In the early 1800s, some textile workers in England opposed the use of certain automated machinery in their work. They destroyed machines in organized raids and were eventually defeated by the powers that be. They called themselves “Luddites” after a mythical character named Ned Ludd. I am a luddite. I did not have a phone with a screen until well after they became common. I have never owned a tablet. I am generally skeptical of new technology.
When new technology is presented, there is a curve of adoption. The innovators are the ones who create the technology and work through the initial bugs. The early adopters are technologically savvy people who want the best and new stuff- think of the people who line up for the latest iPhone. Once a piece of technology has been proven, the early majority are a large bulk of the population who begin to use it and the late majority lag behind due to finances, opportunity, or lack of interest. Finally, the laggards are the ones who resist the technology and only adopt it when it’s necessary for their functioning in society.
To be honest, on some things I am an early adopter. I had a bulletin board system (BBS) that I was the sysadmin for in the early 90s (this was an early version of the internet, very cutting edge at the time). I got on Facebook and Gmail as soon as I could in the early 2000s. I am an Excel and Word super user. We bought an EV last year when less than 8% of cars sold were EVs. I have this love-hate relationship with technology. So, when AI came out, my initial response was “Meh”.
Being somewhat into technology growing up, I was always fascinated with the Turing Test and sci-fi depictions of the future. When ChatGPT came to prominence in 2022 and I thought, “Dang, this thing can absolutely pass the Turing Test”, I was impressed. And then I ignored it.
I have always said, “A smart monkey and a good computer program can do my job as a clinician”. Seriously. Medicine is just a huge list of if-then statements. If you had the patience to program it, you could absolutely put every variable about a patient and situation and come up with a decision tree for every circumstance. I do this on a micro scale for the students- I trace out flow charts for common complications such as hypotension and a vocalizing dog post-op. Once I saw how well ChatGPT did, I thought, “That will definitely replace us some time in the future.”
Fortunately, that future is not tomorrow, and maybe not even in 10 years. It would require someone being very intentional to give an AI enough information to make good clinical decisions. I’m sure it will happen, but I think “veterinary anesthesia” is a pretty low-interest domain to be working on. By extension, I think veterinary medicine is a relatively low priority. They might have decision-helping AI in human medicine relatively soon, but given the conditions and species we have to deal with, I think it will be a while for us.
However, that was all speculation until I decided to play with it. My university has a deal with Microsoft Copilot such that the data is kept on a separate server and isn’t used to further Copilot’s algorithms. One of my concerns about AI had been how the data I enter is used, and also I didn’t want to give them my email address. So this summer I opened up Copilot and played with it.
Someone suggested AI can be used to plan travel itineraries. I asked it to create a 5-day itinerary in London for someone interested in breweries, Neil Gaiman, and the Rivers of London books and didn’t like to visit typical tourist locations. What it came up with was… fine. Nothing that I couldn’t have done with a little bit of searching with Google Maps (my usual way to create an itinerary) and also nothing particularly interesting or remarkable. It didn’t give me any sites related to Neverwhere or the Rivers of London series, for example. It was fairly generic.
Another suggestion I got was using AI to start a letter of recommendation. Although this may feel inauthentic, in reality most of us who write letters pull up an old letter we wrote for someone else, rework the details, and call it good. Is that so different from starting with an AI template? I asked Copilot to write a five-paragraph letter of recommendation for a student with such and such characteristics. What I got was, again, basically fine. It used a fair number of catchy words and phrases, which is fine, but not my preferred writing style. I had to rework a fair amount of it.
Finally, I was asked to write a letter of support for a grant application. I don’t do grant applications. So I went to Copilot and asked it to write such a letter, and gave it the organization, the name of the project, the discipline, etc. It produced a letter that was probably decent. I read it, then went and wrote my own. It gave me a good starting point. It gave me some ideas to use as a jumping off point. But I definitely did not want to use it even as a template- it was just too sterile and, again, used a lot of catchy phrases.
I also asked Copilot to make some novel images with a variety of prompts. These were fine, but there was always something not quite what I wanted in them. But for creating images to use in lectures, rather than searching for hours on Google, it might be a good alternative.
Ultimately, I was pretty unimpressed with Copilot. It certainly wasn’t magic. Yes, it could pass the Turing Test. But that didn’t mean the output was usable off-the-shelf for me. It’s likely that the prompts I gave it were not up to the task to produce what I really wanted- prompt writing is its own separate (and difficult!) skill. I find using AI can be good for inspiration or to build a template. But it will definitely not replace my default method of writing.
Hello! I am a Junior high school student looking to pursue a career as a Veterinarian. I found your blog, and have really enjoyed reading your articles. They’re very helpful, insightful, and specific! Thank you for writing them, and putting them out for us to read. I was wondering if I could ask for some advice: Since my goal is veterinary care, I figured I had to get involved in the medical field in some way during high school, so I joined Nursing 1. In this class, we are taught basic human anatomy, medical terms and conditions, and how to take vital signs. Students taking Nursing 1 had the option to apply for the limitedly available Nursing 2 class, and I figured that since nursing and veterinary medicine are both healthcare related, it would be useful. This class would take up 1 semester of my senior year, including around 8 weeks of clinicals at a local nursing home or rehabilitation center. The benefits include: many hours of experience and a CNA license. Could this be beneficial to me, going into veterinary school? Is a CNA licence at all useful in veterinary medicine? How similar are the pre-reqs for nursing and veterinary? Would I be better off looking into other ways to gain experience? Ultimately, nursing is not what I want to do with my life- but I feel it could be a decent way to make the money I’ll need for veterinary school and other parts of life. I’m just unsure if it’ll be worth it- as it will take up a lot of time that could be spent taking pre-reqs, doing volunteer or internship work, and also conflicts with my music class, which is something I really love. I would really appreciate some advice on this matter!
Thank you!
Glad you have found them to be helpful! The terminology and a lot of the anatomy and general medical knowledge is very useful and applicable. So getting _some_ of that may be helpful. However, for vet school they really want to see that you have time in _veterinary medicine_. If you have the time to spend both learning human nursing and spending time in a vet practice, that’s fine. But if you only have time for one or the other, I would spend time in a vet clinic.