Due to an ill-depicting model of the convective layers below the photosphere and a misrepresentation of the coupling between convection and oscillations, a well-known drift appeared between observed and theoretical frequencies which grows towards high frequencies; the so-called surface effects. With the increasing observational capabilities at our disposal these days with Kepler and TESS, this constitutes the main limitation to accurate seismic probing of solar-like and red giants. In this talk, we will present a brand new and innovative approach that changes the current paradigm by going 3D. It consists in an original non-adiabatic 3D time-dependent convection model for asteroseismology that we are currently developing. To tackle the current issues, we aim at keeping the entire 3D structure of the astrophysical flow in this superficial layer to fully account for the nature of turbulence in our model via the use of advanced hydrodynamic simulation COBOLD and, solve, for the first time, the global non-adiabatic oscillation equations in this full 3D framework. We show the importance of working in a 3D space and how it allows us to stay as close as possible to reality. Finally, we discuss some preliminary results of the model: the impact of this 3D structure on modes frequencies and damping rates as well as, unprecedented, the computation of 3D eigenfunctions.