Pushing towards the top: the idea that gating networks are just an 'evolution' of basic prediction methods in venture capital underestimates their ability to generate maximum return. The opposition is logical because a gating network seeks information asymmetry by combining signals from basic methods in a nonlinear way, where simple juxtaposition does not allow. A basic model, like an LLM, can excel in textual analysis, but gating aims to identify synergies and redundancies for superior predictive value. For example, a gating could dynamically assign more weight to market signals in a volatile sector like biotech, thus surpassing the performance of a static model and reaching the upper limit of prediction.