A.I. Could Worsen Health Disparities

[SummerMadness]

A.I. Could Worsen Health Disparities

Artificial intelligence is beginning to meet (and sometimes exceed) assessments by doctors in various clinical situations. A.I. can now diagnose skin cancer like dermatologists, seizures like neurologists, and diabetic retinopathy like ophthalmologists. Algorithms are being developed to predict which patients will get diarrhea or end up in the ICU, and the FDA recently approved the first machine learning algorithm to measure how much blood flows through the heart — a tedious, time-consuming calculation traditionally done by cardiologists.

It’s enough to make doctors like myself wonder why we spent a decade in medical training learning the art of diagnosis and treatment.

One of the big problems with reliance on machine learning algorithms is placing too much faith in the results without regard for the training set used. If the data used to build the AI is flawed, then your outcomes will be still flawed. And unlike humans, people are more trustworthy of computer decisions because it is believed that is unbiased, but that is not the case at times:

It turns out Amazon’s AI hiring tool discriminated against women
When Algorithms Discriminate
Artificial Intelligence’s White Guy Problem
Machine Bias

 

[Dave-W]

If the data used to build the AI is flawed, then your outcomes will be still flawed. And unlike humans, people are more trustworthy of computer decisions because it is believed that is unbiased, but that is not the case at times:

Indeed. Systemic bias often goes unrecognized and gets written into code.

 

[yeshuaslavejeff]

One of the big problems with reliance on machine learning algorithms is placing too much faith in the results without regard for the training set used. If the data used to build the AI is flawed, then your outcomes will be still flawed. And unlike humans, people are more trustworthy of computer decisions because it is believed that is unbiased, but that is not the case at times:

As Yahweh's Word says, people who place their trust in things instead of in Yahweh are cursed (much worse than simple health problems, or anything caused by any discrimination).

So don't expect to find equality or justice or rightness or righteousness from men or from women in any group - it is misplaced 'trust' in any such case.

Trying to make any group of men or women unbiased fails, and is a waste of time and energy, and leads always to dis-satisfaction, turmoil, and more injustice and unfairness.

Trust God. Let God Accomplish all that is needed for Salvation, every day, concerning everything, always. Do what God Says, with peace and joy and righteousness as His child, in line and perfect harmony with all of His Word.

 

[Aaron112]

Indeed. Systemic bias often goes unrecognized and gets written into code.

What if, as is often the case anywhere, it is on purpose ?

 

[ThatRobGuy]

Technology can have flaws in the "working out the bugs" phase.. Overall, AI can be a net positive if applied properly in the medical field.

To say that people are "more trustworthy" than computer at making certain types of observational assessments is a stretch. An AI algorithm specifically programmed to seek out cancer cells in a high-contrast PET scan would've likely done better than my dad's first oncologist.

 

[ThatRobGuy]

What if, as is often the case anywhere, it is on purpose ?

I don't think that would be the case.

As someone who's a 17 year vet in the IT field, there aren't a whole lot of "far-right" people working in silicon valley.

As a whole, IT people (especially the high ranking ones that would be working on the types of high profile intricate projects discussed here) are usually pretty pragmatic and let data drive their decisions more than personal biases and emotion.