Introducing GUIDE: Teaching AI to Learn Like Humans with Real-Time Feedback

**Turning AI Teaching Methods Upside Down:** Traditional AI training relies heavily on extensive datasets and simulations, often struggling with adaptability and fast decision-making in complex tasks. Researchers at Duke University and the Army Research Laboratory have developed a novel platform called GUIDE, which incorporates real-time human feedback to teach AI more effectively. **GUIDE** was showcased in a study where AI learned to play hide-and-seek, receiving continuous human feedback on its strategies. Unlike previous methods, GUIDE allows for nuanced inputs via a gradient scale, leading to a 30% improvement in the AI's performance after just 10 minutes of feedback. This innovation underscores the potential for more adaptive training methods and opens the door to exploring how individual differences in human trainers can influence AI learning. As a result, GUIDE highlights the potential for creating training frameworks that enhance both AI learning and human cognitive abilities, paving the way for future human-AI collaborations in dynamic environments.