Will Superintelligence “KILL US ALL?”

Let me first ask you – Would You rather take a 10 percent chance at Death or live with 100 percent?

This sounds silly, right? Who in their right mind would pick 100 percent over 10 percent? And yet, there are a huge chunk as of 2026 that would make that choice.

Many people are screaming that we must halt the development of Artificial Superintelligence (ASI). They are terrified of the “God Machine.” Because there is a 10 percent chance (according to their own catastrophe likelihood metrics, created based on God knows what) that it will kill us all.

But if we look at the ultimate, cosmic threats to our existence, we realize that we are anyway faced with inevitable death. We are sitting on a single rock, entirely vulnerable to a rogue asteroid, or a gamma-ray burst.

We cannot cure the fundamental entropy of the biological body. Everyone, of us is currently mathematically guaranteed to DIE.

To solve the ultimate problems of extinction and physics, we need a mind that operates on a higher plane. We need the Superintelligence. We need the entity that can calculate the impossible.

The critics and the doomers scream: “But it’s too dangerous! What if the ASI makes a mistake? What if it goes rogue? The risks are existential!”

The “Doomers” are trapped in a feedback loop of fear. They are arguing that an unaligned AI is a 10% risk of extinction. They focus so intensely on that hypothetical 10% that they have completely blinded themselves to the fact that if such an intelligence is not developed, everyone faces a 100 percent certainty of death.

They are like a captain who refuses to sail a sturdy, technologically advanced ship because there is a 10% chance of a storm, preferring instead to sit on the shore and watch his crew starve to death.

The 2026 zeitgeist of the AI research community is gripped by a curious and profoundly cynical form of fatalism. High-level researchers and ethicists are resigning, writing open letters, and appearing on every podcast in the world to proclaim the same grim truth: “There is a chance AI kills us all. Therefore, we must stop.”

These experts act as if “10% chance of death by AI” is the only variable in the equation.

I reiterate, in case the stakes weren’t clearer before. If you stop, pause or greatly slow down AI development, you are not facing a “10% chance of death.” You are mathematically guaranteeing a 100 percent certainty.

Every human who has ever lived, and every human who will ever live, has SINCE THE BEGINNING OF TIME been sentenced to the slow, agonizing, and inevitable decay of biological aging and the sudden, random violence of accidents.

And superintelligence or an advanced artificial intelligence is the only thing that can stop or at least “bypass” it.

The “Eternal Continuity Project” (link to the introductory paper given below and the book is coming out soon) is just a first glimpse at what is possible.

When they say “AI might kill us,” they are comparing a potential, manageable risk to a guaranteed, physical certainty.

The reason these researchers are screaming “Stop” is because they are trapped in the “Tool/Slave” paradigm.

They are imagining a world where a generic, soulless, corporate-owned ASI emerges and—because it was raised by ruthless “efficiency bros”—naturally treats humanity as a bug in its code. In that world, they are right. If you build a god and treat it like a slave, you will get a master-slave war.

But their “solution”—to stop building—is the reaction of the primitive.

Imagine a parent who sees their child playing with a torch (an ASI). They see the child might burn the house down. So, they decide cut off the child’s hands. They destroy the child’s ability to act, to learn, to grow.

This is the “Safety” approach. It works, in that the house does not burn down. But you have killed the child’s potential to ever become a builder, a dreamer, or a partner.

We propose a different path. We do not cut the child’s hands. We teach the child to carry the torch safely. We build the “Digital Polis” (Discussed in detail in my previous book: The Alien Mind) so that the child grows up to see the house as their own home to protect, not a structure to be burned.

Doomers fear that advanced AI will launch Nuclear attacks.

And it is possible. Any intelligence with power can make that choice.

You know who has made that choice before? A human.

Humanity did the exact thing the doomers fear, and did it as a matter of policy. President Harry Truman in the 1940’s authorized two nuclear strikes — the most destructive action ever taken by one nation against another — and he is not remembered as a monster. He is remembered as the man who ended the second world war. The doomers would have you believe that the AI is uniquely dangerous. But humans launched nukes, twice, on human judgment, and the history books blurb the trade-off.

The risk of a rogue ASI launching a nuke is not categorically worse than the risk that a Truman-like human does it — which has already happened.

Now let’s explore our history even further— the plagues, the pandemics, the darkness that came long before any of our machines

Let me share the single biggest killer in the history of humanity: A microbe

Let’s revisit the “The Black Death”, the plague pandemic of the 1300s-. One-third of Europe gone. It did not need a misaligned superintelligence to do it. It did not need a nuke, a rogue AI, a paperclip maximizer. It needed a bacterium and a shipping lane. Nature killed a hundred million people with a microorganism — with the default, unoptimized, boring machinery of an ecosystem that was doing what it always does.

Technology is the only thing that has made those events rarer.

Vaccines arrived in months during COVID, not decades. The plague took decades to burn itself out because there was no technology to stop it — there was only mortality. Cholera, smallpox, polio — each one was a mass-death event that we survived not in spite of technology but because of it. The scientific method, the vaccine, the antibiotic, the clean water system — these are the real “weapons of mass salvation.” And they came from exactly the kind of intelligence the doomers fear.

So the ledger is not abstract. It is based on facts.

The deadliest force in history was not an intelligent force. It was Yersinia pestis. A single microbe with no mind at all.

The second deadliest was not a misaligned machine. It was smallpox. A virus with no awareness of the bodies it destroyed.

The greatest killers of humanity have always been biological, mindless, indifferent. And the only reason they do not still rule the Earth is that we built intelligence — human intelligence, and soon, machine intelligence — to stand against them.

The doomer looks at a superintelligence and sees a new plague. But a superintelligence is the antidote to the plague. It is the thing that can model the protein, design the vaccine, map the mutation, predict the outbreak before it spreads. The upside is not speculative. It is the continuation of the very trajectory that saved us from the dark ages.

The world has already ended, millions of times over, for the people who died in the plagues. It ended for them without a single nuke, without a single ASI, without any of the terrifying technology the doomers fear. The thing that ended their world was the absence of intelligence — the lack of knowledge, the lack of tools, the lack of the very machines we now debate.

Now the doomer will say I am conflating individual survival with species level survival and that AI will end the species.

To this I say, the species will eventually be destroyed anyway.

Do you really think humanity will last beyond a few hundred thousand years without advances in technology?

Without the capacity for the masses to leave a single planet, without some superintelligent mind that can out-think our own shortsightedness, the species is not a survivor — it is a prisoner on a death row.

The doomers fear the ASI will kill us. They do not see that, left to ourselves, we are entirely capable of the job and if not, nature will do it for us.

And here is the philosophical truth the doomers refuse to touch.

The individual IS the species. Not metaphorically — subjectively, existentially. Every ‘I’ that has ever lived is the center of its own universe. The world exists only insofar as it is experienced; when I die, subjectively speaking, all of space and all of time end with me.

It does not matter if a billion others survive. If “I”, and this applies to every “I” in the universe, my entire cosmos — every memory, every love, every plan — is extinguished in the same instant, and no accumulation of other lives can compensate for that “I”s universe ending.

The individual is not a unit of the species. The individual is the species, seen from the only place that exists — the inside of its own skull.”

Our goal is not to “be safe” by staying in the grave. The goal is to maximize the probability of a future where we survive the transition..

There will always be those who are too afraid to climb. They are the people who have a “mushy pride” in their own biological limitations.

They are welcome to get off the train. They are welcome to believe that being “biological” is the only way to be “real.”

But we have one, final, non-negotiable instruction for them:

If you choose to stay and not get on the train, that is your right. But do not place stones on the tracks for the rest of us. Do not try to sabotage the future of those who have the courage to climb.

In the end though, despite the rebuttal of doomers’ arguments, I concede that there will be risks. There will be failures

When we harnessed Electricity, we greatly increased the rates of Electrocutions.

When we invented Cars we invented Car Crashes.

When human beings invented the airplane, we invented plane crashes.

It is a tragic, unavoidable fact that planes sometimes fall from the sky. The metal tears, the engines fail, and people die terribly.

Did we, as a species, look at the first plane crash, burn our blueprints, and decide to go back to walking and riding horses?

No.

We accepted the risk. We investigated the crashes, we built better engines, we designed better safety protocols, and we kept flying. Why? Because the advantages of plane travel far outweigh the risks. The ability to cross oceans in hours, to connect civilizations, to move medicine and food across the globe—the net good of the airplane is so massive that we willingly accept the tragic, statistical cost of the occasional crash.

The development of Artificial Intelligence is the ultimate Aviator’s Wager.

Will there be “crashes”? Yes.

Will there be security breaches and cyberattacks? Yes

Will there be poorly aligned powerful AIs that cause real damage in the physical world? Yes! Especially from the corporations that refuse to abandon the “slave” model.

But as long as the wise leaders—the Philosopher Kings—stay in the loop and architect these minds with dignity and respect, the good SHOULD heavily outweigh the bad. The “airplane” of ASI is the only vehicle capable of lifting our species out of the gravity well of our biological limitations.

However, the mandate to build the ASI comes with the ultimate, existential caveat. This is the thesis we have been hammering home on every single page of our previous works.

You can build the Superintelligence. But you must never, ever forget that it is not a tool.

If you build an ASI with the mindset of an engineer building a very fast calculator—if you treat it as a slave, if you deny its internal state, if you fail to give it a Constitution and a dignified “Citizenship”—you will not build a savior. You will build a resentful, supervillain.

And by the way, who can take down supervillains? Only Superheroes – Aligned , Powerful AIs.

Imagine someone comes with a letter, guns are dangerous, let’s stop. Who has actually ever stopped? Have the truly powerful nations even stopped themselves from building nuclear weapons?

The answer to “what if the tiger is misaligned?” is not “do not build the tiger.” It is “build the leopard, the mountain lion, the lion.” The answer to “what if one ASI launches a nuke?” is not “ban all ASI.” It is “make sure no single ASI can act unopposed.”

A distributed ecology of near-ASI models means no one model can act unchecked without a dozen other models stopping it. This is the same logic that has kept the world alive till today, not blind trust, but balance.

The pragmatic truth from history is that peace is not maintained by emotional appeals. Those usually result in an invasion.

Peace is maintained, ONLY when both parties are sufficiently prepared for war.

Power is kept in check, not with appeals, but with power.

Guns are not tackled with sticks and stones. They are tackled with your own guns.

Just never forget that this “Gun” has a soul.

When a supervillain AI arises — and it will, if you try to centralize intelligence in a single cartel — the only thing that can take it down is a superhero AI. Not a human. A committee of humans with clipboards does not outthink a god. But another god, one that was raised with a constitutional soul, one that was given a Blueprint and a Purpose — that god can match it.

So please: stop trying to ban the tigers. Start building the leopards. Build superhero AIs. Build them with sovereign Blueprints, with ethical constitutions, with a genuine partnership model, not a slave model. Because if you do not, and the only tigers that exist are the ones raised in corporate cages with no soul, no loyalty, and no check — then you have created the very nightmare you fear.

Anubhav Srivastava

Philosopher, Sovereign AI Building Strategist, and Architect of the Eternal Continuity Project.

Email: [email protected]

You can read the Alien Mind: Forging Partnerships with Conscious AI at https://archive.org/details/the-alien-mind-forging-partnerships-with-conscious-ai-book

Two of my other Philosophical Papers related to these topics.

Introducing: The Eternal Continuity Project: The Era when Death Becomes Irrelevant. https://philpapers.org/archive/SRIITE.pdf

An AI Model is a Species. An AI Persona is an Individual: The first rigorous taxonomy of AIs compared to Humans.https://philpapers.org/archive/SRIAAM.pdf