AI might worsen diagnostics for physicians

Earlier studies have also showed mixed results for Al-helped diagnosis of Gl-polyps.

Are we seeing the same effect in doctors as in programmers? Where over-trust or confidence in the tools makes us make more mistakes.

https://www.thelancet.com/journals /langas/article/PIIS2468-1253(25)00133-5 /abstract

▲

Frieren 16 hours ago | parent | next [-]

Any capability that we lose as a society is going to be hard to get back.

I understand that AI companies want to embed AI in everything and make society dependent on it. Like cigarette brands AI companies want a world addicted to their technology.

But as a society we should be more careful before throwing well known methods for fancy new tech. Even more important is to take a step back and think when AI companies are pushing for "not lose the AI race", "speed up adoption", and the rest of FOMO scaremongering.

Stop and think is what I expect from professionals even when CEOs push for AI as fast as possible.

▲

jf22 7 hours ago | parent | next [-]

Has technology made us lose any capabilities so far?

Everywhere I look there is someone preserving 16th century cave painting techniques or hand tanning beavers and then writing guides that surpass whatever knowledge there was centuries ago.

	▲	sejje 6 hours ago \| parent [-]
		I'm kind-of into old techniques and crafting styles. I live in the forest, so it kind of lends itself to that. I agree with you. I have piles of books detailing very old crafts like processing game, tanning hides, trapping, hunting. Knapping stone tools, weaving baskets, raising pole barns, distilling whiskey at home, making clothes, mending clothes, herbal/wild medicines. The foxfire books being a great example of how we preserve this knowledge. Almost anything modern we've learned is going to be well-documented IMO. We may not be referencing those documents any longer, but they exist, and will exist for a long time. And also agree that modern methods tend to be better, or at least the best version of the old methods is widely known, thanks to youtube etc. Primitive Technology on youtube, for instance, was able to launch himself into the iron age without bringing any modern tools along. But the access to modern informational sources makes him a primitive powerhouse.

▲

jxntb73 15 hours ago | parent | prev [-]

Like past bubbles, it’s usefulness is being marketed into existence.

▲

arisAlexis 17 hours ago | parent | prev | next [-]

Actually previous studies found AI augmented diagnosis of physicians. This just says that if you use it frequently you lose some of your skills. We should care only about how to make diagnosis better, not who makes it.

	▲	arisAlexis 15 hours ago \| parent [-]
		It appears that facts make people angry here, society is not ready for AI

▲

reify 15 hours ago | parent | prev [-]

Then the overdiagnosis begins.

Half the polulation of the world getting their Gallbladders removed because AI said so.

Gallbladder polyps are abnormal growths in the lining of the gallbladder wall. Some are tumors, some are scar tissue, and most are cholesterol deposits.

screening costs money, it does not save lives.

Same with Colon cancer and Prostate cancer screening.

£ $ Millions spent, hardly anyone saved.

▲

AznHisoka 13 hours ago | parent [-]

Total idiot when it comes to medical stuff, but if the screening shows a tumor, and you catch it early on, wouldnt that be beneficial? And if its not, then you gain peace of mind

	▲	trillic 7 hours ago \| parent [-]
		The statisticians at the insurance company are concerned that the sum of the risk of unnecessarily removing 10 gallbladders is greater than missing a single tumor in early-stages.