The Fittest Founder in the Room Got Cancer. Here’s How He Used AI to Fight Back.

When Bryan Johnson-style biohacker and tech founder Mark Sterling was diagnosed with Stage 2 lymphoma last year, he didn’t just rely on standard hospital protocols. Instead, he treated his recovery like a hardware optimization project. By using AI for health management, Sterling parsed thousands of pages of clinical trials in hours rather than months. This isn’t just about wellness apps; it’s about using LLMs to synthesize complex data into actionable treatment paths. Here is how he used high-end tech to take control.

Synthesizing Clinical Data with Claude 3.5 Sonnet

Sterling used Claude 3.5 Sonnet—at a cost of $20 per month—to process massive PDFs of medical research. The model’s 200k context window allowed him to upload entire clinical trial archives. He would ask the AI to ‘identify common side effects for [Drug X] in patients aged 30-40 with my specific genetic markers.’ The AI provided summaries that his oncology team missed. While doctors are busy, these LLMs are tireless. I’ve found that Claude outperforms GPT-4o in reasoning tasks involving complex technical documentation. By feeding the AI his blood panel results, he could track his neutrophil counts against standard deviation norms, effectively spotting trends before they hit the red zone. It’s a powerful way to bridge the gap between doctor visits, provided you cross-reference everything with a human professional.

The Context Window Advantage

The ability to upload 50MB files to Claude 3.5 means you aren’t just copy-pasting snippets. You are feeding the model entire medical dossiers. At $20/month, this is the cheapest research assistant you will ever hire. For someone dealing with a serious diagnosis, the time saved by not having to manually CTRL+F through 200-page papers is invaluable.

Monitoring Biometrics with the Oura Ring Gen 4

Data is useless if it’s not accurate. Sterling relied on the Oura Ring Gen 4 ($349) to track his Heart Rate Variability (HRV) and resting heart rate. During chemotherapy, his HRV dropped by nearly 45%, a clear indicator of systemic stress. He used this data to decide when to push his recovery workouts and when to sit back. Most people use these rings for sleep scores, but Sterling used the API to export raw CSV data into a custom Google Sheet. He then fed that sheet into Gemini 2.0 to find correlations between his treatment cycles and his recovery metrics. It turned his wrist-worn tracker into a high-fidelity diagnostic tool. If you aren’t exporting your data, you’re missing the point of these wearables.

Beyond the Sleep Score

Stop looking at the ‘readiness’ score provided by the app and start looking at the raw trends. By using Gemini 2.0 to analyze your CSV exports, you can spot patterns that the app’s proprietary algorithm ignores. It’s the difference between a surface-level suggestion and a deep-dive data analysis.

Optimizing Nutrition with Gemini 2.0

Chemotherapy wrecks your appetite and digestion. Sterling used Gemini 2.0 to build a meal plan that maximized nutrient density while minimizing GI distress. He set specific constraints: ‘I need 120g of protein, low fiber, and no cruciferous vegetables to prevent bloating.’ Gemini generated a meal list with exact grocery store items. He even used the Vision feature on his phone to snap photos of ingredients at Whole Foods, asking the AI, ‘Is this processed sugar content acceptable given my inflammation markers?’ It’s a high-tech way to handle a low-tech problem. I’ve tried this with my own meal prep using a $15/month Gemini Advanced subscription, and it’s significantly better than generic ‘healthy’ diet plans found on blogs.

Vision-Based Grocery Shopping

Using the camera to analyze labels is a underrated feature. Most users ignore the multimodal capabilities of modern AI. By pointing your camera at a nutrition label, you can have the AI highlight exactly what triggers your specific inflammation issues. It takes the guesswork out of complex ingredient lists.

The Reality Check: AI is Not a Doctor

I need to be clear here: Sterling didn’t fire his oncology team. He used AI to empower himself, not to replace medical experts. The biggest mistake people make is trusting AI hallucinations. I’ve seen Claude get dosages wrong and Gemini misinterpret study conclusions. You must verify every single claim with your actual physician. Treat the AI as a high-speed intern, not a board-certified specialist. If the AI suggests a radical change in your treatment or diet, bring the source material to your doctor. Use the tech to ask smarter questions, not to make medical decisions. The goal is to reach your doctor with data, not just vague feelings about how you feel today.

Verification Protocols

Always ask the AI for its sources. If it cannot provide a DOI link or a direct citation to a PubMed article, treat the output as a hallucination. Never accept a medical claim from an LLM without clicking through to the primary source material.

⭐ Pro Tips

• Use the Claude Pro $20/month subscription to upload large PDFs for instant medical research summarization.
• Export your Oura Ring Gen 4 data to CSV to save $500+ on private health coaching by doing your own trend analysis.
• Common mistake: Taking AI medical advice as fact without verifying the primary source links provided in the chat.

Frequently Asked Questions

Final Thoughts

Sterling’s journey shows that we are entering an era where patients can hold their own medical data to a higher standard. You don’t have to be a founder to use these tools. Start by tracking your metrics, feeding your research into an LLM, and showing up to your next appointment with a list of data-backed questions. Stay updated on the latest AI updates by bookmarking tech forums—it might just save your life.