I have a patient who lives in a house with black mold. He had modest symptoms but wanted to know if they were related to the mold in his home.
I cultured his nose and some skin lesions he had. The skin scrapings showed nothing, but the nasal culture showed growth of penicillium.
Up-to-date, which I spend $500 a year of my own money on, is usually a reliable source of medical information. They stated that people with systemic penicillium infection have a 97% mortality rate or something crazy like that. There was no comment on what to do if you find it in a nasal culture. My patient is feeling fine. He did his own Internet search and came to the conclusion that penicillium is one of the common black molds. A regular Google search tells you that penicillium is harmless to people with intact immune systems.
Good thing I didn’t trust up-to-date blindly in this particular case.
I worry about over-reliance on computers (and especially not-yet-ready-for-prime-time AI) and under-reliance on common sense.











I used Chat GPT to find studies related to a topic I was investigating. It basically quoted studies that didn’t exist with authors that did. I emailed one of the authors of the alleged study she co-authored. “It wasn’t me” she said. I notice lately Chat GPT is now reluctant to quote studies, and it’s database is only current to 2021 anyway.