Normally, when you have two functions $f(n)$ and $g(n)$,
where for all $n \in \Bbb{Z^+}$, you have that $0 < f(n), g(n),$
and you establish a base case $n_0 \in \Bbb{Z^+}$ such that $f(n_0) < g(n_0),$
you then have two separate strategies that you might follow
to inductively prove that for all $n \in \Bbb{Z^+}$ such that $n \geq n_0,$
it is true that $f(n) < g(n)$.
The first strategy would be to examine the general ratios:
$\displaystyle R_n = \frac{f(n+1)}{f(n)}, ~S_n = \frac{g(n+1)}{g(n)}.$
If you can show that for all $n \in \Bbb{Z^+},$ such that $n \geq n_0$,
that $0 < R_n < S_n$, then you can reason as follows:
Suppose that $f(n) < g(n)$.
Then, you also have that $f(n+1) = R_n \times f(n)$
and $g(n+1) = S_n \times g(n)$.
Therefore, since $0 < f(n) < g(n)$ and $0 < R_n < S_n,$
then $R_n \times f(n) < S_n \times g(n).$
The second strategy is very similar to the first strategy, except that instead of focusing on ratios, you focus on differences.
That is,
let $F_n$ denote $f(n+1) - f(n)$ and
let $G_n$ denote $g(n+1) - g(n)$.
Suppose that you can show that for all $n \in \Bbb{Z^+}$
such that $n \geq n_0$, you have that
$F_n < G_n$.
Then, you can reason as follows:
Assume that $f(n) < g(n)$.
Then, $f(n+1) = F_n + f(n)$
and $g(n+1) = G_n + g(n)$.
Then, because $f(n) < g(n)$ and $F_n < G_n$
you have that $F_n + f(n) < G_n + g(n).$
The second strategy above is what the offered solution focused on, differences.
To prove:
For all $n \in \Bbb{Z^+}$ let $f(n) = 2n + 3$ and let $g(n) = 2^n.$
Then for all $n \in \Bbb{Z^+}$ such that $n \geq 4$
$f(n) < g(n)$.
First, the assertion was manually verified for $n = 4$.
Then, (in effect), the offered solution noted that
$F_n = f(n+1) - f(n) = [2(n + 1) + 3] - [2(n) + 3] = 2.$
$G_n = g(n+1) - g(n) = 2^{n+1} - 2^n = 2^n.$
Then, the offered solution noted that for $n \geq 4$, you have that:
$F_n < G_n$.
This conclusion allowed the second strategy above, that focuses on differences, to be invoked.