Last week I shared a few thoughts on why I'm not a fan of having AI write anything that's client- or community-facing.
To be clear, I'm not anti-AI.
Here are 5 beginner-friendly scenarios where I've found AI to be a helpful tool in my work just this week:
- Tech support: AI is very good at this. This week, I've been working on a new website and needed to make some changes to the DNS records. I asked AI to have a look at the changes to tell me if anything was off (it was) and suggest a fix (which worked). I can also ask AI to explain DNS records to me me like I'm 10 (why do I need them, what do they do, what is this type of record v. that type, etc), which helped to wrap my head around what I was doing.
- Actual prompt: Can you look at this page and tell me if I have SPF and DKIM records set up for the domain xxx please?
- Brainstorming partner: As some of you know, I am writing a book about writing newsletters. I have been asking AI to help me with titles/subtitles. I wound up thinking of the title on my own (stay tuned!), but it's been helpful to generate ideas for subtitles. Even if I hate them, having something to work with helps get the juices flowing.
- Actual prompt: I'm starting a paid email newsletter for people who are trying to build their 1:1 clinical work by writing newsletters. I want to highlight a few themes in the subscription: finding your own way to write a newsletter (it doesn't have to be formal, long, referenced, etc., it just has to sound like you), the importance of sticking with it and writing, even when you don't 100% love the results, and using your newsletter as a way to serve your community, even when they can't work with you 1:1. Do you have ideas for a subtitle or tagline I could use?
- Summarize a YouTube video: If I'm looking for a video on a certain topic, AI will summarize the video so that I can see if it's worth it to watch the whole thing.
- Actual prompt: Summarize this YouTube video please.
- Answer quick questions while I'm working on a case: I don't recommend relying only on AI exclusively to do your research, but if you quickly need to look something up while working on a case, AI can be very helpful. Examples: Why might someone be taking XXX medication? Does this medication have any known association with nutrient depletions? What is the stadard medical treatment for XXX? Note that AI tools can gather information from questionable sources, so please double-check/verify what you learn.
- Actual prompt: how much biotin would one need to be taking for it to interfere with a thyroid lab?
- Summarize a consultation/next steps: If you are using Zoom, the new AI summary feature can be handy and does include a bulleted list of action items from the consult that is mostly accurate. It is free and HIPAA-compliant if you have a BAA/HIPAA-compliant Zoom account to start with. I have been using this in some of my 1:1 mentoring sessions (if the client agrees & I remember) and do find the summary to be relatively useful. I haven't started using it for nutrition/herbal clients because I haven't had a chance to update my informed consent form yet, but probably will do this eventually. I would never use this in place of my own notes, and it's a helpful to jog my memory!
- No prompt needed if you turn this on during the visit.
One of the big advantages of using AI is that you can ask follow-up questions and continue the conversation after the initial prompt as well.
I use a tool called Voila for everyday questions, Elicit for academic research help, as well as the built-in AI summaries from Zoom, but you can play around with PerplexityAI, ChatGPT or use Microsoft's CoPilot either as a standalone or in the Edge browser. All have free tiers.
I'd love to hear your thoughts on AI and how/if you're using it in your own work.
Take care,
Camille
