How I Used AI to Navigate My Health Challenges (When Doctors Couldn’t)

(Disclaimer: AI isn’t a doctor—but when doctors aren’t listening, it sure can help you advocate for yourself.) For any one still sceptical, I would still be undiagnosed and a ticking time bomb if I didn’t use AI and get to the bottom of my vascular compressions (yes, multiple). For those with environmental concerns - please see my statement at the bottom of this article.

How I’ve Used AI in My Health Journey

Analysed my symptoms and test results, identifying four possible investigative paths—all of which aligned with insights I’d gathered from patient support groups.
Re-typed an illegible specialist letter—even my GP thanked me because she couldn’t read the original. I uploaded it, asked AI to re-format it, and voila: clarity.
Drafted referral letters by summarising my complex medical history, symptoms, and past tests into a clear, professional structure. My GP appreciated having something ready to go—it saved time in a 15-minute appointment and ensured nothing important got missed.
Prioritised what to investigate next, based on symptoms, results, and likelihoods. This sped up my diagnosis significantly—it would’ve taken much longer if I’d been guessing randomly.
Spotted potential links between blood test results when I asked AI to look for patterns and correlations.

Suggestions I Haven’t Tried—But Others Recommend

AI-powered second opinions – platforms like Q Bio or AI-enhanced radiology tools can interpret test results in new ways or flag things doctors might miss.
Lifestyle pattern tracking – I kept a journal of food, sleep, water, and activity. I could’ve dropped this into AI to spot patterns or any correlation between my bad days and the details in the journal entries.

Worried About Data Privacy?

You’re not alone—it’s a valid concern. But as someone who works in tech, I’ll say this: if someone really wants your data, they’ll find a way. Many major organisations you trust are shockingly lax with security (just Google “Australian data breach” if you need a reminder).

A workaround: Mask identifying info and use a VPN. Most AI tools don’t need your name—just the symptoms or test results.

For me, the risk of doing nothing—the pain, loss of daily function, and potential long-term damage—far outweighed the risk of data exposure. But it’s a personal call, and that’s yours to make.

AI vs. Doctor: When the Machine Got It Right

I suspected a vascular issue (thanks to AI and my research on Google Scholar) and waited three months to see a specialist. The vascular surgeon dismissed me immediately—no new tests, just confidently told me “I don’t think you have any vascular concerns”. I paid the $500 fee, wiped the tears from my eyes and immediately jumped onto ChatGPT to ask which tests I needed to rule out a vascular issue. My GP sent me for those two tests. Both tests highlighted MULTIPLE vascular abnormalities.

In short: the vascular surgeon was confident—and wrong. AI was correct.

Environmental Note

I absolutely care about the environment — but I think we need to be realistic about where the responsibility lies. Blaming the tools that use electricity (like AI or electric cars or dishwashers) distracts from the bigger issue: how we generate that electricity in the first place. That’s where the real environmental damage happens, and that’s what governments, industries, and scientists need to address if we want serious climate solutions. Holding back progress in tech because our grid isn’t clean yet seems like punishing innovation for a problem it didn’t cause. That said, it’s also important that AI systems continue improving their energy efficiency and sustainability, so they grow smarter and greener as adoption increases. But for those still not convinced, asking ChatGPT a question uses roughly the same power as watching a 10-second cat video.