Excellent patient education!! learn by example! May I suggest that your next post is a guidance test for newbies in using AI for solving health problems
Dave, thank you for sharing Sue's story. I'm glad to hear she is doing better! I’d love to know how she prompted the AI—did she start with a Google search before using ChatGPT?
These stories are so important...Very grateful to you and the #PatientsUseAI community for gathering and sharing.
Can Susan share what prompts she used to reach the information that was most meaningful and actionable? Perhaps that can be a part of future posts to help others get a head start on optimal prompting (what didn't work, what ultimately did)?
I googled teeth cleaning & face drooping and this is the first thing that popped up. Glad they figured it, but AI wasn’t needed. Bigger issue is the ER not knowing straight away. I suspect she saw an intern or resident.
Separately AI is incredibly harmful to the environment. The amount of water it uses is staggering.
Erica, thank you for coming here to continue this, from LinkedIn, while I was on vacation. (I'm on the plane home, as I write this.)
You bring up two points. The first is that perhaps AI wasn't necessary in this situation - just googling might have sufficed. As you probably know, I've always been all for googling; I can't speak for why they didn't; or maybe they did and it wasn't enough. I don't know.
But It's not hard to find cases where mere googling isn't enough, as I imagine you'll agree.
Your second point is AI's energy consumption. For that I'll turn to Bart DeWitte of Hippo AI Foundation, which argues for true open-source AI. In May he told me about open-source tools he uses that consume immensely less power. I'll see if I can get him to come join this discussion.
Dave, your posts about AI are wonderful! I remember distinctly an interview on television with a physician, when chat GPT was starting to become "a thing." She was asked if she was afraid of AI. She said, confidently (smugly, actually!) "Oh no, AI is only correct 60% of the time." She left it at that, and the interviewer smiled in agreement with her, or, out of respect for her elevated position! I remember thinking, "This is not true. I know doctors who are correct 0% to 30% of the time, and don't care about their patients!" Again, Dave, thanks for showing the other side. PS, I use Chat GPT all the time, both for medical information and for other information. What do you think of perplexity.ai? A friend of mine prefers that to chat GPT.
Julia, I don't know how I missed this comment six weeks ago. We'd love to hear some examples of what prompts you've used! dave@epatientdave.com if you want.
I just started playing with Perplexity - haven't used it much. @Hugo Campos is the wizard of different LLMs in the #PatientsUseAI world - he literally uses one AI to create better prompts for him to feed to another one!
Excellent patient education!! learn by example! May I suggest that your next post is a guidance test for newbies in using AI for solving health problems
What do you mean by a guidance test? Was that a typo?
yes Dave! meant guidance text.
Some of us are thinking about that. (It doesn't exist yet.) You're welcome to send your thoughts on what it should contain!
Dave, thank you for sharing Sue's story. I'm glad to hear she is doing better! I’d love to know how she prompted the AI—did she start with a Google search before using ChatGPT?
P.s.: I hope she gets her $350 copay refunded.
These stories are so important...Very grateful to you and the #PatientsUseAI community for gathering and sharing.
Can Susan share what prompts she used to reach the information that was most meaningful and actionable? Perhaps that can be a part of future posts to help others get a head start on optimal prompting (what didn't work, what ultimately did)?
I know from side conversations that being newbies, they didn't do anything fancy. But I'll see if we can dig out what they did.
And I absolutely agree that it's time to start teaching each other how to do it!
I googled teeth cleaning & face drooping and this is the first thing that popped up. Glad they figured it, but AI wasn’t needed. Bigger issue is the ER not knowing straight away. I suspect she saw an intern or resident.
Separately AI is incredibly harmful to the environment. The amount of water it uses is staggering.
https://pubmed.ncbi.nlm.nih.gov/27939042/
Erica, thank you for coming here to continue this, from LinkedIn, while I was on vacation. (I'm on the plane home, as I write this.)
You bring up two points. The first is that perhaps AI wasn't necessary in this situation - just googling might have sufficed. As you probably know, I've always been all for googling; I can't speak for why they didn't; or maybe they did and it wasn't enough. I don't know.
But It's not hard to find cases where mere googling isn't enough, as I imagine you'll agree.
Your second point is AI's energy consumption. For that I'll turn to Bart DeWitte of Hippo AI Foundation, which argues for true open-source AI. In May he told me about open-source tools he uses that consume immensely less power. I'll see if I can get him to come join this discussion.
Dave, your posts about AI are wonderful! I remember distinctly an interview on television with a physician, when chat GPT was starting to become "a thing." She was asked if she was afraid of AI. She said, confidently (smugly, actually!) "Oh no, AI is only correct 60% of the time." She left it at that, and the interviewer smiled in agreement with her, or, out of respect for her elevated position! I remember thinking, "This is not true. I know doctors who are correct 0% to 30% of the time, and don't care about their patients!" Again, Dave, thanks for showing the other side. PS, I use Chat GPT all the time, both for medical information and for other information. What do you think of perplexity.ai? A friend of mine prefers that to chat GPT.
Julia, I don't know how I missed this comment six weeks ago. We'd love to hear some examples of what prompts you've used! dave@epatientdave.com if you want.
I just started playing with Perplexity - haven't used it much. @Hugo Campos is the wizard of different LLMs in the #PatientsUseAI world - he literally uses one AI to create better prompts for him to feed to another one!