This is an old one, so bear with me.
A farmer, a biologist and a theoretical physicist meet to discuss how cows could give more milk.
The farmer thinks, and says we need to improve the food content, nutrition and living conditions of cows. Then they'll give more milk.
The biologist thinks, and says we could make a genetically engineered cow that would make much more milk.
The physicist thinks, pauses for a second, and then says....
"Assume the cow is a sphere......"
Ah, the power of assumptions! I remember plenty of physics classes in high school, where we had a bunch of problems to solve, and had to start off our solutions with reasonable assumptions.
Only, not all assumptions I made were reasonable. After all, isn't reasonable a subjective word? I distinctly remember knowing so little in my JEE entrance exam that I solved every single problem with mostly ridiculous assumptions. Assumptions can be powerful starting points to solve problems, but only when accompanied by sound reasoning. If you do solve a problem with an assumption, it is imperative to now go back and test your assumptions to see if they would hold true under varying conditions.