This looks promising, but the inequality we are trying to prove has n−2 instead of n in the exponent on the left hand side.
To fix this, there are (at least) two approaches we can take.
Approach 1: Divide through by (1+1/n)2
The most direct approach is to simply divide both sides of our inequality by (1+1/n)2.
After this, the left hand side is what we want, but the right hand side is a bit of a mess: [1+n1]n−2≥(1+n1)21+n+2n−1.
To clean it up, we simplify the numerator and denominator of the right hand side. The numerator can be written over a common denominator to get
1+n+2n−1=21+2n+n=21(1+n)2.
(For the final step here, we noticed that 1+2n+n is the square of 1+n.)
Similarly, the denominator can be expanded to get (1+n1)2=(nn+1)2=n1(1+n)2.
Substituting these back in we obtain
[1+n1]n−2≥n1(1+n)221(1+n)2=21n.
Approach 2: Manipulate the right hand side to find hidden factors of (1+1/n).
In this less direct (but essentially equivalent) approach we manipulate the right hand side by looking for factors of (1+1/n) that we can cancel from both sides.
This still isn’t quite what we want — on the right hand side we have
(1+n)
, when we want
(1+1/n)
— but we can fix this by noting that
(1+n)=n(1+1/n)
. Thus
[1+n1]n≥21(1+n)2=21n[1+n1]2,
which, after dividing through by (1+1/n)2, implies that [1+n1]n−2≥21n, as desired.
Is it important that n>1?
You might like to check if the inequality is true at n=1, or n=0. You might then consider what would happen if n was allowed to take non-integer real values.
Where do issues arise?
Below we draw the two curves y=(1+1/x)x−2 (in green), and y=x/2 (in blue). We can see that for x>2 (but not for 0<x<2) the green curve is above the blue one.