New research from the IMT School for Advanced Studies Lucca challenges the long-held belief that dreams are random neural noise. Published in Communications Psychology, the study suggests that dream content is a complex interplay of personal traits and shared life experiences. The investigation analyzed over 3,700 dream and waking experience reports from 287 participants aged 18 to 70. Over two weeks, subjects recorded their daily activities and sleep habits, while researchers collected detailed data on cognitive skills, personality profiles, and psychological states. To process this extensive dataset, the team employed advanced natural language processing (NLP) tools. This AI-driven approach allowed for a systematic analysis of the meaning and structure within dream descriptions. The findings reveal that dreams are not chaotic but reflect a dynamic reconstruction of reality. The brain does not merely replay waking life; instead, it reshapes familiar environments like workplaces or schools into vivid, immersive scenes that combine different elements in unexpected ways. This process indicates that dreams actively blend memories with imagined events to create new scenarios. Personality plays a significant role in dream quality. Participants with a higher tendency to mind-wander reported fragmented, rapidly changing dreams. Conversely, individuals who attribute meaning to their dreams experienced richer, more immersive environments. The study also examined the impact of major societal events. Data collected during the COVID-19 pandemic, including lockdown periods from Sapienza University of Rome, showed that dreams became more emotionally intense and frequently featured themes of restriction. As populations adapted to the crisis, these specific patterns gradually faded, demonstrating that dream content evolves alongside psychological adjustment to external changes. Valentina Elce, the lead researcher, emphasized that dreams are shaped by who we are and what we live through. By combining large-scale data with computational methods, the team identified patterns in dream content that were previously difficult to detect. The study also highlights the potential of AI in advancing dream research, noting that NLP models can interpret dream reports with accuracy comparable to human evaluators. This capability offers a scalable path to studying consciousness, memory, and mental health on a broader scale. The research was conducted at the IMT School for Advanced Studies Lucca in collaboration with experts from Sapienza University of Rome and the University of Camerino. It was supported by a grant from the BIAL Foundation and the TweakDreams ERC Starting Grant. These findings collectively suggest that the dream state is a sophisticated mental process deeply rooted in both individual characteristics and the broader context of human experience.