The owner of the dog described the story of its rescue on social media. It all started when the dog was diagnosed with a tick-borne disease. Initially, the treatment seemed positive, but after a short period of time, anemia developed, and the condition worsened. After tests for other related illnesses, the veterinarians were unable to identify the cause and advised the owner to simply observe the course of the disease. This response did not satisfy the owner, and they sought help from another clinic.
During the intervals between appointments, the user turned to ChatGPT for assistance and provided all available test results. The neural network warned that it is not a professional veterinarian and after confirmation, issued a list of possible conditions. Since the dog underwent numerous tests, the owner ruled out the possibility of internal bleeding and other potential causes. As a result, the only remaining option was immune-mediated hemolytic anemia (IMHA) in the animal
At the new veterinary appointment, the diagnosis suggested by the neural network was confirmed. This time, the correct treatment was chosen, and the situation finally began to improve. “We started the proper treatment, and now the dog has almost fully recovered.”
Interestingly, such accurate results were generated by the recently launched GPT4 version, while GPT 3.5 (like the initial veterinarian) proved to be powerless.
You can read full story at https://twitter.com/peakcooper/status/1639716822680236032