Paul, a tech entrepreneur with no medical background, used AI tools like ChatGPT and AlphaFold to develop a personalized mRNA vaccine that significantly shrank his rescue dog Rosie’s aggressive cancer tumors. This groundbreaking case highlights the potential of AI-driven personalized medicine while raising important questions about balancing rapid innovation with the need for rigorous safety and regulatory oversight.
The story revolves around Paul, a tech entrepreneur from Sydney with 17 years of experience in machine learning but no medical background, and his rescue dog Rosie, who was diagnosed with aggressive mast cell cancer in 2024. Despite multiple surgeries, chemotherapy, and immunotherapy, Rosie’s tumors kept growing, and vets gave her a prognosis of one to six months to live. Refusing to accept this, Paul decided to leverage AI tools to design a personalized cancer treatment for Rosie, treating cancer as corrupted data in DNA that could potentially be corrected.
Paul used ChatGPT as a research assistant to explore treatment options and was directed towards personalized immunotherapy, which targets cancer cells specifically by teaching the immune system to recognize mutated proteins unique to the cancer. He collaborated with a genomics university to sequence both Rosie’s healthy and tumor DNA, identifying mutations responsible for the cancer. Using AlphaFold, an AI developed by Google DeepMind, Paul predicted the 3D shapes of mutated proteins to find suitable targets for a vaccine. He then applied his machine learning skills to select the most promising mutated proteins that could trigger an immune response.
The next step involved creating a custom mRNA vaccine based on the selected mutated proteins, similar in technology to COVID-19 vaccines, which would train Rosie’s immune system to attack only the cancer cells. Paul partnered with Professor Paul Thorda at the UNSW RNA Institute to manufacture the vaccine, a process that took less than two months. After navigating complex ethical approvals and driving Rosie 10 hours to receive the experimental vaccine, the results were remarkable: within a month, Rosie’s main tumor shrank by up to 75%, and her energy and mobility significantly improved.
While the story is groundbreaking, experts caution that creating a single mRNA vaccine is relatively straightforward, but proving its safety and effectiveness through rigorous clinical trials and scaling production is the real challenge. The story highlights a tension between innovative, rapid approaches enabled by AI and the slow, bureaucratic nature of traditional medical research and regulatory systems. Some argue that the current institutional framework can delay life-saving treatments, while others emphasize the importance of safety and controlled studies to avoid harmful outcomes.
Ultimately, Paul’s achievement demonstrates the transformative potential of AI in personalized medicine and cancer treatment, showing that individuals with the right skills can now contribute meaningfully to cutting-edge research. Although Rosie’s case is not a complete cure, it represents a significant step forward and raises important questions about the future of medical innovation, the role of AI, and how to balance rapid progress with patient safety. This story has captured public imagination because it suggests a future where personalized cancer vaccines could become more accessible and effective for both animals and humans.