Hi guiys, another question for you :) Lurie define in higher algebra both free algebras and operadic Kan extensions, apparently as two different instances. The main thing one has to figure out to see the equivalence is the following: given a functor $A \to B$ of infinity operads and an element $b \in B$, one can do two things:
1. Take the relative $X \to \Delta^1 \times \Nerve(\Fin)$ and then compute $ A \times_{X} X_{/b}$;
2. Take $A \times_B B_{/b}$;
In the case of operadic extensions one uses the first version, and in the case of free algberas one takes the second. They seems to be equiv…