Meta has filed a patent for a system that continuously records a user's voice and surroundings throughout the day. This system would then utilize an AI to analyze the user's mood. The purported goal is to sell users a wearable device that personalizes workouts based on their emotional state. The patent, published in July after being filed in December 2025, describes an apparatus designed for constant surveillance to enhance workout regimens. It categorizes audible communications and associates them with contextual factors like time, location, and activity. The AI would transcribe these communications and employ machine learning to interpret verbal and nonverbal cues for emotional indicators. To accurately gauge moods, the AI requires data beyond spoken words, including sighs, laughter, physical location, surrounding objects, and even medication intake. The system aims for precise and reliable emotional inference by aligning multimodal sensor inputs on synchronized timelines. This creates a novel data structure for richer emotional analysis, enabling continuous emotional monitoring. The patent explains that humans are less precise than machines in providing corrective workout guidance, necessitating such a technological approach. However, Meta states that patents are often filed to disclose concepts that may or may not be implemented and do not guarantee pursuit of the technology.