X Prize Foundation Wants AI Physician On Every Smartphone

kkleiner writes "One of the exciting ideas being tossed around recently at the X Prize Foundation is the creation of an Artificial Intelligence physician that you could access from your smartphone. Want to know if that rash on your leg is poison ivy or smallpox? Take a photo of it with your phone and ask the AI. The possibilities are enormous, especially for the billion plus people around the world who live more than a few hours' walk or drive from the nearest doctor." This is one of four X Prizes in planning for the future. The other three are for an AI automobile driver, organ generation through stem cell use, and a deep sea submersible capable of exploring the sea floor.

Nice idea, but...

[miaDWZ]

The thing is going to get killed a week after public release after the AI 'misdiagnosis' someone and they decide to sue.

Re:Hmmm

[hedwards]

That's why some health insurance companies provide a complimentary consulting nurse. Basically somebody that's available around the clock to answer those, should I call 911, go to the emergency room, make an appointment or just ignore it suggestions. Obviously they tend to be a bit action biased as doing a screening over the phone isn't easy, but it does help people make better decisions about what is urgent and what can wait.

Yeah...

[SolitaryMan]

Especially for the billion plus people around the world who live more than a few hours walk or drive from the nearest doctor.

Yeah, to all four of them, who actually have a smartphone

As a non AI physician

[Dunbal]

This is a terrible idea. However I guarantee that the AI algorithm will have a "success" rate of around 85%, since that is the rate at which illnesses spontaneously cure themselves. This rate is why homeopathy, snake oil salesmen, faith healers and all other forms of shamans and charlatans manage to convince people of their effectiveness. Too bad that 15% of the patients will suffer permanent disability or die using these methods. That's the part of the statistic we doctors manage to concentrate on and improve, the 15% that really need help...

Re:"AI"

[John Hasler]

> Image recognition is certainly considered a part of "AI" research...

According to the anti-AI crowd AI is whatever it is that computers can't do yet. There was a time when all agreed that a machine playing a credible game of chess would constitute proof that AI had arrived but now defeating grand masters is "mere computation".

Re:Nice idea, but...

[fuzzyfuzzyfungus]

The difference is, "AI" systems are neither advanced enough, nor sufficiently embedded in the public mind to be seen as distinct from their creators.

If my doctor fucks up, I sue the doctor(and possibly the hospital, if they really should have known about his habit of bringing a hip-flask into the surgical suite). I don't sue the medical school that "produced" the doctor.

If an AI expert system fucks up, I don't sue the AI, I sue the company that built the AI. Thus, unless the company wishes to carry some sort of novel "aggregate malpractice insurance" covering all the products it sells, selling that product would be wildly uneconomic.

The same basic conditions apply with the various proposals to automate cars. Humans are shit drivers. It is easy to believe that machines will be able to do better within the decade, and do better cheaply not too long thereafter. However, courts have an easy time with the idea that humans suck, and are morally responsible. Therefore, unless a definite defect in the vehicle is detected, the driver is usually blamed. Even if AI drivers cut vehicular crashes by 80%, saving thousands of lives a year, the companies building them would be sued into oblivion.

Re:Internet hypochondria is already a phenomenon

[Kilrah_il]

It's even more problematic than you think. As a doctor, many times I have patients coming to me with "agonizing abdominal pain" - they are sure it's appendicitis. If you check their stomach or ask them "does it hurt here?" they jump and cry and wail, etc. etc. But if I start talking to them on other subjects (What do you do in your life? Are you married? Children?) and get them diverted, I find out many times that they "forget" about their pain and the stomach is as soft and non-tender as can be.
This is one of the reasons that no app can replace a physical exam by a doctor. You need the doctor-patient relationship to strip away the anxiety and find out the true magnitude of the symptoms. So, yes, an app like the one in TFA could be nice as a handy reference, but nothing beats the good-old face-to-face meeting.

Re:Internet hypochondria is already a phenomenon

[bitflip]

I was having some pain in my chest a couple of years back, and did some research on the internet, but none of the symptoms fit. So, I went to my doctor, and told him I'd done some research, but it didn't seem like a heart problem. When he looked surprised, I asked him if I was the only person who'd said they'd looked stuff up on the internet, and decided nothing was wrong. He said, pretty much, yea.

(turned out to be a minor stomach problem, all fixed now - thanks doc!)

Really, AI on a smartphone is the best solution?

[haggais]

If I already have my smartphone with me, or indeed any phone with camera, and I can take pictures of my skin condition (or whatever), and write or talk about it, perhaps I could SEND the photos and comments to a REAL LIVE doctor, even one who works more than a few hours drive away, but within phone coverage, and get his NATURAL INTELLIGENCE comments, instructions, etc. over the phone? Is AI going to be better and cheaper than human doctors any time soon? And assuming it does get that way, why does it need to run on my smartphone? Through the magic that is the interweb (of which I believe smartphones are a part), would it not be easier to send the data to a slightly more powerful SERVER, and get my AI GP's advice from there? Slapping "AI" and "smartphone" on a problem does not make for a brilliant futuristic solution.