Key Takeaways:
GitLab co-founder Sid Sijbrandij used AI to self-direct a personalized cancer treatment protocol after standard care failed for his recurring osteosarcoma, and is now in remission.
Developer Paul Conyngham used genetic algorithms and AI tools to screen millions of compounds and design a custom mRNA vaccine for his dog Rosie's aggressive cancer, shrinking her tumors.
Sam Altman called Conyngham's work "amazing" and said it should be a company, but the bottleneck is paperwork, not science.
Two stories surfaced this week that share a pattern. In both, a person facing cancer turned to AI when institutional medicine ran out of options. In both, AI delivered results the institutions had not.
Sid Sijbrandij, co-founder of GitLab, faced recurring osteosarcoma. Standard care failed. He self-directed a protocol using AI for deep research, maximal diagnostics, repurposed drugs, and personalized treatment decisions. He published his entire process, raw data and treatment deck included, publicly. His cancer is currently in remission.
Paul Conyngham's dog Rosie had aggressive cancer. He used genetic algorithms to screen millions of compounds, ChatGPT to build his research pipeline, Gemini for genetic construct design, and Grok for validation. He worked from 300GB of sequencing data to design a custom mRNA vaccine. When university ethics approval would have taken until mid-2026, he drove to an approved lab and got the injection done. Rosie's tumors shrank. One new mass turned out to be a different cancer, meaning the vaccine worked on the original.
Sam Altman met with Conyngham and said the work was amazing, that it should be a company, and that AI had empowered an individual to act with the power of a research institute. Conyngham is now building one.
The bottleneck in both cases was not scientific. It was institutional. Ethics board timelines, regulatory approval processes, patent blocks, and the friction between what an individual can figure out and what an institution will allow. As Conyngham put it: the science is solved. We are bottlenecked on paperwork.
None of this replaces clinical trials or regulatory oversight. What it does is demonstrate that the tools for understanding disease and designing interventions are no longer locked inside research institutions. They are available to anyone with the technical literacy to use them and the urgency to try.
People Also Ask
Q: How did the GitLab CEO use AI for cancer treatment? A: Sid Sijbrandij used AI research tools to analyze his osteosarcoma, identify repurposed drug candidates, and direct a personalized treatment protocol after standard care failed. He is currently in remission.
Q: What is the Rosie mRNA vaccine story? A: Developer Paul Conyngham used AI tools to screen millions of compounds and design a custom mRNA vaccine for his dog's cancer from 300GB of genetic sequencing data. The tumors shrank after vaccination.
Q: Did Sam Altman comment on AI-designed cancer treatments? A: Yes. Altman met with Conyngham and said AI had empowered an individual to act with the power of a research institute, adding that the work should become a company.
Q: Can AI replace traditional cancer treatment? A: AI tools can accelerate research and personalize treatment design, but they do not replace clinical trials, regulatory oversight, or medical supervision.
DoorDash pays 8M gig workers for AI training (prior WYDE coverage)
World + Coinbase launch AI agent identity layer (prior WYDE coverage)