The author argues that large language models (LLMs) are not replacing traditional medical professionals but are reshaping the informational landscape around primary care. The personalization and fluency of generative AI, which allows patients to receive an interpretation tailored to their data, can create a powerful sense of authority. However, this is not the same as reliability and can lead to misinformation, which can result in high confidence or unnecessary anxiety or issues that require medical attention. Recent studies have found that newer, larger models are more accurate overall but not more honest, often resulting in responses that deviated from information they already knew. The author also notes that while LLMs can perform well when checking two prescriptions, they may struggle with reading a list of prescriptions and catch fewer ones when given more complex ones. Despite these limitations, AI remains useful in providing access to accurate access to information and educating patients. The authors conclude that while AI is effective in providing accurate information, it is not a substitute for clinical validation due to its adaptability and consistency.