When doing the regression model:

when age squared has a negative coefficient and the p value is significant, it proves that they are non-linear.

when people getting older, then age is not a premium.

But what does it mean if I get a negative coefficient for age squared but p-value is insignificant?