My research examines how humans and AI systems can collaborate in data-driven decision-making, leading to human-machine hybrid intelligence that augments and builds upon the capabilities of each. Following a design science research methodology, I focus on how to design and implement such systems in various contexts and settings. Furthermore, I aim to extend this research to develop a framework for Provably Beneficial Hybrid Intelligence (Provicial HI), extending Russell's (2019) notion of provably beneficial AI to socio-technical systems in which intelligence emerges from human-machine collaboration. Some of the main research questions include: (1) What core elements and interactions define provably beneficial hybrid intelligence? (2) How can hybrid intelligence systems be designed to remain transparent, adaptive, corrigible, and aligned with human goals and values across changing contexts? (3) How does hybrid intelligence evolve over time through collaboration, feedback, and knowledge refinement?