ChatGPT does not cure dog cancer

When an Australian tech entrepreneur with no background in biology or medicine said ChatGPT helped save his dog from cancer, the story exploded with the kind of validation that tech companies have longed for: proof that AI will revolutionize medicine and take on one of the deadliest diseases. The reality, as usual, is more complex.

The version of the story that spread online first I mentioned by Australianit was relatively clear. In 2024, Sydney resident Paul Cunningham learns that his dog Rosie has cancer. Chemotherapy slowed the disease but failed to shrink the tumors. After vets said ‘nothing could be done’ for Staffordshire Bull Terrier Shar Pei, Cunningham He said “I took it upon myself to find a cure.”

Cunningham said he used ChatGPT to brainstorm treatment ideas. The chatbot surfaced as an option for immunotherapy, directing it to experts at the University of New South Wales, who then genetically profiled Rosie’s cancer. Then use ChatGPT and Google Protein structure AI model AlphaFold To help understand the results. With the help of UNSW professor Pall Thordarson, A Personal mRNA vaccine Specifically designed for Rosehip tumor mutations. Thordarson said Australian This is believed to be the first time such a treatment has been designed for a dog.

A few weeks after Rosie’s first injection last December, Cunningham said her tumors had shrunk and she was feeling better, even chasing rabbits in the garden. However, they did not go away completely, and one tumor did not respond at all. “I’m under no illusions that this is a cure, but I think it has given Rosie significantly more time and quality of life,” Cunningham said. Australian.

This nuance was lost as the story spread. Newsweek The headline ran, “Owner with no medical background invents cure for terminal dog cancer,” instead New York Post A tech pro saves his dying dog by using ChatGPT to code a custom cancer vaccine, he announced. On social media, a lot the accounts exaggerated Rosie’s case as a “cure” and a sign of the arrival of a new era of personalized medicine. some, In particular OpenAI CEO and founder Greg Brockman certainly should have known better, and others, like Google DeepMind CEO Demis HassabisI did and shared it without any fuss. Elon Musk Join in Also, keen to point out that XAI Your puppy also played a role -Details that were absent from much of the original coverage.

The story also gives a lot of credit to the AI. Not only was Rosie not cancer-free, it’s not clear that the mRNA vaccine was responsible for her improvement. It was personal therapy administered together Another form of immunotherapy is known as a checkpoint inhibitor, which is designed to help the immune system target tumors, making it difficult to know whether a vaccine will have any effect at all. One of the participating scientists, Martin Smith, He said The team performs tests to check the immune response.

The vaccine itself was not produced by the chatbot. ChatGPT did not design or create Rosie’s Therapy; Human researchers did. At most, the chatbot served as a research assistant helping Cunningham analyze the medical literature — impressive, but far from implicit hacking.

Reports are also vague about AlphaFold’s role. David Asher, professor and director of biotechnology programs at the University of Queensland in Australia, said Edge The model “can contribute to structural hypotheses about proteins, but it is not a ready-made system for cancer vaccine design.” He noted that official guidance also cautions that AlphaFold has not been validated to predict the effects of certain mutations and does not represent “many biologically important contexts” either.

It is difficult to determine your puppy’s contribution. On the 10th, Cunningham books “The final composition of the Rose vaccine was designed by Grock,” but it is not clear what that means in practice or what input was given to the model. Grok realistically falls into the same category as ChatGPT: a tool that “can help search the literature, summarize research, translate terms, suggest workflows, draft code or documents, and help the user consider options,” Asher said. It’s a useful role, but not exactly what the design of a cancer vaccine suggests.

Overall, Asher said Rosie’s case is “better viewed as unusual and very specific evidence that this can happen rather than as a model that ordinary people can easily reproduce.” He said it would require a “significant” workforce of experts, “not just a chatbot and some prompts.”

This distinction is particularly important in medicine, where success depends not only on the production of plausible information, but on the physical work done by experts to produce, test, and deliver the actual treatment. Alvin Chan, an assistant professor at Nanyang Technological University in Singapore, who builds artificial intelligence for biomedical and pharmaceutical discoveries, said: Edge The “AI made this” framework ignores this massive human effort, without which “AI output would still be just text on a screen.” In Rosie’s case, AI is best understood as a tool to draw a blueprint rather than as the creator of the treatment itself.

The whole thing has the faint whiff of a PR stunt that’s hard to shake. Bold claims made on questionable foundations using arcane methods fit comfortably into the world of tech fundraising. mRNA vaccines—much like the broader promise of personalized medicine—remain largely unproven as cancer treatments in humans, let alone dogs, and while the case may be real, it seems highly tidy and conveniently covers up the tens of thousands of dollars and significant expertise required to turn the idea into a viable treatment.

I’ve reached out to Conyngham to ask to chat on X but have yet to hear back. His profile says “Eliminate Cancer for Dogs” and Links to a Google form describing his “dream of making this process something accessible to everyone.” The form asks if your dog has cancer, whether you are a researcher or scientist who would like to participate, and whether you are an investor.

I think it would be a mistake to dismiss Rosie’s story as completely meaningless. AI may not replace the laboratory any time soon, but it is making science more accessible to ordinary people. However, this is not the same as making care easier, and few patients – or pet owners – have ready access to the world-class experts, specialized equipment, and deep pockets needed to turn this information into real treatment.