Overnight star entrepreneur Paul Conyngham is scaling his company based around his experiences mixing LLM medicine, analysis, established treatments, and private laboratory services to bring tailored treatments, such as he brought to his dog Rosie, to a wider, human audience. The Rosie saga has gone viral and birthed a business, and LLM companies are thrilled. Will any of it work?
Last week’s viral story of Rosie the dog, whose ambitious owner leveraged LLMs (among many other things) to create a custom-tailored mRNA vaccine continues to instruct, maybe edify, but surely divide.
The medical and health care industries are among the most deeply corrupt.
Enter the human version of the AI self-medicating story: GitHub founder and billionaire Sid Sijbrandij was diagnosed several years ago with osteosarcoma. Sid pursued standard treatments, but they weren’t enough to halt the cancer’s progress. So Sid countered with a truly impressive, even inspiring, quotient of agency — throwing himself into gathering his own medical team, deploying AI where possible, maximizing every diagnostic test he could find, and open-sourcing his records.
With this approach, Sijbrandij says he was able to advance something workable in terms of a treatment protocol. His cancer is in remission, and he’s talking about the process on X and his website.
First the good news
The stories have remarkable parallels. Both involve successful, wealthy tech entrepreneurs. Both involve cancer, and both offer hope. The timing of both is curious.
There are a few very hard — even uncomfortable and offensive — but necessary questions to ask with respect to media reality. Boomer elites hope to keep the all-important economic and financial line going comfortably up as they exit into retirement en masse. How much of the self-guided vaccine push, and the rosy vision behind it, is real?
At the end of 2024, Sijbrandij transitioned from CEO to executive chair of GitLab, saying, “I want more time to focus on my cancer treatment and health.”
RELATED: A man used Grok to save his dog. Is intellectual property about to die?
wildpixel/Getty Images
Paul Conyngham details in a long essay posted to X exactly what the chatbots did to help. Also detailed are the various other cost-prohibitive treatments he leveraged to save his dog. Conyngham admits he was putting in an extra hundred hours of work per week on complex paperwork. This is time most people don’t have to spare, obviously. He does not mention the costs, but in America at this point few have any cash after paying the monthly nut. In Paul’s essay, we read “what the chat bots did NOT do”:
They did not collect samples. They did not isolate or sequence the DNA. They did not physically manufacture the vaccine. They did not administer it. Many brilliant scientists were required — including Professor Pall Thordarson at the UNSW mRNA Institute who manufactured the vaccine, Professor Rachel Allavena & Dr. José Granados at the University of Queensland who administered it, and Professor Martin Smith who provided expert guidance on the bioinformatics throughout.
Enter the dissenting opinions and scam artists.
Peeling away the hype
Never one to leave credit, cash, or free and mistaken public goodwill lying on the table, Sam Altman weighed in last week. “The coolest meeting I had this week,” he posted, “was with Paul, who used ChatGPT and other LLMs to create an mRNA vaccine protocol to save his dog Rosie. It is amazing story.”
At about the same time, just after the euphoria dissipated and Altman chimed in, a series of critical posts took swings at the general, and admittedly largely dilettante, story.
Patrick Heizer, a working biomedical engineer, called Sid Sijbrandij’s approach and presentation of the evidence in his self-management of osteosarcoma “extremely impressive.” However, he also estimated that Sid spent “tens of millions” to make it happen.
In Heizer’s learned opinion, the era of personalized, AI medical utopia is not here. With respect to the evidence presented in the story of Rosie the dog, Heizer was dismissive — citing a general lack of risk protocol and evidence for what did and didn’t work.
Another biochem Ph.D./founder figure on X, Egan Peltan, echoed Heizer’s doubts regarding the Conyngham/Rosie story. “There’s no evidence his process (beyond FDA approved doggie α-PD-1) had any impact on disease progression. The most parsimonious explanation is a partial response to α-PD-1.”
In our previous examination of this heartwarming tale, I suggested that the era of AI medicine poses certain glimmers of hope with respect to the future of decentralized — and, potentially, more affordable and effective — medical treatments.
But this sunny future must first punch through a systemically corrupt and increasingly inept system — with medical “errors” leading causes of death, pharmaceutical regulatory capture well entrenched, overall care suffering long-term degradation, and institutional scams outstripping even the power of the federal courts to fight.
Will we arrive at a two-tiered privilege scenario where regular Americans are once again supplying their data to be used merely for the benefit of those at the top, who can afford to leverage the panoply of treatment options? We’ve gotten far too accustomed to being left out in the cold, and the medical and health care industries are among the most deeply corrupt.
Tech
