Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

OpenAI’s o1 correctly diagnosed 67% of ER patients vs. 50-55% by triage doctors

theguardian.com

Post reply on HN
Print view View original

391–400 of 500 posts

Re: OpenAI’s o1 correctly diagnosed 67% of ER patients vs. 50-55% by triage doctors

#391

Post by Hobadee »

Obviously annecdotal, but a couple years ago my friends kid was sick, and doctors were trying to figure out what was going on. My friend threw the symptoms and test results into ChatGPT, and it said the likely cause was leukemia. A few hours later the doctors handed them an official leukemia diagnosis.

I think AI, like in all other fields, will become a great tool to help augment. Throw the patient data in and get a response and that can be the first thing the doctor checks for, but they shouldn't simply take AI as truth.

P.S. friends kid is doing great - it was caught early enough. They are due to be completely done with treatment in just a couple months!

Re: OpenAI’s o1 correctly diagnosed 67% of ER patients vs. 50-55% by triage doctors

#392

Post by theshrike79 »

troupo wrote: post #378
theshrike79 wrote: post #302

Earlier quoted context omitted.

Step 2 prevents that. It's not there by accident. They need to write down their (initial) diagnosis before the AI answer is shown.

Step 2 doesn't prevent it, because of step 4. AI becomes "upon further testing/examination/review we conclude that..."

And then if the patient isn't cured or has an adverse reaction, the answer given by the doctor in step 2 is examined compared to the post-AI resolution.

If #2 is correct and #4 wrong, the doctor has to answer for stuff.

Re: OpenAI’s o1 correctly diagnosed 67% of ER patients vs. 50-55% by triage doctors

#393

Post by smithoc »

kaikai wrote: post #184
thaumasiotes wrote: post #151

Earlier quoted context omitted.

> My daughter is in her 20s now and is still small -- it's just the way she is. When she goes to see her primary, do you know what their first question is? "When was your last period." Is that supposed to be a problem? How does it connect to the story in your comment? The question seems to be warranted to me, since being underweight can stop you from menstruating. So if you find someone thin and her last period was o…

I’m a normal weight, and get asked the same question. More importantly, I can tell them, “I have a regular cycle” and they WILL NOT take that as an answer. I HAVE to give them a date, and they will ask me to make one up if I can’t remember or want to decline giving them that information. Particularly given the alarming stories of people being prosecuted for having miscarriages, it feels ridiculous. If anything I hope…

> Particularly given the alarming stories of people being prosecuted for having miscarriages

You need to delete your social media accounts and change where you're getting your news from. Nobody is "being prosecuted for having miscarriages". A few people have been investigated for drug abuse during pregnancy which led to the baby's death, which sensationalist news stories twisted into attention-grabbing headlines.

A doctor asking about cycle is just a core piece of diagnostic data like taking blood pressure and temperature, not some conspiracy to harm you.

Re: OpenAI’s o1 correctly diagnosed 67% of ER patients vs. 50-55% by triage doctors

#395

Post by xaxfixho »

OptionOfT wrote: post #45

As a 37 year old male with 2 THRs I'm glad the AI was NOT used in my diagnosis. All the models that I used to look at my x-rays said nothing was wrong, even when adding symptoms. When adding age it said the patient was too young. (I was ~3 months away from wheelchair bound in those x-rays). The worst one was Gemini. Upload an x-ray of just the right hip, and it started to talk about how good the left hip looked like.…

[dead]

Re: OpenAI’s o1 correctly diagnosed 67% of ER patients vs. 50-55% by triage doctors

#396

Post by __mharrison__ »

xoofoog wrote: post #150
pixel_popping wrote: post #25

Earlier quoted context omitted.

I agree with you on this specific study, however, I can't really wrap my head about the fact that doctors will be better than AI models on the long-run. After all, medicine is all about knowledge, experience and intelligence (maybe "pattern recognition"), all those, we must assume that the best AI models (especially ones focusing solely in the medical field) would largely beat large majority of humans (aka doctors),…

I would love to replace my doctors with AI. Today. Please. I have had Long Covid for over a year now, which is a shitty shitty condition. It’s complicated and not super well understood. But you know who understands it way better than any doctor I’ve ever seen? Every AI I’ve talked to about it. Because there is tons of research going on, and the AI is (with minor prompting) fully up to date on all of it. I take treatm…

(disclaimer: not a doctor, sample size one)

My friend with long Covid fatigue (and no taste since late 2020) saw good improvements from nicotine patches.

Re: OpenAI’s o1 correctly diagnosed 67% of ER patients vs. 50-55% by triage doctors

#397

Post by gosub100 »

gpm wrote: post #11

I'd be very very hesitant to trust studies like this. It's very easy to mess up these benchmarks. See for example this recent paper where AI managed to beat radiologists on interpreting x-rays... when the AI didn't even have access to the x-rays: https://arxiv.org/pdf/2603.21687 (on a pre existing "large scale visual question answering benchmark for generalist chest x-ray understanding" that wasn't intentionally mess…

Or the case where supposedly radiologists couldn't see a gorilla in the image [1]

I know it might look like a loss for radiologists, but I don't see it that way. More like you can't trust these studies.

1. https://www.npr.org/sections/health-shots/2013/02/11/1714096...

Re: OpenAI’s o1 correctly diagnosed 67% of ER patients vs. 50-55% by triage doctors

#399

Post by xaxfixho »

Kuyawa wrote: post #233

As a 60yo I developed my own AI medical assistant [1] and I've used it extensively for many conditions, I can't be happier. After analyzing some lab tests it even recommended a marker that was not considered first by the doctor, so yes, it won't replace doctors but it is a very helpful tool for self-diagnosing simple conditions and second opinions. [1] https://mediconsulta.net (DeepSeek)

[dead]

Re: OpenAI’s o1 correctly diagnosed 67% of ER patients vs. 50-55% by triage doctors

#400

Post by n8henrie »

jmalicki wrote: post #347
n8henrie wrote: post #257

Earlier quoted context omitted.

So why not order your own labs? I'm sure you can think of ways to get your own medications if you are sufficiently convinced that this is the best course of action for your health.

Because you can't order many of your own labs, and then insurance won't pay for them.

> you can't order many of your own labs

Really? Which ones?

> insurance won't pay for them

Non sequitur, replacing doctors with AI will not help you pay for the preposterous US healthcare system. Vote!

Post reply on HN