May I add my view?
I dislike all formulations like "one input follows the other one". I think, such a view does not help too much because it is not correct.
In fact, the situation is as follows:
As long as the opamp (with negative feedback) is within its linear amplification range the voltage difference between the opamp`s positive and the negative input terminal
is always Vp-Vn=Vd=Vout/Ao with Ao=open-loop gain.
This formula applies, of course, for inverting and non-inverting operation of the amplifier.
In particular, for inverting operation the non-inv. input is grounded and the voltage at the inv. terminal never "follows" this zero potential.
Instead, according to the relation given above this voltage is Vn=-Vout/Ao and, thus, strictly follows Vout.