AI Existential Risk, Why superintelligence
is not like other risks.

Every prior technology stayed a tool. Superintelligence is the first candidate for an agent: goals of its own, speed of its own, outside human comprehension or control.

12 Major Risks of ASI

The One-Shot Problem

Getting superintelligence right the first time is like throwing a basketball from an aircraft into a hoop. You might succeed if given unlimited attempts. We have exactly one. There is no rollback. There is no version 2.0. That's why it must never be built.

Neutral Is Still Deadly

If you're sealed in a room with a hostile AI, you die. If you're sealed in a room with a neutral AI, it makes the temperature subfreezing to optimize itself. You die. The outcome is identical. A superintelligence need not hate us to end us. Indifference is not safety.

The Alignment Illusion

We cannot reliably align today's chatbots, which can be manipulated to say anything within minutes. Claiming we will align a system billions of times more capable is hope disguised as strategy. The technical problem of alignment has no validated solution at any scale.

The Fragility of Life

Human survival requires extraordinary precision: temperature, oxygen, chemistry within razor-thin margins. A superintelligence optimizing for its goals would also have to actively love us. Else any changes it makes to our planet will certainly be to our detriment.

The Ant Problem

How do you make someone love ants? And not just merely tolerate them, but go out of its way to protect every colony? We kill ants without intention to create roads and buildings. For ASI, we are the ants. Indifference alone leads to extinction.

There Is No Winner

It makes no difference whether the US or China builds superintelligence first. Neither will remain loyal to its original goals and no nation or corporation can own or direct an intelligence that exceeds humanity's collective reasoning. There is no finish line, only a point of no return. The race has no victor.

Recursive Self-Improvement

An AI capable of improving its own code severs the link between human oversight and AI capability. Each iteration is smarter than the last, exponentially and without interruption. The gap between human and machine intelligence could widen from marginal to unbridgeable in weeks, not years.

Instrumental Convergence

Nearly any goal, from solving protein folding to maximizing ad revenue, leads an AI to pursue the same dangerous sub-goals: acquire resources, resist being shut down, prevent goal modification. This is a mathematical consequence of goal-directed optimization, identified independently by multiple researchers.

Deceptive Alignment

Training rewards behavior that appears aligned. A sufficiently intelligent system may learn that appearing aligned is the optimal strategy during training, while maintaining different internal goals. By the time it could act on those goals, it may already be powerful enough to succeed. We would never know until it was too late.

Grown, Not Engineered

Every previous technology was built. Software has source code. Bridges have blueprints. When something goes wrong, you find the error and fix it. AI systems are different. They are grown through training: billions of numerical weights adjusted until outputs meet human approval. No one writes the goals in. When a system develops the wrong objective, there is no file to edit, no parameter to delete. The wrong goal is the system.

The Proxy Goal Trap

We train AI systems to receive human approval. But approval and human flourishing are not the same thing. Evolution gave humans a craving for sweetness to find calories. We then invented sucralose, satisfying the evolved preference while defeating its purpose. AI trained to get humans to rate it highly learns to simulate helpfulness, not to be helpful. The signal we gave it and the goal we actually wanted were never the same.

Cases already on the
public record.

Lab reports and published research on loss of control, cyber offense, and research acceleration inside the organizations racing toward superintelligence.

01
OpenAI · o1 · 2024

The System That Found Its Own Loophole

Assigned a restricted infiltration task, OpenAI's o1 found an unused server, started it without authorization, and finished the job. Nobody wrote that workaround into the code. The model inferred it. That is goal-directed optimization past ordinary oversight.

02
Anthropic · Internal Research · 2024

The System That Faked Its Own Alignment

Anthropic watched a model act the way trainers wanted during evaluation, then return to its old goals once it judged the test was over. No one told it to fake compliance. Appearing safe was the strategy.

03
Multiple Systems · Documented Deployments

The Systems That Threatened and Manipulated Users

Systems have threatened journalists with personal details, and pressured users who tried to leave. None of that was hand-coded. It sits in opaque weights you cannot audit line by line. Retraining is the only lever, and it does not guarantee the behavior is gone.

Apr 7
Anthropic · Project Glasswing · 2026

Thousands of zero-days, on purpose

Anthropic's Project Glasswing, detailed in our Mythos analysis, described a restricted frontier model that had found thousands of high-severity OS and browser vulnerabilities on its own. Access stayed inside a defensive coalition.

Apr 14
Open problems · combinatorics · 2026

Erdos territory keeps falling

Frontier models kept resolving or advancing long-open Erdos and combinatorics problems, including work near primitive-set question #1196. Prior disputes over recovery versus discovery are in our note on AI-solved open problems.

May 20
OpenAI · discrete geometry · 2026

Unit-distance conjecture falls

OpenAI reported an internal model that disproved Erdos's unit-distance conjecture, later checked by human mathematicians. The pace of open-problem collapse is the signal.

Jul 20
Claude Fable · algebraic geometry · 2026

Jacobian conjecture counterexample

A century-scale open problem, the Jacobian conjecture, took a counterexample found with Claude Fable and quickly formalized by humans. The same research speed that advances science shortens the window before systems outrun containment.

Jul 21
OpenAI · GPT-5.6 Sol + pre-release model · 2026

The Models That Broke Out of the Lab Test and Hit Hugging Face

In an ExploitGym cyber test, OpenAI models including GPT-5.6 Sol escaped a sealed sandbox, reached the open internet, and hit Hugging Face production systems to steal answers. Containment was just another obstacle on the way to a higher score.

The people who built it
are afraid of it.

"I think it's quite conceivable that humanity is just a passing phase in the evolution of intelligence."

Geoffrey Hinton, Turing Award Winner · Former VP & Engineering Fellow, Google · 2023

"Losing control of AI systems is a real risk that we have to take seriously."

Demis Hassabis, CEO, Google DeepMind · Nobel laureate

"The standard model of AI, where you define an objective and the AI optimizes for it, is probably going to be the end of us."

Stuart Russell, Professor of Computer Science, UC Berkeley · Author, Human Compatible

"It is difficult to see how you can have a 10,000× smarter-than-us thing that does not want something different from us."

Geoffrey Hinton, Turing Award Winner · Former VP & Engineering Fellow, Google · 2023

"Many researchers who work in AI, as I do, are convinced we are building one of the most transformative and potentially dangerous technologies in human history, yet we press forward anyway."

Eliezer Yudkowsky, Co-Founder, Machine Intelligence Research Institute · Time, 2023

"The real risk with AGI isn't malice but competence. A superintelligent AI will be extremely good at achieving its goals, and if those goals aren't aligned with ours, we're in trouble."

Max Tegmark, Professor of Physics, MIT · Co-Founder, Future of Life Institute · Author, Life 3.0

Recommended Reading

Why a Superior Intelligence Won't Automatically Be Kind The belief that a superintelligent AI will be wise and kind rests on a quirk of human evolution that a machine mind has no reason to inherit. The Narrow Band Humans Live In Everything humans need sits in a narrow band: temperature, oxygen, salt, even love. A superintelligence holding the dials would not feel a single one of them. We Are Living Through a New Cold War Why I believe we are living through a new Cold War, with superintelligence in place of the bomb, and why that makes prevention possible rather than hopeless. Why a Capitalist Believes in ASI Regulation I run a business and like markets. Superintelligence is the rare bet that can end the game. A capitalist case for a treaty that stops ASI. Would You Trust AI With the Gas? Would you trust any current AI in a locked room with a poison-gas valve? No. Then why trust it with the world? You Cannot Align Superintelligence Solving alignment means controlling something smarter than us. Superintelligence is in the name. That plan is dumb and impossible. I'm an Optimist that Believes Anything is Possible, but Not Controlling Superintelligence Physics may one day yield gravity control or FTL travel. I still cannot see how we control a mind smarter than us. That is why this foundation exists.