Intuition behind Probability Integral Transformation
If $F$ were invertible with inverse function $F^{-1}:(0,1)\to\mathbb R$, the event $[Y\leqslant y]=[F(X)\leqslant y]$ would coincide with $[X\leqslant z]$ with $z=F^{-1}(y)$, hence its probability would be $F(z)=y$. A random variable such that $\mathbb P(Y\leqslant y)=y$ for every $y$ in $(0,1)$ is uniformly distributed on $(0,1)$.