The work aims to address the dilemmas of inheritance and development within the Batik art, analyze the current deficiencies in the integration of AI technology with Batik art, and resolve the challenges faced by AI technology in the inheritance and development of Batik art. During a three-month fieldwork in Malaysia, in-depth interviews were conducted with local Batik artists, and the creation experience of our new media artwork "Neural Batik" was analyzed. As a result, "Neural Batik", a machine learning algorithm-based convolutional neural network painting system, was developed, enabling users to transform offline paintings into online artworks with traditional Batik styles in real time, visualizing the thoughts of experiencers and presents them in a digital interface, fostering interactions between artificial intelligence, generative art, and traditional Batik crafts. This innovation not only provides diversified digital expression methods for Batik art, and ensures the inheritance and development of ancient skills, but also transcends communication barriers imposed by time and space, and enhances communication and exchange between different cultures with Batik as a medium.