Aurornis
12 hours ago
There is a lot of authoritative sounding text here but no real explanation of how this huge process solved the mystery fatigue. In fact she says in the middle that the specialist she saw was more valuable. Toward the end she mentions experiments that were run, but the heavy hitters appear to be from specialists adjusting the medication related to her tumor treatment and a real person reviewing her food diary and noticing she was in a caloric deficit.
For all of the writing about collecting data and using AI, she never explains what the AI did for her. Maybe it suggested trying iron supplements because her ferritin was slightly low on a blood test? I hope it explained that these blood tests can have transient highs and lows and that excess iron from supplements can have major negative health effects if not monitored.
For all of the lead up about using AI, there was never really a delivery where the AI was used to solve the mystery fatigue.
I’m glad this person is feeling better, but this post reminds me of those LinkedIn posts where someone is trying really hard to show their AI skills but used an example where the AI didn’t contribute to or accelerate the outcome.
haldujai
11 hours ago
My gut as a physician says AI will revolutionize primary care the most and the premise of AI having more specialized knowledge than a PCP holds water - I think the future of primary care is AI equipped midlevel providers - so I was very excited by the intro only to be let down. A lot of buzz about nothing sadly.
The only takeaway here is logging helps detect patterns, we already knew that.
OutOfHere
11 hours ago
For simple things, if an AI (combined with blood tests) can do what the doctor can do, then technically I don't need the doctor. I also don't need the friction that comes in working with a doctor, considering the AI is close to frictionless.
blooalien
11 hours ago
> ... "considering the AI is close to frictionless" ...
I feel like this bit is more of a problem than many folks realize. I know that people see "frictionless" as a desirable trait, but for many of the things folks are using AI for these days, you're gonna want "push-back" sometimes. Especially in cases of health care and computer programming, there are times when the user is straight-up wrong about what they're trying to do, say, or believe, and the last thing you wanna hear when your life or job or the lives or jobs of others are on the line and you're about to make a huge mistake is "You're absolutely right! Let me get right on doing that for you!"
OutOfHere
11 hours ago
You're absolutely right! What I do is have the LLM rank and score candidates for me given my situation. This helps me focus on what's more likely versus less likely. Blood tests are not free; they're in fact fairly expensive in aggregate, so this filtering has been essential.
haldujai
9 hours ago
> technically I don't need the doctor
That’s the goal for doctors too. It would be great to get simple things off the system.
But I think the more realistic intermediate step is a trained person cheaper than a doctor - nurse, PA, etc - aided by AI. The current generation of agentic AI doesn’t seem to be there yet and is too agreeable from RL.
“you’re probably fine sleep it off combined with: drink more water, eat healthier, exercise more, sleep better, consume less alcohol and quit smoking/vaping” +/- “we’ll check some labs to make sure” is the correct answer for probably 95%+ of encounters so it’s not hard for an automated system to handle most simple things, even without AI.
OutOfHere
an hour ago
Is it really the goal for doctors? I don't believe it. They run an exclusive protectionist racket.
See https://www.msn.com/en-us/health/other/this-state-is-testing...
Title: This state is testing out AI doctors—and actual doctors aren’t happy about it (Wall Street Journal)
New York state residents get the worst of it with the state bending over backwards to take away the people's rights.
Also, let's not confuse independence with AI. Independence is where I can do some extra bloodwork and diagnostics to satisfy my well-being. I don't need an AI limiting what I investigate either.
yarekt
11 hours ago
I get it, and it draws parallels with debugging and incident response in software dev. Your best tool is good data, if you have more data, you’re able to make good judgement in short time. If AI helps you collect good data about your condition, then presenting that to a clinician is certainly better than “it hurts, idk” because there’s only so much they can do at the time
Aurornis
11 hours ago
But where does that happen in the article?
Even for the part where she discovers she was in a caloric deficit and that carbs helped, the AI didn't seem to have a role:
> Figuring out what to track is an iterative process. At first, I logged only daily calories. But when I noticed that a small amount of carbs sometimes helped me recover faster mid-episode, I tracked carbs on the same spreadsheet.
The carb realization happened before tracking carbs. And why did it take a specialist to notice she was in a 300 calorie deficit? That seems like something that any AI process would trivially notice.
I don't know. This article is puzzling. I think using AI to analyze health records is interesting, but this article didn't really have anything supporting that other than to mention that she put everything into Claude on the side.