The automation of chemical synthesis presents opportunities to enhance experimental reproducibility and accelerate discovery. Traditional closed-loop approaches, while effective in specific domains, are often constrained by rigid workflows and the requirement for specialized expertise. Here, we introduce a chemical robotic explorer integrated with an artificial intelligence (AI) copilot to enable a more flexible and adaptive synthesis, simplifying the process from inspiration to experimentation. This modular platform uses a large language model (LLM) to map natural language synthetic descriptions to executable unit operations, including temperature control, stirring, liquid and solid handling, filtration, etc. By integrating AI-driven literature searches, real-time experimental design, conversational human-AI interaction, and feedback-based optimization, we demonstrate the capabilities of AI in successfully synthesizing 13 compounds across four distinct classes of inorganic materials: coordination complexes, metal-organic frameworks, nanoparticles, and polyoxometalates. Notably, this approach enabled the discovery of a previously unreported family of Mn-W polyoxometalate clusters, showing the potential of AI-enhanced robotics as a generalizable and adaptable platform for material innovation.