Is anyone else using AI like ChatGPT

If I were asking a question about my lab results

‘Have prostate cancer got these results, What do they mean for me? Taking current drugs ….

In a query you could include a list of all of the drugs you’re taking Should be better than talking to a pharmacist about lab results. It could do a synopsis of all the side effects Compared to the lab results.

You are spot on, I upload my blood work monthly and it shows areas or issues I can talk to my team about! I have been fighting bad fatigue for the past couple of months and Chat gave me talking points.

If it helps then use it. You just have to remember that it's fairly subjective and not private in the least, you are lending your data to the model. That may be of no concern, but don't let yourself get too comfortable with the answers, they aren't always right.

I use AI quite a bit and I actually spend a bit of time beforehand typing the most comprehensive query I can, then save the project and refine as needed. But even with all of that, I still verify what's puked out at me because I've been burned more times than I can count!

"Conversation", I think, would be most accurate, but ChatGPT also keeps a general cross-conversation memory of things you've told it (not as detailed as in specific conversations).

There actually is a job called a prompt engineer for just the purpose @jeffmarc is describing. Every AI tool has its strengths and weaknesses and as these tools start to identify the sources and score their database validity, the trust for the answers will increase but generally ai is a great tool to collaborate with and get some medical insight especially for those that don't know about Mayo connect or have doctors who do not have the time or experience to answer a slew of questions.

My husband uses ChatGPT for many work related things with astonishing results and he also used it for his diagnosis and possible treatment options for his particular case. It was interesting to see AI choosing our preferred method for treatment (RP) over and over again no matter what new parameters we added. It is probably because it is really the best way in his case with IDC and cribriform but doctors were suggesting that other methods could also give results. I guess they "could" but probably with some lesser degree of success. All data points to RP as the best approach for my husband's case, all things considered.

I always add new information, like studies I read or other variables, like history and other issues I've had, and for me has done a good job.