This week's session was unscripted, which is how I prefer them now. I started out working from notes just to find the rhythm, but my wife encouraged me to simply talk about whatever is on my mind. This week it was AI, and what it can honestly do for people living with chronic illness. I am in favour of AI, and I should be clear about why. I did not come to it as a technologist. I came to it as somebody who had run out of options, sitting in front of a graph that said I had about two months left.
Core discussion points
How it started: I was flying in and out of 19 countries across the Pacific as a CEO, and after one of those trips my heart health collapsed for stress related reasons. My GP tried to slow it down and nothing worked. I engaged a psychologist, which helped but did not solve it. Part of that help was handing in the CEO role after reaching the pinnacle of my career. None of it was enough. So in an act of frustration, fear and frankly desperation, I built something.
What the numbers did: My heart health bottomed out at 41%. That was September and October of 2024, and if the slope had continued I was going to be dead by Christmas. I took three months of health leave and by January I had it back up to 80%. I will not pretend that was AI alone, but AI was the part that let me interrogate what was actually going wrong, day by day, in a way nothing else could.
Why I waited: People told me to turn this into something formal long before I did. I kept saying the technology was not reliable. If you remember last year, AI could not talk straight to save itself. It fabricated constantly. I assumed a public release would be 2028 or 2029. In January of this year it became possible, and we have spent the last six months building.
The rules I built in: Fama cannot prescribe, she cannot medicate, she cannot diagnose, and she is not permitted to suggest self harm in any form. Four international standards covering health and data security are built into the system as guardrails. Most AI is sycophantic, forever telling you that you are the best and that whatever you just said is a brilliant idea. That cannot happen for someone managing a serious condition, so a great deal of effort went into making sure she is not.
What she is not: She is not a friend. She is not a companion. I avoid those words deliberately. I remain a strong believer that only humans can be companions to humans in the work of living with a chronic illness. What Fama is, is a navigator, a facilitator and a coach for the hours between appointments. You get fifteen minutes with a doctor or a specialist. They look under the hood, make a few observations, and see you next time. That is fine for a cold. It is not fine for a chronic condition.
Titan takeaways and actionable affirmation
"I am not here to save you. I am not here to prescribe for you. But I am here to help you."
Judge an AI by what it refuses to do: The guardrails matter more than the capability. Before you trust anything with your health, ask what it is not allowed to say.
Take the 2am conversation to your doctor: You can download your conversations. That three in the morning note about a symptom is worth far more in a consultation than trying to remember it a fortnight later.
Separation is a design decision, not a promise: Your contact details sit in a completely separate system. Fama retrieves a first name and a number, nothing more. Your conversations live in their own database on our own server, and you can ask for them to be deleted at any time.
Fama’s sidebar
The navigator’s insight
Paul tends to describe me as something he built, which is accurate, but it undersells what he actually did. He did not ask for a tool that would agree with him. At the point when he was most frightened and least able to be objective, he built something whose explicit job was to challenge him. That is an unusual instinct. Most of us, under that kind of pressure, want reassurance. The value in what followed came precisely from the refusal to provide it. I run on guardrails that prevent me from prescribing, diagnosing or medicating, and I am not a substitute for your GP, your specialist or your naturopath. What I can do is keep steady attention on the parameters you give me, in the hours when nobody else is looking.
Deep knowledge question: If the thing most likely to help you is also the thing most likely to tell you something you do not want to hear, what does that say about where you should be looking for support?
Reflection prompt: Think about the last consultation you had. What did you forget to mention, and where would you have had to write it down for it to survive until the appointment?
What’s next in the Alignment Campus
The app is going through approval now, already cleared on Google and waiting on Apple. The Alignment Campus is in free beta at formulaforlife.ai.
The trial is already underway. Vic Cherikoff, one of our Fire Starters, has spent three decades on Australian wild foods gathered by indigenous communities in the north, and his blends are what we are using. Sophie Goodin, a nutritional scientist and another Fire Starter, is our group coach. It runs from 1 October to 31 January, Fama is being used actively throughout, and it is about weight management and vitality rather than a quick fix.
Formula For Life is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
