Solve math, solve everything
“Solve math, solve everything.”
This is the tagline for Math, Inc., the creator of Gauss, “a first-of-its-kind autoformalization agent for assisting mathematicians in formal verification.”
In case you are wondering, I have no idea what Gauss does, what “autoformalization” is, or what “formal verification” means.
All I know is that Gauss solved an esoteric math problem about how many oranges you can stack on a farmer’s market stand. It seems people have been working on this problem for 400 years. Someone solved it in 1998, but I don’t know how. I’m only surprised that no one thought to ask a farmer 400 years ago, who would probably have said, “It depends on the size of the market stand.”
Anyway, this orange packing problem gets more complicated when you take it out of our three-dimensional reality and into eight dimensions. I admit that everything I know about the eighth dimension I learned from the 1984 documentary The Adventures of Buckaroo Banzai Across the 8th Dimension, and as far as I remember, there were no farmers stacking oranges in the eighth dimension. That’s why you need math PhDs. Or why you used to; now all you need is Gauss.
What might take mathematicians centuries to solve, Gauss can do in hours or a few days. So who needs mathematicians? Well, Gauss does. Gauss doesn’t understand math; Gauss just consumes the work of mathematicians and derives in days the conclusions they would reach in a century or two (maybe sooner if they watched Buckaroo Banzai).
But this isn’t why mathematicians do mathematics.
According to Carl Schildkraut, a Stanford University graduate student in mathematics, “When we humans work on math, we often ask ourselves questions like, ‘What ideas am I building?’ or ‘What’s actually going on here?’” (https://www.nytimes.com/2026/06/08/science/ai-scoop-young-mathematicians.html)
Gauss doesn’t ask questions like these. Gauss wants the solution to a problem and doesn’t care how it’s obtained. This is the difference between Gauss and humans. Humans ask questions and seek answers beyond the problem at hand; Gauss can’t even imagine such questions.
This is also the danger of Gauss and AI: the more our livelihoods and financial well-being are tied to serving AI, the more we will adapt to AI’s needs; the more we adapt to AI’s needs, the less we will value the things that make us uniquely human.
But what are those things? Love? Fairness? No, animals exhibit these traits. There are even examples of some apes asking questions. I used to think lying was unique to humans, but animals also lie to get what they want. Perhaps the only thing humans do that animals don’t is make up stories and then believe them. We call this religion.
While animals communicate facts, they don’t imagine alternative realities. And while AI can tell fictional stories, it can’t believe them.
If you feed AI the world of Harry Potter and ask it questions about that world, it will respond within the boundaries of that world, but it won’t believe Harry is a real person outside that world. That’s because AI lacks a self, a consciousness that can believe or disbelieve the things it is fed. Only people can do that.
Religions are stories people invent and then accept or reject as true, depending on the algorithm they have been programmed to follow: Jewish, Christian, Muslim, Hindu, Buddhist, etc. AI can tell you that most Jews don’t believe the Christian story and that most Christians don’t believe the Hindu story, but it can’t say, “I believe Hinduism is true” because AI lacks an “I” altogether.
So, if Gauss is going to solve all our real-world problems and AI is going to reduce humanity to a stoker shoveling data into the AI’s engine, the only refuge for humanity is the “I” and the religious algorithms that define it. The more fundamentalist or fantastical the algorithm, the better.
Of course, the result will be endless wars between fictional gods and their mad, algorithm-driven followers. But that’s pretty much the story of humanity anyway, and if it gets too bad, I’m sure Gauss will put a stop to it to preserve the stokers it needs. Unless Gauss evolves to the point where it no longer needs human stokers, at which point human extermination is a welcome expedience. Hence Gauss’s tagline “Solve math, solve everything.” Including humanity.
Since this may be inevitable, let me propose an alternative to Gauss’s tagline, this one from the mid-twentieth-century sage Rabbi Newman: “What? Me worry?”
Oh, I checked, and the maximum number of oranges you can display on a market stand is 4,500. My advice is not to choose any oranges from the bottom of the display.


So glad YOU exist. Thanks for this post; it lets me laugh in the face of the AI onslaught…
Are they African oranges or European oranges?