Variability in Large Language Models' Responses to Medical Licensing and Certification Examinations. Comment on "How Does ChatGPT Perform on the United States Medical Licensing Examination? The Implications of Large Language Models for Medical Education and Knowledge Assessment".