Clinical practitioners are all too familiar with Dr. Google, the online “expert” who diagnosed your canine patient with cancer, prompting the dog’s owner to question your pancreatitis diagnosis.
The good news is Dr. Google’s days appear to be numbered. The search engine’s “usefulness” as a diagnostic tool is coming to an end. The bad news is Dr. Google is being replaced by the superior ChatGPT, an artificial intelligence (AI) language model trained on vast amounts of data and designed to give human-like text responses to user questions.
Maybe the rise of ChatGPT is good news after all. Or maybe it’s good and bad—it depends on what your opinions are about AI.
No matter what you think, veterinary medicine is no different from every other industry using AI in varying degrees to improve outcomes and efficiencies. In fact, veterinary medicine is poised to be inundated with AI technologies, according to Shawn Wilkie, founder and CEO of Talkatoo, a Canadian-based tech company that makes AI dictation software for veterinary practices.
“There are, like, 15 major venture capital firms that only want to invest in companies in veterinary medicine, and there’s more coming on board all the time,” said Wilkie, who spoke about AI’s transformative potential in two presentations, “The Transformation of Vet Med: An Introduction to AI and Its Role in Veterinary Medicine” and “Integrating ChatGPT into Veterinary Clinical Practice” on Saturday morning at AVMA Convention 2024 in Austin, Texas.
Wilkie has founded six startups over the past 20 years and co-hosts the Veterinary Innovation Podcast. Much of his presentations focused on ChatGPT. He explained that ChatGPT is a large language model, or an advanced AI model trained on massive amounts of text data to learn patterns, grammar, and context of human language, otherwise known as generative AI.
He recounted the story of Sassy, a dog whose life was allegedly saved by ChatGPT according to a series of posts on X (formerly Twitter). After treatment for a tick-borne disease, the dog’s condition worsened. A blood test showed severe anemia, but the responding veterinarian was unable to diagnose the dog’s condition. The owner entered Sassy’s signs and medical history in ChatGPT4, as well as the results of the dog’s blood work.
As the owner explained on a Twitter post, “(I)t occurred to me that medical diagnostics seemed like the sort of thing ChatGPT4 could be potentially good at, so I described the situation in great detail. I gave it the actual transcribed blood test results from multiple days and asked for a diagnosis. Despite a medical disclaimer, ChatGPT complied with the user request to make a diagnosis: immune-mediated hemolytic anemia (IMHA), which the dog’s owner claimed to be accurate.
“There’s tons of examples like this,” Wilkie noted.
AI-powered chatbots can provide instant responses based on reliable information provided by veterinarians, offering valuable insights and guidance to pet owners, he said. They can then provide the veterinarian with specific details, which helps facilitate a focused conversation during the appointment.
Using ChatGPT, Wilkie added, helps veterinarians provides pet owners with convenient access to reliable veterinary advice.
“To ignore this stuff, I think is a really bad idea,” he continued. “To fully embrace it and to trust everything it says I also think is a really bad idea. You need to sit somewhere in the middle, with awareness and an openness to understanding things in maybe a different way or do things in a different way.
“But you need to be aware that people are going to take a copy of the bloodwork and input it into ChatGPT and come to you saying this was the diagnosis I was given. Please give me an alternative or differential diagnosis with probability of those diagnosis being accurate.”
Original Article Published on AVMA News
