Explore the challenges and opportunities of diarization in real-world settings through this informative talk by Alex Cristia from CNRS. Delve into the potential of daylong recordings from wearable devices to provide unique insights into spontaneous speech use. Learn how these recordings can offer ecological measures and facilitate the study of individual differences in language use. Discover applications in early language acquisition, seniors' well-being, child development assessment, patient recovery monitoring, and intervention effectiveness evaluation. Examine the technical, ethical, and legal considerations surrounding such recordings, with a particular focus on automatic event labeling (diarization). Gain insights from Cristia's diverse background in linguistics, psycholinguistics, and neurobiology of language as she discusses her research on linguistic representations, their formation, and the influence of learnability biases on world languages.