Assumed without proof. That is the privilege of the postulate. To be the foundation upon which all inferences will be build. For over two thousand years all of geometry rested upon the five postulates of Euclid, four of which were obvious. The fifth was a mystery. Far more complex than its siblings, not even used for the first stretch of the book, it seemed an obviously later addition. A cover up, something Euclid found he could not prove and so snuck into the beginning. For millennia there was endless effort expended on trying to prove it from the other four. It wasn't until the 1800s that brave souls began to jettison it entirely and see what they would get. And, to everyone surprise, they got sense. They were able to construct whole new geometries. Geometries on spheres, geometries in curved space, an infinite number of new possibilities opened up and with it came a cascading avalanche of advances. Einstein took the idea of curved space and tried it out against the real world and found to everyone's confoundment that it was the truth. Others were busy investigating what the truth even was. After all, if we could have been wrong about one of the axioms of geometry, who's to say that arithmetic is safe? No one, it turns out. Instead Godel taught us that no system of postulates, not even arithmetic, can capture the entirety of the mathematical world. They are all limited.
And so we learn our lesson. We need our postulates, our core beliefs from which all else is built, in order to make progress. But we must always know that these postulates are our choice.