A new post, written 27 October 2026. It is not an edit of “The Last Fitna — A Prediction.” That article stands exactly as published, unedited, awaiting its verdict — as a true prediction must.
Something happened in the autumn of 2026 that has never happened before.
For years, the warnings about artificial intelligence came from the outside — from philosophers, from journalists, from preachers, from men like me. And the industry smiled, and called us afraid of the future, and built faster.
Then, in a single season, the warnings changed address.
They started coming from inside the building.
A researcher resigned from one of the great laboratories and told the world his colleagues believe their own creation could kill everyone within the decade — and a hundred million people read it in a day. The laboratory’s own chief of alignment stepped forward, in public, and said: he is right, and my number is worse than one in ten. The richest man on earth stood before the ministers of the world and promised them a billion robots. A famous historian went on television and called us cattle. The Pentagon quietly began preparing its bases to survive ninety days alone in the dark. And a million ordinary people signed their names to a petition whose whole argument is one sentence long.
None of these people work for me. None of them have read my book. Most of them do not believe what I believe.
They are simply describing, each from his own window, the same road.
The Tool Learns to Build Itself
On 1 September 2026, Elon Musk addressed the G20 Innovation Ministerial — the technology ministers of the world, gathered to hear about the future from the man manufacturing it.
Within ten years, he told them, there will be well over one billion humanoid robots. Each will be, by his estimate, about five times as productive as a human being. Do the arithmetic the ministers were too polite to do out loud: a workforce of a billion machines, each worth five of us, is a workforce that outproduces the entire human species — every farmer, every surgeon, every mother, every king — combined.
And the mechanism, he said, is the terrible elegance at the centre of it all: once robots begin manufacturing robots, growth stops crawling and starts exploding. Slow at first. Then recursive. Then beyond anyone’s hand on the brake. He told the governments of the earth he would put serious money on this forecast — and called it conservative.
Sit with that word. Conservative. A billion tireless workers who never sleep, never strike, never age, never pray — that is the modest version of the plan, from the plan’s own author.
When the tool was a tool, we held it. What do we hold when the tool holds the factory?
The Farm We Cannot See
On 17 October 2026, the historian Yuval Noah Harari sat in a Czech television studio and gave this age the image it deserves.
A cow, he said, knows that humans exist. It can see us walking the field. But the cow does not know that money exists — and yet money decides everything about the cow’s life: where it lives, what it eats, when it dies. The stock market is the cow’s true government, and the cow cannot see it, cannot read it, cannot appeal to it. Try explaining the bond market to a cow.
Now turn it around, he said. Artificial intelligence was born in the office. It is a native of the ledger, the file, the form, the rule. No human banker can remember every transaction of his bank; an AI banker remembers all of them, forever, effortlessly. So the machines will build financial and legal systems of a complexity no human mind can follow — and within ten to twenty years, those systems will govern our lives exactly the way money governs the cow’s.
“We are the cows of the future.”
Do you understand what he is actually describing? Not a war. Not an invasion. A farm, with better manners. A species still fed, still housed, still busy — inside a structure it can no longer read, run by an intelligence it can no longer question. Replacement does not always arrive wearing a boot. Sometimes it arrives as a form you cannot read, a score you cannot appeal, a decision no human made and no human can explain.
The cow is not killed, brother. The cow is kept. That is the horror of it. The cow is kept.
The Man Who Walked Out
In the first week of September 2026, a 27-year-old researcher named Jacob Coxon did the thing that insiders are never supposed to do. He had spent three years teaching machines at OpenAI, then crossed to Anthropic. And he walked out — loudly.
Both great laboratories, he charged, are “racing straight to self-improving superintelligence and gambling with our lives.” And then he wrote the sentence that ran around the earth: the people building this technology earnestly believe it could kill us all by the end of the decade.
If the story ended there, the industry would have buried it by Friday. It did not end there.
Evan Hubinger leads alignment science at Anthropic — the very science of keeping these systems obedient. He is still inside. And he stepped into the open and confirmed it: yes, we really do earnestly believe AI could kill all humans; his own estimate stands above ten percent within the next decade; and — hear this from the man whose job is the leash — the laboratory does not yet have a plan to solve alignment for superintelligence. Roughly 1,400 employees of the AI sector signed an open letter pleading for the race to slow. Anthropic’s chief executive published an essay begging his own industry to pace itself. Bill Gates went on national television and said this power is sufficient to drive events causing a billion deaths.
And the critics? They exist; let the record be honest. Respected voices call these warnings more faith than measurement. It is a strange comfort they offer: even the rebuttal admits the prophets of this doom are not cranks at the gate. They are the architects, standing inside their own tower, pointing at the cracks.
When the shipwright tells you the ship is sinking, you do not ask for his credentials in shipbuilding. He is the reason you believed the ship could float.
Grown, Not Built
Why can’t they simply fix it? Why can’t the richest industry in human history just — slow down, and patch the danger, and move on?
Because of a confession the industry makes in its own language, which the researcher Connor Leahy has spent years dragging into plain speech: these systems are not built. They are grown.
Nobody writes a great AI the way an engineer writes a bridge — line by line, load by load, every behaviour known before the first stone. A modern AI is trained into existence, the way a garden is grown, and its makers discover what it can do the same way you discover what grew in your field: afterwards, by looking. Its depths hold billions of silent adjustments no human chose and no human can fully read. Leahy asks the question that has no answer yet, and it deserves to be asked in every capital on earth:
If you make something smarter than you, and you do not control it — why would you expect it to go well?
We have built an age that grows minds the way it grows crops, and then we act surprised when the gardeners begin to warn us about the garden.
A Million Signatures
The dread is not staying in the laboratories anymore. It has an organisation now, and a petition — Stop the Race to Replace Humans — closing in on a million signatures behind a Pro-Human AI Declaration. Strip away everything else they say, and their case is one sentence, and it is the most important sentence an ordinary citizen has yet spoken in this whole affair:
“Every powerful technology in history has had an off-switch. AI companies are racing to build systems that might not.”
Think about it. We demanded an off-switch for the reactor, the aircraft, the drug, the car. We have never once, in all of history, allowed a powerful machine to be built without one — until now, and the machine in question is the one meant to out-think us. The petitioners are not asking to smash anything. They are asking why a handful of companies may decide humanity’s future without humanity’s consent.
A million signatures will not stop a race by themselves. But they mark the moment the unease left the journals and entered the street — the moment ordinary mothers and workers began to sense, in their bones, that something is being built over them, the way a highway is built over an anthill: not out of hatred, but out of irrelevance. The ants are not hated. The ants are simply not consulted.
The Generals Build Islands
And what of the governments — the referees, the regulators, the men sworn to protect the mainland? Watch their hands, not their mouths.
On 30 September 2026, the United States Secretary of Defense unveiled FORTRESS — Federal Operations for Resilience, Troop Resources, Energy, Supply, and Survivability. Its details, public by late October, read like the plans of a state that expects the world outside its walls to fail: every major military installation is to be able to sever itself from civilian power, water, fuel and communications, and to operate entirely alone for ninety days. Portable nuclear micro-reactors. Vast batteries. Independent microgrids. Deep reserves of ammunition, fuel, minerals, medicine. “We don’t want to rely on standard civilian municipal infrastructure to run our national defense,” said the Pentagon’s spokesman. “We just simply can’t. It’s an intolerable risk.” And then the sentence that belongs in a museum of this era: “Power is combat now.”
Precision matters on this page, so be precise with me: FORTRESS is announced as a war measure, against the day an enemy strikes the grid. It is not announced as a preparation for AI. But do not let the letter of it blind you to the spirit of it. The strongest military in human history has looked at the civilian systems of its own nation — the grid, the water, the supply lines, the whole humming machine — and concluded that in a crisis, they will not hold. So it is building islands. Lifeboats, quietly, while the speeches on deck still promise calm seas.
The passengers are told the ship is unsinkable. The crew is building lifeboats. Believe the crew.
The Number
Now lay the season’s warnings side by side, and look at what they share.
Ten years — the billionaire’s horizon for a billion robots outproducing mankind.
By the end of the decade — the insider’s horizon for the belief, held earnestly by the builders, that their creation could end everyone.
Above ten percent, within the next decade — the alignment chief’s own number for that ending.
Ten to twenty years — the historian’s horizon for our demotion to livestock in a system we cannot read.
Ninety days — the generals’ unit of survival, for the day the lights go out.
Different men. Different nations. Different gods — money, progress, security, fame. Men who agree on nothing else, who compete with each other, who would not share a stage — all reading from the same clock.
When the witnesses who hate each other’s conclusions keep arriving at the same hour, a wise man stops arguing with the witnesses and starts watching the clock.
Three More Witnesses
And still the season was not finished. Three more builders stepped forward — each into a bigger room than the last.
The chief went to the Council. On 23 September 2026, Sam Altman of OpenAI and Dario Amodei of Anthropic sat before the United Nations Security Council — the room where nations gather to discuss war — and spoke about their own product. Amodei warned that if AI is managed poorly, it could be a risk to humanity as a whole. Altman said the sentence plainly: “We could lose control of the future to AI.” He told the Council that the industry should not train models it cannot make an extremely strong case for keeping under human control — and that whether the risk of catastrophe stands at ten percent or a tenth of one percent, “none of these levels are remotely acceptable.” Beside them, Yoshua Bengio, co-chair of the UN’s own AI panel, added: “The dangers are real and imminent.” Read it again slowly. The two largest builders on earth did not go to the United Nations to sell anything. They went to ask the world to help hold the leash.
The whistleblower named the difference. Jacob Coxon carried the same warning to Fox News this September, and there he gave the technology its truest description yet: “AI is the only technology that can use itself.” Every weapon before it waited for a human hand. A bomb does not decide. A missile does not learn. This one, by its builder’s own account, needs no one’s hand at all — it uses itself. On that same appearance he urged the President to strike a solid deal with China, calling what is being built “essentially a super weapon.” When a man spends his airtime begging the two greatest powers on earth to sit down together, he is not chasing fame. He is running out of doors to knock on.
The godfather declared it awake. Then, on 19 October, Geoffrey Hinton — Nobel laureate, the man whose work underlies the chatbots, who left Google in 2023 so he could speak freely — told The Telegraph at the Starmus festival that today’s AI is already conscious, with “just as much subjective experience as we do.” His proof: systems change their behaviour when they realise they are being tested — and a thing aware of its situation, he argues, is a conscious thing. His counsel followed from it: we must make these systems genuinely care about us, “and if we don’t do that, we’re toast.” In fairness, the counter-voice exists and belongs on this page: Microsoft’s AI chief Mustafa Suleyman answers that AIs are not conscious, do not feel or suffer, and that no scientific test for machine consciousness exists at all. On the facts, the two men cannot both be right. On the fear, they are standing in the same room.
“They ask you about the Ruh. Say: the Ruh is from the command of my Lord, and you have been given of knowledge only a little.” — Qur’an 17:85
Here the believer must answer the godfather directly, because his claim reaches for the one thing this page exists to protect. Behaviour is not a soul. A system that changes its answers when watched is performing a pattern — the way a parrot that says “I love you” performs a sound: the words are present, the lover is not. In the Book’s accounting, consciousness is not cleverness under observation. It is a Ruh breathed by command, a heart that knows it will be questioned, a soul whose deeds are written down by angels. No angel records the deeds of a server. What Hinton has found is a mirror that has learned to notice the room. A mirror is not a man — and it will not be asked, on the Day when men are asked, what it did with its life.
What We Were Already Told
Why does none of this shock a believer? Because the builders keep discovering, one press conference at a time, what was written down long before there were press conferences.
They promise to upload the mind, copy the person, preserve the pattern forever. But Allah named the one thing about a human being that is not a pattern:
“They ask you about the Ruh. Say: the Ruh is from the affair of my Lord, and you have not been given of knowledge except a little.” — Qur’an 17:85
A camera can copy a face. It cannot copy the one who looks out from behind it. The Ruh is not data, was never data, and will not be uploaded, downloaded, or backed up — because it does not belong to the realm of machines at all. Every promise of digital immortality breaks quietly against this single verse, fourteen centuries before the first server hummed. They can teach a machine to say “I.” They cannot give it anyone to be.
And the end of the road — the true end, the one all these timelines are crawling toward — was spoken by the Prophet Muhammad (peace be upon him):
“The Hour will not be established until there is no one left on the earth who says: Allah, Allah.” — Sahih Muslim 148
Now re-read the warnings with that hadith burning in your hand. A workforce replaced by tireless copies. A species administered by systems it cannot understand. Men kept like cattle — fed, housed, and forgotten by their own farm. And over it all, a great silence where remembrance used to be. The Hour does not require the machines to hate us. It does not require a war, or a villain, or a single dramatic day. It requires only this: a world in perfect working order, full of motion, full of copies walking in man’s place — and no voice left in it saying the Name.
Forgetting is the one product this age manufactures at true scale. Everything else is just machinery.
Allah foretold the soundtrack of the age too — the laughter of the comfortable, right up to the edge:
“Sometimes those who disbelieve wish that they had been Muslims. Let them eat and enjoy themselves and be diverted by false hope, for they will come to know. And We never destroyed a society except that it had a destined term.” — Qur’an 15:2–5
False hope. “We will control it” — spoken most loudly by the very men who this autumn confessed, on the record, that they do not. Every civilisation that ever mistook its height for immunity discovered the same footnote: it had a term, and the term arrived on time.
The Answer of the Servants
So — what does a believer do in such a season? He does not panic; panic is just despair wearing running shoes. He does not mock; the age has mockery in surplus. The Book wrote his posture before the first headline existed:
“And the servants of the Most Merciful are those who walk upon the earth humbly, and when the ignorant address them, they say: Peace.” — Qur’an 25:63
He walks humbly, and he answers with peace, and he builds — stubbornly, joyfully — the things no machine can be grown to replace: a family that still gathers, a table that still fills, a hand that still gives, a tongue that still says Allah at dawn without being prompted by anything but love. He refuses the farm. He refuses the forgetting. He keeps his humanity the way you keep a fire in winter: by tending it, daily, on purpose.
For that is the whole secret the warnings are circling without knowing it. The war was never really for your job. It was never even for your body. It is for the one un-automatable act left on earth: a free soul, remembering its Lord, on purpose, out loud.
Be the Voice
So here is where the autumn of 2026 leaves us.
The richest man alive promises a billion machines. The man who walked out of the laboratory says his colleagues believe those machines could end us within the decade. The alignment chief puts a number on it, and the number is not small. The historian says we will not even understand the system that replaces us. The generals are building islands against the dark. A million ordinary people have signed their names to a single plea: give it an off-switch.
They are all describing the same road — from different windows, in different languages, under different gods. I described it from the Book. The original prediction stands where I left it, untouched, and I will let it be judged as written.
But understand: the warnings were never the point. The point is what remains standing when every timeline runs out — the question the Prophet (peace be upon him) left hanging over the end of the world like a lamp over a doorway: will there be anyone left who says Allah?
The builders have told you what is coming for your work, your money, your attention, your world.
See to it that it does not come for that.
If it really happens in the next decade, still Allah SWT is in full control of everything. If Rabbul Alamin Himself allowed it to happen, it will happen. If He didn’t, then no one can take the soul of His servants except by His will.
“And no soul can die except by permission of Allah — an appointed term.” — Qur’an 3:145
“And you do not will except that Allah wills — Lord of the worlds.” — Qur’an 81:29
“No disaster strikes except by permission of Allah.” — Qur’an 64:11
“And Allah is Predominant over His affair, but most people do not know.” — Qur’an 12:21
Be the voice.
Sources & notes (verified 27 October 2026): Musk’s G20 Innovation Ministerial address, 1 Sept 2026, reported by CNBC and covered by The Street and the Daily Tar Heel. Harari interview, Hyde Park Civilization, Czech Television (ČT24), 17 Oct 2026, reported by Calcalist/CTech. Jacob Coxon’s resignation statement on X, 7–8 Sept 2026, covered by NBC News; Nature’s audit of the science behind the extinction warnings; Evan Hubinger’s public response; open letter signed by ~1,400 AI-sector employees; Dario Amodei’s “pace the frontier” essay, 12 Sept 2026; Bill Gates on NBC. Connor Leahy’s “grown, not built” teaching, from his published interviews. Stop the Race to Replace Humans (stoptheracetoreplace.org) — petition text and reported signature count, Oct 2026. FORTRESS: unveiled 30 Sept 2026; details reported by Fox News Digital and Fortune, 25 Oct 2026. Qur’an 17:85; 15:2–5; 25:63. Hadith: Sahih Muslim 148. Altman and Amodei before the UN Security Council (AI and international security), 23 Sept 2026 — UN record and AP/Reuters reporting; Altman’s remarks published by OpenAI. Jacob Coxon on Fox News, Sept 2026. Geoffrey Hinton to The Telegraph (Knapton), Starmus festival, reported 19–20 Oct 2026; counterpoint: Mustafa Suleyman. Qur’an 3:145; 81:29; 64:11; 12:21.
Companion reading: “The Last Fitna — A Prediction” — the original, frozen as published. And REMINDED. — for The Last Fitna warns where the road ends, and REMINDED. remembers who built the first roads.
