The paper presents the conceptual background on which the research project Media Art Live Archive: Intelligent Interface for Interactive Mediation of Cultural Heritage (TL02000270) partly supported by the Technology Agency of the Czech Republic, is build on. The main goal of the project is to develop specially trained artificial neural networks, which will be programmed to serve as an innovative tool for research into the videoart, specifically the work of Steina and Woody Vašulka. In the text, the following questions are discussed: Can artificial neural networks serve as an epistemological tool providing a new perspective on the early videoart and the work of Vasulkas particularly? What known research methods can they simulate? And what can they bring methodologically inovative to the research of the videoart?