We need to face the dark health care reality behind the AI-fueled cancer treatment stories
Anecdotal Bait-and-Switch
How They Deceive You
Propaganda
Uses positive AI treatment anecdotes to pivot into broad corruption claims while treating disputed statistics as established fact.
Main Device
Anecdotal Bait-and-Switch
Opens with inspiring cases then indicts the entire system without evidence linking the two.
Archetype
Progressive anti-corporate health skeptic
Views medical and pharmaceutical institutions as inherently captured and prioritizes systemic indictment over case-specific evidence.
Opens with uplifting stories only to pivot into systemic corruption claims by citing disputed death statistics as settled fact.
Writer's Worldview
“Progressive anti-corporate health skeptic”
2 findings · 1 omission · 9 sources compared
What is your news hiding from you?
Same analysis. Any article. Completely free.
Narrative Analysis
The article uses two documented cases of AI-assisted cancer interventions as an entry point to argue that U.S. healthcare suffers from entrenched corruption and unequal access, but it does so by pivoting away from case-specific outcomes toward broader systemic claims without direct evidentiary links.
Key Findings
- Framing technique: The piece opens with the verifiable stories of Paul Conyngham’s work on a personalized mRNA vaccine for his dog Rosie and Sid Sijbrandij’s AI-supported osteosarcoma management, then immediately asserts that “the medical and health care industries are among the most deeply corrupt.” No data from either case is presented to substantiate the corruption claim; the transition instead projects a two-tier privilege scenario onto the successes.
- Use of disputed statistics: The article treats “medical errors leading causes of death” and “pharmaceutical regulatory capture” as settled facts supporting its thesis. The cited error statistic traces to a 2016 analysis whose methodology has been questioned by subsequent health reporting for overcounting and definitional issues.
- Selective emphasis on agency: Both cases involved wealthy individuals supplementing standard care with additional resources. The article correctly notes this pattern but does not connect specific actions (genomic sequencing, open-sourcing records) to measurable results before shifting to systemic critique.
What Was Missing
The article omits documented, case-level outcomes that are independently reported. Rosie’s treatment produced measurable mobility gains confirmed by university collaborators, and Sijbrandij’s records were open-sourced with genomic data made available for wider research. These facts are verifiable through participant timelines and institutional reporting; their absence narrows the reader’s view of what the interventions actually achieved before the systemic argument begins.
Source Context
Blaze Media produces opinion and news content with a consistent right-leaning editorial stance. The author, Andrew Edwards, is not identified with prior specialized health reporting in available records. The outlet’s revenue model relies on subscriptions and advertising rather than academic or medical institutional ties.
Coverage Differences
Other outlets covering the same primary events stayed closer to process details and participant statements. The Scientist and UNSW News focused on sequencing methods and laboratory collaboration without systemic claims. The Conversation included an oncologist’s cautionary perspective on the approach. Sid Sijbrandij’s own Substack post supplied the raw timeline of diagnosis, recurrence, and AI use without broader institutional indictment.
The article accurately identifies that both successes involved substantial personal resources and that questions of scalability remain open. At the same time, its reliance on contested statistics and rapid pivot from individual results to systemic corruption reduces the precision of the analysis.
Further Reading
- The Scientist: ChatGPT and AlphaFold Help Design Personalized Vaccine for Dog With Cancer
- The Conversation: A man used AI to help make a cancer vaccine for his dog — an oncologist urges caution
- UNSW News: Meet the man who designed a cancer vaccine for his dog
- Century of Bio: Sid Sijbrandij’s cancer timeline and information-maximalist approach
- Sid Sijbrandij Substack: Founder-mode documentation of diagnosis and AI-assisted protocol
Neutral Rewrite
Here's how this article reads with loaded language removed and missing context included.
AI-Assisted Cancer Treatments Draw Attention Amid Questions on Evidence and Access
Overnight interest has centered on entrepreneur Paul Conyngham, who combined large language model analysis with established veterinary treatments and private laboratory services to develop a tailored protocol for his dog Rosie. The case has circulated widely online. Separately, GitLab executive chair Sid Sijbrandij has described his own use of similar tools following a diagnosis of osteosarcoma several years ago. Both accounts have prompted discussion of the role of artificial intelligence in medical decision-making and the practical limits of such approaches.
Conyngham detailed the process in a post on X, noting that the language models assisted with literature review and protocol suggestions but performed none of the laboratory or clinical steps. Samples were collected and DNA sequenced by external teams; the mRNA vaccine itself was manufactured by Professor Pall Thordarson at the UNSW mRNA Institute; administration was handled by Professor Rachel Allavena and Dr. José Granados at the University of Queensland; and bioinformatics guidance came from Professor Martin Smith. University collaborators reported measurable quality-of-life improvements for Rosie, including restored mobility, following the intervention.
Sijbrandij, who stepped back from the CEO role at GitLab at the end of 2024 to focus on treatment, has stated that standard therapies alone did not halt disease progression. He assembled an expanded diagnostic and research team, incorporated AI-supported data analysis, and made extensive genomic records publicly available. His account indicates the cancer is currently in remission. The open-sourced dataset has been referenced by others as a potential resource for additional research.
Both individuals are technology entrepreneurs with substantial resources. Conyngham described dedicating roughly one hundred additional hours per week to documentation and coordination. Biomedical engineer Patrick Heizer reviewed the publicly shared materials from Sijbrandij’s case and characterized the effort as extensive while estimating costs in the tens of millions of dollars. He noted that the broader vision of routine personalized AI-driven medicine remains unrealized given current evidentiary standards.
Biochemist Egan Peltan examined the Rosie case and stated that available information does not isolate the contribution of the custom mRNA component beyond the effect of an FDA-approved canine PD-1 inhibitor. He described the most straightforward explanation as a partial response to the approved agent. Heizer similarly pointed to the absence of controlled risk protocols and outcome data that would allow clear attribution of results.
OpenAI chief executive Sam Altman posted that he had met Conyngham and described the account as notable. Coverage of both stories has appeared across technology and general-interest platforms, often highlighting the individuals’ initiative while noting the specialized resources involved.
Documented barriers to wider application include the requirement for specialized laboratory capacity, regulatory approvals for novel agents, and sustained clinical oversight. Sijbrandij’s decision to release genomic data provides one concrete mechanism by which individual cases could contribute to collective research without requiring equivalent private expenditure by others. University teams involved in the Rosie protocol have likewise published aspects of the work, adding to the available record.
Questions remain about reproducibility. Neither case included randomized controls that would separate the effects of multiple concurrent interventions. Cost estimates from reviewers indicate that replicating the full suite of diagnostics, manufacturing, and monitoring would exceed typical insurance or public coverage limits in most jurisdictions. At the same time, the open release of data and methods offers a pathway for independent verification or adaptation by research groups.
The accounts illustrate both the capacity of individuals to coordinate complex information and the practical constraints that currently limit broader deployment. Further peer-reviewed reporting on outcomes, costs, and comparative effectiveness will determine whether elements of these protocols can be standardized or scaled.
Investigation Log · 33 steps
Starting investigation...
Investigating The Blaze
Investigating Andrew Edwards
Searching for "Paul Conyngham Rosie dog mRNA vaccine AI LLM"
Verify the viral story details and outcomes
Searching for "Sid Sijbrandij osteosarcoma cancer treatment AI GitLab"
Verify the billionaire's cancer story and claims of remission
Source: Andrew Edwards
Search results identify multiple individuals named Andrew Edwards but none as a media source or author with documented expertise in medicine or AI. The closest match is a LinkedIn profile for a communications professional and writer with ~20 years in public affairs and journalism, currently in a strategic role; no publications, credentials, or subject-matter focus are listed. Other results refer to unrelated figures such as athletes, a musician, and an artist.
Source: The Blaze
Blaze Media is a privately held U.S. media company producing news, opinion, and entertainment via theblaze.com and BlazeTV. Formed in 2018 via merger of TheBlaze and CRTV LLC and headquartered in Irving, Texas, it has no documented independent fact-check ratings or health/medical reporting track record. Output focuses on original shows, podcasts, and articles.
Searching for "Patrick Heizer Sid Sijbrandij cancer critique OR analysis"
Verify the biomedical engineer's assessment of costs and evidence
Searching for ""medical errors" leading cause of death US statistics"
Check accuracy of corruption/inept system claims
Comparing coverage of "Paul Conyngham Rosie AI dog cancer vaccine"
Comparing coverage of "Sid Sijbrandij GitLab cancer AI treatment"
Coverage comparison completed
Coverage comparison completed
Framing
Uses inspiring AI cancer stories (Rosie dog vaccine, Sid Sijbrandij remission) as lead-in to assert "The medical and health care industries are among the most deeply corrupt" and warn of two-tier privilege, without linking specific evidence from the cases to systemic claims.
Creates impression that these successes prove entrenched corruption and inaccessibility rather than individual agency or tech potential, steering readers toward anti-establishment distrust.
Cherry-Picking
Cites disputed "medical errors leading causes of death" statistic and "pharmaceutical regulatory capture" as established facts to support corruption thesis, while other outlets covering the same stories emphasize innovation and caution without systemic indictment.
Selects contested data points that amplify negative framing; 2016 BMJ study on errors is widely questioned for methodology.
Missing Context
The Rosie treatment produced documented quality-of-life improvements (mobility) per university collaborators, and Sid's case involved open-sourcing extensive genomic data for broader research benefit.
Omitting positive measurable outcomes from the cases themselves makes the pivot to "dark reality" appear more totalizing than the evidence warrants.
Writing analysis narrative
Writing verdict summary
Writing neutral rewrite
Investigation complete. Preparing report...
Analysis narrative ready
Neutral rewrite ready
The Compass
You see how this outlet sees the world.
How do you see it? Find your political shape in a few minutes.
Take the testOr check your own article