The presence of Artificial Intelligence (AI) is growing in all areas of life and has an increasing role in both students' learning and their future work lives. Learners themselves should be active participants when learning and working with AI, not algorithms, and to do that, they need to develop their self-regulated learning (SRL) skills to be able to be active, adaptive learners. I argue for a theory and empirically driven approach to human-AI collaboration. I discuss how multimodal learning process data can be used to understand core human learning mechanisms to build Hybrid Intelligence systems that augment rather than replace human intelligence, systems that leverage our strengths and compensate for our weaknesses. I claim that human SRL skills empowered by AI are needed for future society, together with Hybrid intelligence solutions for education.