3 ms·
Fermat's Last Theorem and the Principle of Duality
Let A,B,C,K be natural numbers. Let T signify true. Let F signify false. So we have ∀K(K≤2→∃A,B,C(A^K+B^K=C^K ))=T. But this is equivalent to ∀K(K>2∨∃A,B,C(A^K+B^K=C^K ))=T. By the principle of duality we then have ∀K(K>2∧∃A,B,C(A^K+B^K=C^K ))=F . Also consider the following. Again, let A,B,C,K be natural numbers. ∃K(K≤2∧∃A,B,C(A^K+B^K=C^K ))=T. By the principle of duality we then have ∃K(K≤2∨∃A,B,C(A^K+B^K=C^K ))=F. But this is equivalent to ∃K(K>2→∃A,B,C(A^K+B^K=C^K ))=F
- okokwhatever 3y agoAnd just with a single post you've demonstrated why Hacker news isn't the place to paste an equation. Anyway, interesting.
- Kabbalist 3y agoAre you able to read and understand it though and is it the correct use of the principle of duality?