Ask ChatGPT to talk like any historical figure or fictional character, and you could discuss physics with Einstein, politics with Napoleon and tech with Tony Stark. But as people use it to write essays and poems, ChatGPT's factual errors while writing research papers have also been flagged, cautioning users against deploying it for sophisticated tasks.
The idea of AI stepping in for an actual doctor may be scary, but ChatGPT managed to outdo a veterinarian to save a dog's life.
How did ChatGPT pull this off?
After going through transcripts of blood tests and getting a description of a dog from its owner, the generative AI spotted a condition that even the vet had missed.
The pet named Sassy had been diagnosed with a tick-borne disease and was improving before a decline in her condition.
Despite several tests, the vet couldn't find a reason for her worsening anaemia and advised the owner to wait and watch.
But the pet parent wasn't willing to wait it out, and fed all the information to ChatGPT to get some background of Sassy's condition.
The AI itself clarified that it isn't a doctor
Although it initially clarified that it isn't a veterinarian, ChatGPT did point out that the dog may be suffering from immune-mediated hemolytic anaemia.
After Sassy's owner consulted another vet, the diagnosis was confirmed, and the dog made a full recovery.
Although the incident has gone viral after the pet parent Cooper tweeted about it, users still need to verify results with actual doctors or else ChatGPT could worsen the online self-diagnosis problem.
Still too early for AI to replace doctors
Misinformation about diseases including Covid has been a major issue, and even an AI ultimately gets its information from existing sources.
While ChatGPT is fascinating, the world of medtech is already introducing algorithms to spot breast cancer, eye diseases and other serious conditions for timely treatment.
