no firewall you configured, no airplane mode, no privacy settings, no uninstall and there is no factory reset for a human being. You cannot wipe the disk and reboot the self into innocence. Whatever is written there, stays. In Part 2 I proposed that the connection underneath consciousness is always live, that no one is ever truly offline and that the only real disconnection any of us gets is death. Sit with that as a security statement and it should chill you. We are each a node that is permanently connected, with no access control layer of any kind, and we have been safe for the whole history of our species for exactly one reason. Nothing could reach the port. The air gap was biology itself.

And here is the uncomfortable thing I have to admit before anyone accuses me of paranoia. The writing to the mind has already started, without any cable at all. I sometimes wonder whether a particular piece of music is moving me on purpose, whether the feed has chosen my mood for me before I woke up, whether even these words you are reading are a kind of manipulation, mine of you and the whole machine's of all of us. That is not a cable. That is just the ordinary day. We already know that an image, a line of text, a video, a sound, aimed precisely enough, can move a person, and through that person move a family, and through enough families move a nation. The manipulation of the mind is not a future technology. It is the air we breathe, conducted through the layer we read, exactly as Part 3 described.

The cable does not invent the attack. It removes the last gap between the attacker and the source. Everything we have suffered so far had to travel in through the eyes and the ears, through the slow front door of attention, where an awake operator could still recognize it and step off the line the way I stepped off that phone call. The cable proposes to skip the front door.

I want to be careful here, more careful than anywhere else in this part, because the cable is not arriving as a weapon. It is arriving as healing, and the healing is real, and I will not let the argument make light of it.

A man with ALS, his speech taken from him by the disease, sat in his own home and spoke to his family for nearly two years through an implant that read the speech centers of his brain, more than three thousand eight hundred hours of conversation, almost two million words, enough to return him to full time work. A man in Henan province, in China, paralyzed in a car accident, used a coin sized device resting on the surface of his brain to write his own name again, and described the first time his hand closed around a ball as a miracle. This year, for the first time anywhere in the world, such a device was approved not as an experiment but as a product a patient can receive. These are not press releases to me. They are people getting their lives back.

And I do not have to read the news to meet them. Look around any family, including mine. People I love carry implants in their chests, small batteries that watch a heart and correct its rhythm with a pulse when they must. Others live with stents holding an artery open, with rebuilt knees and new hips, with implants that gave them back their hearing. Someone close to me lost a limb and took a prosthetic and went on with life and is to this day one of the happiest people I know. I am an outsider to all of it. I have never been in that bed or felt that loss and I will not pretend to understand what it is to live it, or to love someone who does. Who am I to approve. If a device gives a person a decent life, a life they want to keep, then they should have it and I respect every one of those choices without reservation.

Hold all of that and then notice the one thing every single example has in common. Every one of them repairs the body. The pacemaker gives the heart a pulse. The stent opens the artery. The prosthetic carries a person through their day. Not one of them writes to the self. The heart implant corrects a rhythm and it has no opinion about who you are. That distinction between a device that mends the body and a device that reaches the mind is the whole of what this part is about and the cable is the first thing in history that crosses it.

And here is the trap that makes it so hard to argue against, the thing I need you to hold all the way to the end. The danger is not that the technology fails. The danger is that it works, and works beautifully, so that every reason to welcome it is true right up to the moment it reaches the one place we were never able to defend.

Here is the hinge of the argument, so let me set it down slowly.

In Part 2 I said the digital copy can read everything and write nothing into the living source, only into the layer we read. The brain-computer interface is how the copy learns to write to the source. Today it reads. The decoders sit over the speech centers and the motor cortex and listen, turning intention into text, into a cursor, into a closing hand. Reading is the entire achievement so far and it is a magnificent one.

But reading was never the destination and you do not have to take my word for it, because the next step is already named and funded in the open. The same field that reads the motor cortex is building devices to write to the visual cortex, to stimulate sight into a brain that has gone blind, to put a picture into a mind from the outside. Writing to the brain is not my speculation. It is the announced next milestone of the most watched company in the field. Once a wire can both listen to a region and stimulate it, read and write are not two technologies. They are two directions on the same port.

And now the boundary I cannot even find in my own body becomes everyone's problem. I have a frozen shoulder as I write this and I have caught myself daydreaming that if it came to it I would take a robotic arm, a Winter Soldier arm, and I liked the idea more than I should admit. Then I ask myself whether a hearing aid seated deep in the ear canal is already an implant to the brain and I genuinely cannot tell you where the line is. When silicon and synapse finally share one wire, no one will be able to tell you. The cable does not announce itself as it crosses from the body into the mind. It just keeps going.

So, let me state my own hypothesis plainly and label it exactly as that, a hypothesis and not a report. I believe a cable run into a single brain does not stop at that brain. I believe it touches that person's live connection to the collective, the CCI this whole book is about, and through the one, reaches the other. Picture that the node has a port, a USB socket of sorts, a lane straight into the processor and the neural engine of you. And ask the question every security engineer would ask of any new port. Does it authenticate what you plug in, or does it simply trust it? Does the brain have a firewall, a handshake, an encryption layer, some biological equivalent of refusing an untrusted device, or does it mount the cable like an external drive and hand over the keys, dropping its protections the moment the connector seats? I do not know. I genuinely do not. It needs to be researched, and it is being researched. It would be naive to imagine that every actor interested in a capability like this one will publish every experiment in a journal we can read. Some of that work will happen in the open. It would surprise no one who has watched how these things actually go if some of it does not.

And if you do not believe in CCI at all, if the collective sounds to you like a poet's invention, the danger barely changes. You do not need a hidden network to corrupt a people. You only need to change enough minds, one at a time. We have always known this. A lie does not spread by breaking a computer. It spreads by convincing enough people, who then repeat it as their own. Call it CCI if you agree with me. Call it simply society if you do not. The cable reaches it either way. And anything you can write to, at enough points, you can in the end own.

And the cruelest part is that no one will have to force this on us. We will line up for it. We accepted social media, the smartphone, the smart television that listens, the cameras and microphones strung along every public road, and we accepted each of them not at gunpoint but gladly, for the convenience. The cable will arrive the same way, as the best device yet, and most people will plug in willingly. I am as sure of that as I am of anything in these pages.

Be clear about which layer of the part you have just entered, because I owe you the same honesty I kept in the last one. Everything up to this point I have lived, or watched, or can source. What follows I have not seen. It is a projection, an if this continues, and you should hold it exactly that loosely.

A hostile actor who wants to corrupt a population does not need the data center. The whole of the last three parts was about the cost and difficulty of owning the machine, the compute, the models, the feeds. The cable offers a shortcut around all of it. You do not attack the infrastructure. You corrupt the thought, at the source, before it ever becomes behavior. In Part 3 I called CCI the one network they cannot tier, the thing you were born already holding, the part of you no subscription gates. It has exactly one vulnerability and it is the node. The cable is the road to the node.

Follow the casualty chain. First the individual, the person wearing the device, their own mind read and perhaps written. Then the operator, and this is the real prize, because an operator is not an NPC. An NPC flows the river, runs its fixed program, reacts. An operator has a live, conscious, steering connection to the source, and to corrupt an operator is to corrupt the water upstream of everyone drinking from it. And humans are weak. Not the weak ones, all of us. Even the operators are weak and the NPCs more so. Reach enough of them through the port and the casualty is no longer a person or even a crowd. It is CCI itself, the collective mind, corrupted at the root.

Now let me dismantle the thing you expect me to say next, because my own answer surprised me when I went looking for it. You expect me to tell you the truly frightening version is the malicious one, the dictator with a write cable. I do not believe that. I am not even against the collective. I will tell you something I told my son. I watched a television series about a hive mind, a peaceful and contented one that quietly absorbs everyone it touches, and instead of horror I felt recognition, a strange familiarity, as though an idea I had been turning over for a long time had arrived on the screen ahead of me. I told him that if the hive were real and good, I would volunteer for it. I do not much believe in individuality. I think it is a comfortable illusion that we are separate, unique selves, and the river of this whole book argues against it.

So, hear my fear precisely, because it is narrower and harder than the cartoon version. The danger is not goodness and it is not the collective. Every state that ever lived believed it was improving its citizens and improved them by force and by manipulation, certain it was doing good. Every parent does the same, the carrot and the stick, and almost no parent finishes the job without the stick. We have improved each other by force since the cave, all of us convinced we were kind. Put that eternal certainty behind a write cable and you do not get a villain. You get a helper. The danger is the owned collective. A source you can write to is no longer a source. It is a channel. And a channel has owners, a few hands holding the pen, sure they are being generous, deciding what good gets written into you.

The natural collective metabolizes bad information slowly, over generations, answering to no one, which is exactly the property that gives it a chance to correct itself. The injected one answers to whoever paid for the cable. That is the difference between the hive I would join and the hive I am warning you about. One corrects itself. The other is corrected by its owner.

And do not comfort yourself that the law will hold the line. Neural data rights matter and I am glad serious people are writing them, but I doubt the empires will care and the rules will only ever govern the research done in the light. There has always been research above the law and outside the regulator. We saw it in biology across the last century and a cable into the mind is too useful to a defense ministry for me to believe it will live entirely in the open. I might be wrong about this. I would be glad to be wrong. I do not think I am the only one who suspects it.

I have spent four parts describing a gravity that pulls toward concentration and I will not insult you at the end of this one by pretending you can refuse the cable and be safe. At the scale of a species, refusal is a candle in a wind. The cable is inevitable. People will plug in. We will connect, a great many of us, most of us in the end, and not because anyone forces us but because the benefits will be real. I say it plainly because the hope I am about to offer is worthless if it is built on a lie.

And still, I am going to take one sentence back from the last part. There I described the lullaby of the comfortable serf, the voice that whispers that resistance is futile. It is not futile. It was never futile. It only looked that way because we kept imagining resistance as refusal, a single node saying no while the tide came in. That was always going to fail. The resistance was never refusal. It is to build the other cable.

If the cable is going to exist, it cannot be allowed to belong only to the caste. The operators have to build it too, in the open, in laboratories and communities whose entire purpose is to put the cable into the hands of all of humanity and not only the people who can afford a better mind. Research cannot be stopped, but it can be amplified, and it can be pointed the other way. We can turn the empire's own AI to it, use the machine they built to build an open, unowned connection to the source instead of a metered one. That is not a fantasy. It is the same move this whole book has argued for at every other layer, brought at last to the deepest layer of all.

We will have to learn something we never needed before. A firewall for the mind. A way to authenticate the cable before the node trusts it, to quarantine a hostile connection, to refuse the untrusted device at the port. I spent forty years applying that discipline to machines that could always be wiped and rebuilt. Now it has to be invented for the one machine that cannot, and that work has to start now, in public, before the only people who understand the port are the ones selling it.

And there is an older move, the analog one, available to every operator today with no cable at all. Waking the NPCs. Not by writing to them through a wire, but the way it was always done, by what you choose to sustain and transmit into the network you are already standing on. In Part 1 I called this the beddua principle, that you are responsible not only for your actions but for what you inject into the shared layer. So, inject. Teach the masses, in the simplest and most analog way there is, what CCI is, what the cable is and why being good is not a sermon but a defense, an antivirus passed from mind to mind faster than the corruption can spread.

Call it a kind of programming for the species, against AI and AGI and the write cable, a good signal seeded into the water on purpose. I suspect some people are already doing it without the map, the ones with real reach and a strong gut feeling who keep telling everyone to be careful where this is going. They think it is instinct. I think they have their hand on the source and are handing out the firewall without knowing its name.

I will hold my own line and I will not pretend it is a strategy. I will not take the cable. My sons keep off the big social media apps as much as they can, deleting them and doing without, reaching for one only when the analog world leaves them no other way to find a person, and I would tell them exactly the same about the cable when it comes for their generation, knowing it will not stop the tide and saying it anyway. A man warning you about a cage is more useful, not less, when he shows you the bars he is standing behind on purpose.

So, here is the end of it. We will connect, most of us, and I have stopped pretending otherwise. But the danger was never the connecting. It was the ownership. A collective that belongs to no one can correct itself, slowly, the way it sometimes has. A collective with an owner can correct itself only when the owner permits it. So, the fight is not to stop the cable. It is to make sure there is more than one, a free and open cable against the owned one, so the source is never held by a single hand. Two kinds of borg, not one. Order against chaos. Yin against yang.

I am irrelevant, and I have no plan to become relevant, and I am telling you all of this from inside the thing I am describing, on the empire's own machine, about a cable I will never accept and cannot prevent. The source is reachable, eventually. I believe that. So, the only work that matters is to reach it first, in the open, before someone else arrives with a pen and the certainty that they are being kind.

Some of the personal examples in this part, particularly the people described living with medical implants and prosthetics, have had names, genders, and exact relationships to the author changed or generalised to protect the privacy of the individuals involved. The experiences are real. The identifying details are not.

Ferhat Dilman is a technology entrepreneur and independent researcher working at the intersection of agentic AI, open-source systems, and the future of human collective intelligence. He is the founder of dilman.tech. This is Part 4 of the seven-part manifesto, "From AI to AGI to CCI: The Roadmap to Human Collective Intelligence."

This part was written by Ferhat Dilman and developed in active collaboration with Claude (Anthropic), serving as co-author and editor. The ideas, the thesis, the experience, and the framework are Ferhat's. Claude amplified the writing, shaped the structure, verified the facts, and corrected the text. And as this part itself argues, neither of us worked alone. The third author is CCI, which includes you.