Dear All,
As someone who’s used ChatGPT since its release — and developed advanced Python tools with its help — I want to share a word of caution, especially for colleagues in rural or resource-limited settings.
ChatGPT doesn’t lie, but it does try to please. If you’re not already familiar with the topic you’re asking about, it can be hard to judge the accuracy of its responses. You still need domain knowledge to use it safely and effectively, particularly in health care.
In my own case, I’d already developed a predictive algorithm from 100 million medical records over 16 years. I used ChatGPT to help refine this into a tool that can predict mental health outcomes based on early diagnoses—and compare diagnostic sequences between those with and without mental disorders. It’s still in beta and not public yet, but it shows what’s possible when you already know what you’re doing.
The key takeaway? ChatGPT can be a powerful assistant—but it’s not a substitute for expertise. Use it to extend your reach, not replace your judgment.
Happy to share more if you want to contact me directly.
Best
David
HIFA profile: David Cawthorpe is Adjunct Assistant Professor at the University of Calgary, Canada. His professional interests include: Human Development, Developmental Psychopathology, and Delivery of low bandwidth medical education curriculum. cawthord AT ucalgary.ca