Cancer sucks.
There’s no other way to say it. Almost everyone reading this knows someone they care about who’s gone through the fear, the treatments and the uncertainty.
In many ways, the uncertainty is the hardest part. Because the fight against cancer has always been defined by trial and error, with doctors trying one treatment after another and hoping something works.
But what if that guesswork is finally coming to an end?
A heartwarming story out of Australia suggests that it just might be.
Rosie’s Story
When veterinarians told Sydney tech entrepreneur Paul Conyngham that his rescue dog Rosie had only months to live, he refused to accept it.
Rosie had been diagnosed with advanced mast cell cancer in 2024. Chemotherapy slowed the disease but didn’t stop it. So Conyngham, a data scientist and machine learning engineer with no formal training in biology or oncology, turned to ChatGPT and other AI tools to do something no one had done before.
He built a personalized cancer vaccine for his dog.
Image: The Australian
Conyngham worked with researchers at the Ramaciotti Centre for Genomics, UNSW and the University of Queensland to sequence Rosie’s tumor, identify mutations and create what multiple reports described as a canine-specific personalized mRNA cancer vaccine.
Rosie received her first treatment in December, and within a couple of months some of her tumors had shrunk by about 50%.
This one hit close to home for me. My husky–Australian shepherd mix has been living with a terminal neurological disease called GME. Thankfully, she’s been in remission for five years. But when you’ve lived through that kind of diagnosis, you know how meaningful any sign of progress can be.
To be clear, Rosie’s case doesn’t prove ChatGPT has “cured cancer.” Even some of the coverage around the story has warned against turning a remarkable case study into a miracle headline.
But it does reveal two incredible things.
The first is that a single person using modern AI tools was able to cross a knowledge gap that would have been almost impossible to bridge just a few years ago.
I’m finding this in my own interactions with AI.
The second is that AI is getting good enough to help personalize cancer treatments.
You see, for most of modern oncology, doctors have treated cancer based on categories. They identify the type of cancer, match it to a standard therapy and adjust if it fails.
That approach has saved a lot of lives. But it has limits.
Because cancer isn’t one disease. It’s many.
Two patients can have the same diagnosis and respond in very different ways. That’s why treatment can feel unpredictable.
But AI is starting to change that.
At UC San Diego, researchers developed a model called DeepHRD that can detect a key biomarker, homologous recombination deficiency, or HRD, directly from routine biopsy slides.
HRD helps determine whether a patient is likely to benefit from platinum chemotherapy or PARP inhibitors. But until recently, finding that signal required additional genomic testing that could take time and sometimes failed.
DeepHRD pulls HRD from pathology images doctors already use, which means faster and more confident treatment decisions.
In studies, the model was able to predict HRD in breast and ovarian cancers from standard H&E slides across multiple patient groups.
I understand if all this sounds a little too technical. But what it means in practice is simple.
The faster you can read the biology of a tumor, the faster you can stop guessing.
And it goes beyond diagnostics.
Researchers are also using deep reinforcement learning to personalize treatment schedules in prostate cancer. In a 2024 study, AI-guided treatment more than doubled the time before the cancer got worse compared with current approaches.
The idea here is to use the same treatments with better timing, smarter sequencing and closer feedback based on how the tumor is actually responding.
That’s why Rosie’s story could be the first of many like it, as AI helps doctors move from what usually works to what works best for each patient.
This shift is showing up in drug development too.
AstraZeneca (NYSE: AZN) is using AI to develop a biomarker that helps identify which lung cancer patients are more likely to benefit from one of its drugs.
In a Phase III study, patients who tested positive for that marker did noticeably better than those who didn’t.
In other words, AI isn’t just helping find new drugs. It’s helping decide who should get them.
Then there is the infrastructure layer.
Companies like Tempus (Nasdaq: TEM) are helping build the foundation for this new kind of medicine. It has created a large library of patient data and
tools that turn it into useful insights for doctors.
Image: Tempus.com
Its platform includes genetic testing, tracking how treatments are working and matching patients to clinical trials — often in days instead of months.
Tempus certainly has Cathie Wood’s attention. This month, Wood’s Ark Genomic Revolution ETF (ARKG) bought around $2.1 million worth of shares in TEM.
That might look like a bet on a single company, but it’s also a bet on where medicine is heading.
Here’s My Take
I saw many of these AI healthcare use cases firsthand at Nvidia’s GPU Technology Conference in Washington, D.C. last fall.
And it reinforced something I’ve come to believe.
The future of cancer care isn’t likely to arrive as one dramatic cure.
Instead, it’ll arrive as a series of tools that make treatments more specific, more adaptive and more personal.
I believe we’re moving toward a model where AI becomes a standard part of cancer care. Not replacing doctors, but helping them make faster and better decisions.
As that happens, the companies building the data and infrastructure behind this shift could become very important. Tempus is one of them, as Strategic Fortunes readers already know since it’s in our model portfolio. But ARK’s recent investment suggests that more investors are starting to see the same thing.
Which means Rosie’s case might look like an outlier today…
But I believe it’s better understood as a preview of what’s to come.
Regards,
Ian KingChief Strategist, Banyan Hill Publishing
Editor’s Note: We’d love to hear from you!
If you want to share your thoughts or suggestions about the Daily Disruptor, or if there are any specific topics you’d like us to cover, just send an email to [email protected].
Don’t worry, we won’t reveal your full name in the event we publish a response. So feel free to comment away!
