The problem arises from a solution to an exercise in Walter Rudin's Real & Complex Analysis.
The Problem: Let $X$ be an uncountable set, let $\mathfrak{M}$ be the collection of all sets $E\subset X$ such that either $E$ or $E^c$ is at most countable. Define $\mu(E)=0$ in the first case, $\mu(E)=1$ in the second. It can be shown that $\mathfrak{M}$ is a $\sigma$-algebra and $\mu$ is a measure on $\mathfrak{M}$.
Now, let $f: X\to\mathbb{R}$ be a measurable function. Since $X$ is uncountable, we have $\mu(X)=1$. For every $n\in\mathbb{Z}$, we know that $f^{-1}([n, n+1))=f^{-1}(\mathbb{R})-[f^{-1}((-\infty, n))\cup f^{-1}([n+1, +\infty))]$, thus $f^{-1}([n, n+1))\in\mathfrak{M}$. Let $E_n=f^{-1}([n, n+1))$, then either $\mu(E_n)=0$ or $\mu(E_n)=1$. Note $X=\bigcup_{n=-\infty}^{+\infty}E_n$, and $\{E_n\}$ is pairwise disjoint, thus $\mu(X)=\mu(\bigcup_{n=-\infty}^{+\infty}E_n)=\sum_{n=-\infty}^{+\infty}\mu(E_n)$. Then $\mu(f^{-1}([n, n+1)))=1$ for some $n\in\mathbb{N}$; WLOG, suppose $n=0$, i.e., $\mu(f^{-1}([0, 1)))=1$. If we write $[0, 1)=[0, \frac{1}{2})\cup[\frac{1}{2}, 1)$, then it is clear that either $\mu(f^{-1}([0, \frac{1}{2})))=1$ or $\mu(f^{-1}([\frac{1}{2}, 1)))=1$. It follows that there is a sequence of intervals $\{[a_n, b_n)\}$ such that $\mu(f^{-1}([a_n, b_n)))=1, 0\leq b_n-a_n\leq\frac{1}{2^n}$. In other words, it means that $\mu(f^{-1}(a))=1$ and $\mu(f^{-1}(b))=0$ for some $a\in\mathbb{R}$ and all other real numbers $b\neq a$.
My Question: Why do we have "In other words, it means that $\mu(f^{-1}(a))=1$ and $\mu(f^{-1}(b))=0$ for some $a\in\mathbb{R}$ and all other real numbers $b\neq a$"? Indeed, $\underset{n\to\infty}{\lim}b_n-a_n=0$; but how exactly do we find the $a\in\mathbb{R}$ such that $\mu(f^{-1}(a))=1$?