We have enjoyed much success with target centric drug discovery as the leading approach in the past few decades. Target centric drug discovery starts with specific protein targets which are typically identified through diseases associated with human genetic changes such as mutations, or through genetic knock down or knock out methods, followed by additional functional validation. Despite successes, significant challenges remain: we often identify new protein targets that are deemed "undruggable"; also, for certain diseases, lack of clear understanding of underlying mechanisms prevents a target centric approach. Target agnostic cellular screening, as an alternative approach, has also been utilized in drug discovery to uncover unknown mechanisms behind a specific disease phenotype. While there have been noticeable successes, its application has been limited by the often perceived daunting process of mechanism-of-action deconvolution. Recently, a number of publications revealed examples of small molecules, identified from target agnostic cellular screenings, eliciting their function "chemically induced proximity". In many cases, the small molecules enable new protein-protein interactions that do not exist in the native cells, thus creating a "gain-of-function" effect that may not be recapitulated from knock down or knock out methods. In this perspective article, recent findings and their implications for drug discovery are discussed. Additionally, a general framework for target agnostic screening in the context of "chemically induced proximity" is proposed, aiming to maximize the efficiency of hit finding by balancing screening efforts with mechanism-of-action deconvolution, including key factors such as the nature of compounds for screening, assay format, and readout.