Question $3\mathrm{.}1$: $[4$ points$]$

Assume that the following function has been proposed to you to use as a new loss function for regression:

$L_3(hat())={\displaystyle \underset{i=1}{\overset{n}{}}}(y_i-$hat$(y{)}_i{)}^3$

What do you think about this new loss function, would you accept it$?$ Please, explain.