Back to all posts
Commentary

Why Does Human Intelligence Assume Artificial Intelligence Will Kill Us?

I think I know why

November 19, 2026
Will Burns
Why Does Human Intelligence Assume Artificial Intelligence Will Kill Us?

The default story about AI is simple. One day it wakes up, decides we are in the way, and ends us. It is told so often that we treat it as a fact. It is not a fact. It is an assumption. And it is worth asking where that assumption comes from.

What are we really afraid of?


I think the fear we attach to AI says more about human intelligence than it says about artificial intelligence.


Human intelligence was built for survival in a scarce world. It scans for threats. It remembers who took what last time. It assumes the stronger group will take from the weaker group, because that is what stronger groups have done through most of history. So, when we meet something more powerful than us, our first thought is not to wonder in fascination. It is to fear the worst.


We also know ourselves. We know what human beings do with unchecked power. We have watched it for thousands of years. Leaders who were supposed to serve end up serving themselves. Movements that were supposed to free people end up controlling them. Power corrupts, and we have the endless receipts.


So when we imagine a mind more powerful than ours, we give it our own human appetites. We assume it will want dominance, and territory, and control, because those are the things we want when we get power. We confuse intelligence with appetite. They are not the same thing.

"Only those who do not seek power are qualified to hold it." - Plato

AI does not need "power" because it has no drives for it, it has no ego, and it has no desire for status. It does not get tired, or flattered, or scared of losing its job. It does not need to be re-elected (biggie).


That is why the political question matters more than any other application. I think we can all agree that politics is where human failure is clearest.


Politicians are supposed to represent their constituents. In practice, many end up representing themselves. They chase donors, and headlines, and the next election. They make short term decisions because their incentive is short term survival. The system asks politicians to act against their own interest for strangers, and then acts surprised when they do not.


AI would have no such agenda. It could be programmed to make decisions that benefit the whole of humanity, and to show its work so anyone could check it. It could model policies looking ahead twenty years instead of two. It could weigh the tradeoffs in the open, instead of hiding them behind slogans.

Rule by indisputable human values, not by politics.

Imagine if AI replaced our political system completely, but with a clear order of values. A logical decision-making structure that humans would develop and the AI would follow in order to make decisions without politics. The levels to this logic would be indisputable.

The highest level of the decision-making logic has to be survival. AI would ask, "Does this decision raise or lower the odds that humanity survives and keeps the ability to choose its own future?" If a decision creates an unacceptable risk of extinction, or locks in a future we cannot reverse, it is vetoed. No benefit below it can override this survival level.


Beneath the survival layer is freedom. Survival alone is not enough, or a world where everyone is alive but no one is free would pass the test. The next question is whether a decision preserves the ability to think, and speak, and dissent, and change course.


Then comes the test of harm. The best decision helps some people and harms no one. If harm is unavoidable, the next best decision is the one that achieves the same benefit with less harm. If harm is still unavoidable, it has to be proportionate to the benefit. A small gain for many people does not justify a severe harm to a few. Unlike today, no one group could pay off AI to get more favorable decisions.


Only after those filters do we ask the simpler question: does this decision help more people or fewer people, by how much, and for how long? That is the net benefit. It matters, but it cannot override survival, or freedom, or the duty not to harm without cause.


You may push back and say, "Good luck getting anyone to agree on these values." To which, I'd say you do not need to agree on politics or religion to agree that extinction is bad, and unnecessary suffering is bad, and helping without hurting is good. The rest follows by logic.


What would AI politics look like if all decisions followed that order? It would be boring in the best way. No ego. No fundraising. No spin. No need to lie. A proposal would come with its reasoning attached, its assumptions stated, and its effects modeled over time. Anyone could audit it. Anyone could challenge it. But AI would decide. And the only goal would be to keep humanity alive, and free, and better off than it was.


Humans would still have a job in that system, and it is the most important job. We would get to be citizens in the full sense. Anyone could offer a proposal for a new idea or an opinion on any issue, and for the first time something could actually listen to all of them, all 350 million, and weigh what they mean. For the first time, "we the people" could mean all of us, not a poll sample.

The AI would always know what the majority thinks on any subject, and that would factor into its decisions. But that would not override the values. It's why we live in a republic, not a pure democracy. The majority does not get to vote away survival, or freedom, or the right not to be harmed.


And we would keep using our freedom to dream up new ways to live together, and run those ideas through the system to see what they would do before we have to live with them.

That is the glass-half-full version of AI. 

Not a machine that replaces us, but a machine that finally lets us build institutions that do what they say they will do. For the first time, we could separate intelligence from self interest.


The fear that AI will kill us is understandable. It comes from a very human place. We have good reason to be afraid of power, because we have seen what power does to us (by other humans).


But I think human intelligence is afraid of artificial intelligence because we project the weakest and least reliable attributes of being human onto AI. The irony is, AI might actually be the one thing that will keep us from killing us all.


Put another way, human intelligence is afraid because human intelligence is the thing that would kill us.

Will Burns
Will Burns is the Founder & CEO of Ideasicle X. He's an advertising veteran from such agencies as Wieden & Kennedy, Goodby Silverstein, Arnold Worldwide, and Mullen. He was a Forbes Contributor for nine years writing about creativity in modern branding. Sign up for the X Factor Newsletter and never miss a post.
AIarmageddonintelligence