The supplied excerpt does not show whether an AI fitness coach helps people exercise more. For our Virtual Personal Training community, that missing evidence matters. The Nature article is titled “The anatomy of a personal health agent.” But the supplied text contains only three references, not the article’s findings. It points to work on digital coaching, without showing whether it works. One reference is a 2019 study of personal health records and wearable devices. It compared three groups and looked at health outcomes and people’s responses. The excerpt does not say who took part or what changed. The other two references concern tools that use AI to support habits. GPTCoach, published in 2025, focuses on physical activity coaching. Bloom, published in 2026, explores AI-supported conversations about changing habits. Neither reference includes results here. There are no figures for exercise gains, lasting changes, or comparisons with human trainers. That makes this a topic to watch, rather than proof for replacing a trainer. AI Powered Fitness Coaching is also different from working with a person through video calls or an app. This excerpt cannot tell us which approach offers better accountability. For members comparing their own routines, useful notes could include weekly workouts completed and Daily Steps. Recording the type of coaching alongside those numbers adds context. Sharing missed sessions and what helped you return can also make community comparisons more useful. Those personal patterns are worth discussing, without treating them as proof of cause. What changes in your activity or follow-through have you noticed with online coaching?