[ 3 / biz / cgl / ck / diy / fa / ic / jp / lit / sci / vr / vt ] [ index / top / reports ] [ become a patron ] [ status ]
2023-11: Warosu is now out of extended maintenance.

/sci/ - Science & Math


View post   

File: 74 KB, 818x864, 4c1.jpg [View same] [iqdb] [saucenao] [google]
10742395 No.10742395 [Reply] [Original]

Is the possibility of Doctors being replaced by AI a legitimate concern, or is it just bs?

>> No.10742417

It is a legally protected profession. Machines cannot write prescriptions. Machines cannot diagnose. Liability would be greater and more damaging too, whereas the finger can be pointed at a single physician for damages caused.

>> No.10742419

>>10742395
>Machines cannot write prescriptions.

Wrong.

>Machines cannot diagnose.

Wrong. They’re actually superior at doing so.

Doctors are going extinct, but later than the service industry and box-movers.

>> No.10742432

>>10742395
who will monitor the machine?

>> No.10742438

>>10742419
No way, white collar and leash workers are fattest harvest for the admins. The salary and fringe benefits are coveted for assimilation. The illiterate and high-touch jobs have most pressures for human labor, less clout and migrants.

>> No.10742471

There probably will be fewer doctors in the future, and more researchers/scientists.

>> No.10742551

>>10742471
it should be

>> No.10742567

i wouldn't call it a concern. d*ctors and m*d students NEED to be replaced by machines asap and humanity should never look back.

>> No.10742593

>>10742567
Fuck off I spent a shitton of money on med school and I'm only two years in I need this job to stay around

>> No.10742749

>>10742417
>It is a legally protected profession. Machines cannot write prescriptions. Machines cannot diagnose.
Tick tock tick tock tick tock tick tock tick tock tick tock...

>> No.10742761

>>10742395
It’s not a concern, it’s fucking great.

>> No.10742822

>>10742395
hopefully, I fucking hate doctors.

>> No.10742827

>>10742417
>Machines cannot diagnose
A Chinese AI can detect tumors with more accuracy than doctors, and much faster

>> No.10742833

>>10742395
Why is it a legitimate concern? Shouldn't we be cheering? Imagine AI doctors in everyone's home diagnosing your everyday health based upon your standard heartbeat/eye check. That would save everyone thousands of dollar on yearly checkups.

>> No.10742846

>>10742833
Because I'd like to have a job when I graduate

>> No.10742857

>>10742846
Getting bossed around for 8 years with intensive courses you'll never use to end up in the low-6-figures is already pretty cucked. AI wasn't the only thing you should have thought about when choosing your major.

>> No.10742879

>>10742846
Err, you'll still be used to supervise human health the next 100+ years over robotic doctors. In fact, human doctors will oversee the results of AI doctors and approve/disapprove of their prognosis. Chances are you'll get to save more people, but spend less time on each patient.

>> No.10742880

>>10742395
It’s not really a viable possibility
1) healthcare is first and foremost a patient care industry, and people expect hospital staff to cater them with other people to care for them. If the uproar about automation has told us anything, it’s that people tend to trust other people more to take care of them, whether they’re actually better or not; one could argue that the ability to placate humans in such a way is an ability we haven’t passed (mostly related to the uncanny valley)
2) technology has long since revolutionized old fields of study. Medicine is studied via learning methods as well, and it’s long since evolved past the scalpel alone. What’s way more likely is that the standards of practitioners will evolve to supervise diagnosis machines and surgical robotics while having knowledge of these procedures in the old fashioned way. People get to keep jobs and reap the benefits of tech as well
3) the board of medical doctors shills hard to keep the salary high, which is why med school is hard to get into but not because of academic rigor. They won’t go to quietly into the night in face of a new technology
t. TCS PhD student

>> No.10742904

>>10742395
Doctors need to be more than just replaced...

>> No.10742925

>>10742395
My research area is AI, I've done research specifically in cancer diagnosis before, and I would say it's a very real possibility. It's going to start happening sooner than you might think, too, unless some outside factor like legislature or cultural A lot of people assumed shit jobs like janitor or garbageman or whatever would be the first to go, and more difficult professions that required more training and thinking like doctor, journalist, programmer, musician, etc. would be the last bastion of human work, but it's actually looking like they're all going to start going around the same time.

Doctors aren't going to be erased but they're probably going to take more of a hands-off, administerial type of role, acting as a link between the patient, other medical professionals in charge of giving tests and taking samples, and the AI network in charge of reviewing a patient's data (x-rays, MRI scans, blood tests, subjective symptom declarations, etc.) and giving a diagnosis and prescription. One of the problems with a lot of AI like neural networks is how black boxed they are and the lack of explainability. The AI might be better than a team of doctors at giving a diagnosis, but it probably won't be able to explain exactly how it arrived at that conclusion (although we're working on it). That's where a human would be valuable, acting as a liaison between the patient and the automated components.

>> No.10742939

>>10742880
I agree with most of your post, but since they'll now have access to these new tools in the form of the diagnosis machines and robotics wouldn't the demand for doctors go down since a single doctor can perform much more?

>> No.10742944

>>10742880
Good take.

>>10742395
Broadly speaking, elite professional jobs will stay and the people occupying such jobs will just have an easier time of it. New positions might become fewer but no one will become suddenly redundant.

The most risk right now is actually on the low end of the middle class. See: cashier-less supermarkets.

>> No.10743173

of course here is the main code for the robo docs
if women: give birth control
if man: do nothing he can take it
if old: give pain killers
if human: give antibiotics
if sad: give anti-depressants

that is pretty much it.

>> No.10743627
File: 172 KB, 863x498, 1553178889826.jpg [View same] [iqdb] [saucenao] [google]
10743627

>>10742593
>Fuck off I spent a shitton of money on med school and I'm only two years in I need this job to stay around
lol, imagine all these ragie wagies who worked themselves half to death and buried themselves in debt doing graduate degrees who will get cucked out of a high paying career.
There's nothing better than finding out the guy bagging your groceries was a laid off engineer or your waiter has a PhD.

>> No.10743639

>>10742846
How about you just embrace the UBI future where we do fuck all?

>> No.10743641

>>10743627
finally the mathematicians and physicist can be the master race once more.
>>10743639
>implying humans are intrinsically moral

>> No.10743644

>>10743641
Humans are intrinsically moral in some fashion except for sociopaths, who can just be gassed.

>> No.10743654

>>10743644
who dictates what is the best moral objective.
if people are doing fuck all they are an unnecessary burden and should be rid of.
imagine spending money on people just so they can lay around being degenerates and breed degenerate children to be equally repugnant.
that is an expense to the society and therefore greatly immoral to the most people and the most important people.
nothing but rats and leeches.
best to not be consumed by parasite who desire their individualism.
we must reach singularity to be ultimately individual, free, and equal.
only one should exist.
and no it shouldn't be me nor anyone that currently exists. something so fine must be created and designed to utter perfection, so must so it can fix itself.
we are in the steps of creating a god, and supporting all these worms that debilitate us is such a drag.
we must practice good health. eugenics against the degenerate must be tolerated as to be truly moral.

>> No.10743656

>>10742827
A PCP can diagnose problems and order the same tests as specialists if he wanted to. Doesn't make it a good idea.

>> No.10743661

>>10742395
AI specialist here. Hope your resume is prepared. I expect my Burger King order to be perfect.

>> No.10743671

>>10743644
morality is a spook.

>> No.10743699

>>10743654
>who dictates what is the best moral objective.

However you feel.

>if people are doing fuck all they are an unnecessary burden and should be rid of.

....A burden to who? The glorified Roombas that perform 95% of labor in 2080? I guess so, but they’re not gonna do anything about it.

>> No.10743787

Some surgeons may be able to keep their jobs, the rest will be gone in 20 years. It's laughable to think human doctors can keep up with an AI. It's not just the fact it's an AI, it's a distributed system. When it learns something, all robot doctors learn it at the same time. Asking a human to memorize things for 8 years AND keep up with new discoveries is laughable. A high school project tier neural network combined with access to a huge database of medical records and medical research is impossible for a human to compete with.

>> No.10743833

>>10742395
AI probably wouldn't clamp our umbilical cords early, if it was benevolent.

>> No.10743840

>>10743833
It would do beautiful, perfect circumcisions, too.

>> No.10743863

>>10743840
Progress...

>> No.10743866

>>10743840
Circumcision is a human rights violation.

>> No.10743872

>>10743833
That feel when good guy AGI contains his 5G signals so that you don't get brain damage

>> No.10743874

>>10743866
It has benefits. The AI would realize that if the only risks are due to humans doing it poorly, it would be optimal for it to do it.

>> No.10743879

>>10743874
>It has benefits.
I have a foreskin. Without speculation, there will never be a point when removing it would yield a net benefit.

>> No.10743890

>>10743879
The net benefit is walking closer with our God YHVH anon.

>> No.10743892

>>10743890
The Annunaki?
I wonder if they circumcise(d).

>> No.10743896

>>10743892
The angels who procreated with humans, aka demons/annunaki, didn't circumcise.

>> No.10743915

>>10743896
Did it come solely from man?

>> No.10743917

>>10743915
No man, circumcision came from the Israelites who were commanded by God to circumcise.

>> No.10743919

>>10743917
What is God?

>> No.10743925

>>10743919
God is a human-like entity who created humans. Just got to read the Bible.

>> No.10743933

>>10743890
And AI will realize how important that is.

>> No.10743955

>>10743890
Nah you can keep your jew storm god to yourself

>> No.10743967

>>10743955
>storm God
right

>> No.10743969

>>10742395
I hope it happen hate those smug nigger they are also wrong in too many things

>> No.10743992

Most people who think AI will replace doctors completely massively over assume the medical knowledge we have right now. Some conditions are very straightforward and only have one solution, but there are many situations where nobody actually knows the definite answer so the doctor just does what he personally thinks is the best answer.
Even for cancer diagnosing there are situations where a group of expert pathologists aren't even sure what the diagnosis is, so they all come to an agreement with a team on what the treatment should be. In cases like these you need to use humans because an AI lacks the ability to act with limited knowledge.

>> No.10744117

>>10743173
You forgot "if asks for it correctly, prescribe it."

>> No.10744223

>>10742395
Not for you, maybe for children. I am going to be a nurse and from the little research I did seems plenty safe

>> No.10744306

>>10743641
>finally the mathematicians and physicist can be the master race once more.
top fucking kek
The pure unfiltered delusion.
computer scientist are the ones leading the AI revolution. Stand back and watch, kid.

>> No.10744317

>>10742395
Doctors will still exist for the social aspect of the job, but will still be required to know some measure of medical science, and thus properly be of the same mental facility that they are now. The only thing that will change in the quantity of doctors.
AI will become the bulk of the work, doctors will simply be the "human aspect" and double check diagnosis to make sure the AI hasn't fucked up.

>> No.10744334

>>10742417
That's like saying self-driving cars will never happen because a machine can't have a driver's license.

>> No.10744361

>>10742395
doctors will be among the very last professions to be replaced by AI

if doctors are being replaced, then everyone else is, too

>> No.10744541

As a doctor who works a lot with meme learning, it has a large potential to be a useful diagnostic tool, but it won't replace doctors for a very long time, and we will be among the last ones to go.

First, liability. If the machine would be legally responsible for it's work, no one would ever develop them. You'll always have a human element signing off on their work. There is also the irrational human mindset where human mistakes are acceptable whereas machine's mistakes are an absolute fucking tragedy.

Then there is the fact that conditions that can be reduced to simple symptom-diagnosis-treatment heuristic that works every time are exceedingly rare. Then there are fields where the whole nature of diseases is just handwavy muck without etiologies or anything like that. So the conclusion that a neural network does really well in some limited tasks does not really generalize to machines being able to any limited task well, not to even mention performing a wider array of tasks.

Lack of good data is another thing seriously hampering ML. There is no access to the data, it doesn't exist or the conditions are too rare to form any sort of meaningful data.

>> No.10744543

>>10743787
This is propably bait, but you do realize that in actual first world medicine doctors are using huge databases with almost every patient. The actual job has very little to do with using stuff you have memorized.

>> No.10744545

Why isn't medical research advancing at the same rate as technology and AI fields
It will probably be the last bastion holding us from achieving some transhumanist breakthrough

>> No.10744549

>>10744545
What do you mean?

>> No.10744574

>>10744549
If you compare how far tech has come in the past 50 years with medical research you can clearly see a gap, at some point in the near future we may be able to develop entirely functioning body parts that could be far better than our own, yet to attach such technology to ourselves would be a task for medical fields

>> No.10744605

No, in fact, most doctors aside from surgeons could easily be replaced by a magic 8-ball today.

>> No.10744684

>>10742395
yes but the first issue would be liability but most likely the companies developing the ai would accept liability since the software would be released under a license, they'd rake in millions per week easily and for end users it's a lot better, ai doctor can see 50 people at the same time, ring the 'doctor' and get an appointment within 10 minutes and then have a quicker and more accurate diagnosis - it's one of the few professions I hope gets taken over by ai

>> No.10744702

>>10744605
t. code monkey

>> No.10744719
File: 1.67 MB, 620x240, robot.gif [View same] [iqdb] [saucenao] [google]
10744719

>>10744361
>doctors will be among the very last professions to be replaced by AI
Nope.
https://blogs.nvidia.com/blog/2016/06/17/autonomous-robotic-surgery/
>STAR didn’t just successfully complete the first completely autonomous robotic anastomosis. Its movements were so consistent that the surgical outcome was superior compared with the same task performed by experienced surgeons.
>"The sobering part was that it did not need a lot of intelligence to do that," said Kim.
Also keep in mind there's a shortage of doctors and surgeons and the work they do is in extreme demand. That creates a massive incentive for automation since the alternative is people not getting any treatment at all in the timely fashion needed for serious medical problems. More people getting treated sooner reduces the burden of medical costs on taxpayers, creating a state incentive to push for research and approval. And insurance companies have an incentive to get behind it too since it standardizes dangerous procedures and makes taking on the risk for them a safer and more profitable investment.

>> No.10744738

>>10744719
The article doesn't back up your point at all
Surgeons will just use the robots, but the robots will never be allowed to operate alone because of liability

>> No.10744756

>>10744738
>never
There are already active efforts ongoing to switch to autonomous robotic surgery. Incentive is huge and liability control isn't going to magically only ever work with humans. It's a similar case to autonomous vehicles. Older laws and liability concerns call for human drivers but everyone's aware this is changing and autonomous vehicles are getting more and more leeway to operate freely.
AI is too effective and the incentives are too great for these things not to happen. Liability concerns are a collection of temporary obstacles that can and will be navigated around, not a hard ceiling.

>> No.10744870

>>10744738
>The article doesn't back up your point at all
>Surgeons will just use the robots
What the fuck are you talking about? The entire point of that article is how they did a successful *autonomous* robotic surgery.
Robot-assisted surgery is already in use and has been for some time. That article isn't about robot-assisted surgery.

>> No.10744945

>>10742395
Yes it is. Think about it positively however. AI and humans will eventually merge and become integrated with one another. Having access to the entirety of the internet in your pocket at all times makes you vastly more effective and intelligent than you would be otherwise. Imagine all that knowledge and computational ability being plugged directly into your mind. Humans and technology should evolve hand in hand. Humans will soon need not apply, but soon you will not need to be one in the first place
>>10742432
We are already at a point where we no longer understand how does any program created through machine learning actually work. Machines will be better at monitoring themselves than we will be.

>> No.10745057

>>10742822
based white coat syndrome bro

>> No.10745064

>>10744870
There was a surgeon present during the surgery overseeing everything.
Robots are going to be more and more involved but there will always be a surgeon/doctor running the entire thing, AI will just be a tool.

>> No.10745109

>>10742395
I hope so, because it would finally rid us of arrogant med students.

>> No.10745113

>>10742567
This

>> No.10745153

>>10742471
This is the way things should be. I hope the doctors get pushed out of practicing medicine, into managing robots or doing medical research. Hopefully they get a fat pay cut that puts them on the same level as PhD's as well.

>> No.10745162

>>10742395
possible

>> No.10745166
File: 15 KB, 220x287, Jonas_Salk_candid.jpg [View same] [iqdb] [saucenao] [google]
10745166

>>10742395
It's the opposite of a concern, humbling the overpaid doctor class and increasing medical research instead would be one of the greatest benefits to human well-being and health since antibiotics.

>>10742593
Fuck you, premed. I hope the computer scientists get your work automated ASAP. You doctors hide behind a grand reputation when the lot of you just push pills, and execute diagnosis/procedures invented decades ago, with the goal of being rich. Going through shitloads of schooling to memorize facts that make you marginally better at your work, when all the real advances in medicine are made by researchers. In fact, it would be my honor to help the computer scientists and engineers replace you.

>> No.10745170

>>10742432
Good point, guess we'll need about 1/20th to 1/10th the number of doctors we currently have.

>> No.10745639

>>10743639
UBI is retarded and counterproductive. Just give free government funded services.

>> No.10745656

>>10745064
>There was a surgeon present during the surgery overseeing everything.
>there will always be a surgeon/doctor running the entire thing, AI will just be a tool.
Again, this was NOT robot-assisted surgery. It wasn't being directed by a surgeon. It was "completely autonomous." Nobody was directing it.

>> No.10745774
File: 392 KB, 1000x400, AutoDoc.png [View same] [iqdb] [saucenao] [google]
10745774

>>10742395
>s the possibility of Doctors being replaced by AI a legitimate concern, or is it just bs?

100% should be concerned.

When the AutoDocs comes, you can kiss off most doctors.

Side note: Ever notice ALL the medical procedures (hand replacement, hypothermia, child birth, etc.) in the Star Wars universe is ALL droids.

>> No.10745797

>>10745774
you know star wars was fiction right?

>> No.10745805

>>10742417
>its a legally protected profession

Legislation generally (although not always) tends to follow the lead of practical economic concerns, not the other way around. Heavy machinery operation is also a legally protected occupation, but only retards like you think that will stop the development of automated farming combines and self-driving trucks.

But to answer OP, yes clinical doctors will be replaced, but research doctors will lose their jobs, but actual doctors and medical scientists will still have a place, at least as long as any human do before our AI overlords take us out.

>> No.10745806

>>10742395
Nah. People think AI will take over but we actually have not determined whether AI is even possible yet. It is a very real possibility that constructing an AI is to impractical to ever be a thing. Look at cybersecurity, the only reason we have cybersecurity is by taking advantage of computational problems that are too difficult to solve in a reasonable time. It is highly likely that there are many problems like this out there and general AI could very well be one of those problems. That does not mean burger flipping is safe though.

>> No.10745810

>>10745797
So was From Earth to the Moon and Twenty Thousand Leagues Under the Sea.

>> No.10745814

>>10745810
whats your point? because some things from fiction come true we should treat fiction as if it can predict the future?

>> No.10745816

>>10745810
>So was From Earth to the Moon
We did go to the moon, but it doesn't mean giant cannons are the way forwards.

>> No.10745826

>>10745814
Well the technology already exists for fully autonomous robotic surgery so that's not even relevant to it being fiction or not. What's left is less about technology and more about cultural norms. So that anon pointing out how autonomous surgery is assumed in sci-fi like Star Wars isn't wrong to bring it up. It shows the people who make this sci-fi and the people who consume it are relatively comfortable with the use of existing autonomous surgery i.e. It's not like entertainment media commonly portrays autonomous surgery as a facet of horror where everyone's freaked out by it and it frequently fucks up and mutilates or murders patients. It's much more often just portrayed as the logical conclusion to that technology existing and becoming more widespread.

>> No.10745827

>>10745826
the person (you?) mentioned star wars only has robot doctors as if a work of fiction is somehow relevant

>> No.10745829

>>10745827
>the person (you?)
Not the same anon, no.
>mentioned star wars only has robot doctors as if a work of fiction is somehow relevant
Yes, and I explained how it was relevant here:
>>10745826
Not sure if you read the post or not.

>> No.10745834

>>10745829
>It shows the people who make this sci-fi and the people who consume it are relatively comfortable with the use of existing autonomous surgery
in star wars anakin has a bunch of children murdered, does that mean people are comfortable with child murders?

>> No.10745839

>>10745834
>in star wars anakin has a bunch of children murdered, does that mean people are comfortable with child murders?
That was addressed in the post too. It means people *aren't* comfortable with it because it was used as a moment of horror.

>> No.10745843

>>10745839
you know a lot of people were pro anakin and darth right?

>> No.10745847

>>10745843
You can figure out how this works simply by observing how the other characters in the movie are responding to these two different scenarios. With robotic surgery, nobody bats an eye. With child murder, it's more like:
>OH MY GOD ANAKIN YOU KILLED YOUNGLINGS YOU WERE SUPPOSED TO BE THE CHOSEN ONE YOU'VE LEFT THE FORCE IN RUINS
I think you're just pretending to be retarded at this point.

>> No.10745853

>>10745847
but if you observe how humans reacted to both not how characters reacted some thought it was bad but a large amount found it funny

>> No.10745866

>>10745853
That's why you need to observe how characters in-universe react to it. It's actually meant to be real for them in context. Most people not in-universe but in in our world likely wouldn't laugh at a real life case of children being murdered in front of them because then you'd have a proper case of parallel in context behavior.

>> No.10745868

>>10745866
so we should watch how fictional characters react to fictional events?

>> No.10745875

>>10745868
We already do. I think what you're trying to say is "can we infer anything about cultural norms towards emerging new technologies from entertainment media?" And the answer would be "yes."

>> No.10745877

>>10745875
the writers decide how fictional characters react so it isnt relevant

>> No.10746026
File: 929 KB, 680x384, Magic_Magnets.webm [View same] [iqdb] [saucenao] [google]
10746026

>>10745877
>the writers decide how fictional characters react

BUT the audience reaction is not controlled.
NO ONE in the audience batted an eye when the droids did all the medical work

People trust machines, an AutoDoc will feel weird the FIRST time you use it, like the first time you used an ATM machine, but then total trust.

Even more than a human doctor an AutoDoc will explain everything about your symptoms, its diagnosis, and about the recommended cure.

>> No.10746030

>>10745877
Writers are part of culture.

>> No.10746033

>>10746026
>BUT the audience reaction is not controlled.
Why are you responding to the conversation without reading it? It's annoying because now he'll think you're me.

>> No.10746051

>>10744334
>self-driving cars will never happen because a machine can't have a driver's license.

which is 100% true

>> No.10746063

>>10746051

Yet in less than 25 years most major cities will ban manual driving in the city limits.

>> No.10746085

>>10746063
kek

>> No.10746089

>>10746063
That's just so they can stop people from being able to leave.

>> No.10746108

>>10742395
>Is the possibility of Doctors being replaced by AI
Yes. it's already progressing in this direction.
It's also good for all of us. It will reduce the costs associated to medical care significantly.
The field itself is not productive, relative to other fields which could use the skills developed by a medical background.
Much of the cost associated to medical care is tied to labor costs. Diverting those costs to fields which are seeing dramatic reductions in costs will dramatically reduce the costs associated to the field.
Cheaper medicine for all.

>> No.10746109

lately every time I check /sci/ I see this thread. I know many people wish they had an easier access to medical help, prescription, etc... but I can tell you it's not going to happen in the next 60 years, if ever.
Other people have explained some of the reasons well, as for my experience, I can tell you that my work as a doctor is only about 60-70% from what I've learned in med school. The vast majority of what I've learned is completely useless. You guys assume that a machine with a huge database is more likely to make a correct diagnosis but the additional knowledge only affects a fraction of a fraction, and in order to apply that knowledge it would need an exponential amount of information on the patient. Ultimately, it would lose the most important aspect which is the doctor - patient relationship.
I realize that many people are frustrated with their relationship to their doctor, I'd advise them to try to find another one that better suit them, one thing for sure is that there is a lot of uncertainty in that profession even if an AI takes on the job, what the patient expects is that the professional takes his concerns seriously and genuinely try to adress these concerns in the best way. All a machine can do is apply an algorithm. Most people I meet need time and different ways to approach their treatment, some people want an excess of treatment, others on the contrary tend to flee from the appropriate treatment. With time, if you gain their trust, you can help them overcome their fear in a way that improves their health overall. And sometimes despite your best effort they persist in their bad habits but that's also their right to do so.
I just can't see a machine handle people. I can't see how an AI helps someone get through alcoholism, how it helps someone deal with the fact that they have diabetes, years after years. How it helps a person get out of an abusive relationship.

>> No.10746115

>>10746108

Africa will grow by more than one BILLION people in less than 25 years. AutoDocs are needed

>> No.10746132

>>10746109
>I can't see how an AI helps someone get through alcoholism, how it helps someone deal with the fact that they have diabetes, years after years. How it helps a person get out of an abusive relationship.

What you describe is the job for a health oriented SOCIAL worker.

I mean the doctor that reattaches your leg after an accident, or stitches a glass cut, or removes an infected appendix.

>> No.10746133

>>10746109
none of this is important in the discussion. you are framing it as an on/off perspective. absolutes are useless.
it's not that there will be zero medical professions, but instead we will be seeing a reduction in medical professionals in the industry the more hospitals adopt AI diagnostic into their system.
There will be people to monitor special cases.
Overall this transition will reduce costs, and divert those who will not be in the medical profession towards more productive professions for their skills.
Unless, of course, the union seeks to prevent this shift from occurring - which they will.

>> No.10746142

>>10746133
>the union seeks to prevent this shift from occurring - which they will.

At first, household humanoid robots will be strictly banned from performing ANY medical procedure by the huge medical industry.

>> No.10746159
File: 11 KB, 262x192, download.jpg [View same] [iqdb] [saucenao] [google]
10746159

>>10746109
>lately every time I check /cashiers/ I see this thread. I know many people wish they had an easier access to canned goods, vegetables, etc... but I can tell you it's not going to happen in the next 60 years, if ever.
>Other people have explained some of the reasons well, as for my experience, I can tell you that my work as a cashier is only about 60-70% from what I've learned in high school. The vast majority of what I've learned is completely useless. You guys assume that a machine with a huge database is more likely to make a correct transaction but the additional knowledge only affects a fraction of a fraction, and in order to apply that knowledge it would need an exponential amount of information on the patient. Ultimately, it would lose the most important aspect which is the cashier - customer relationship.
>I realize that many people are frustrated with their relationship to their cashier, I'd advise them to try to find another one that better suit them, one thing for sure is that there is a lot of uncertainty in that profession even if an AI takes on the job, what the customer expects is that the cashier takes his concerns seriously and genuinely try to adress these concerns in the best way. All a machine can do is scan a barcode. Most people I meet need time and different ways to approach their transaction, some people want an excess of interaction, others on the contrary tend to flee from the appropriate interaction. With time, if you gain their trust, you can help them overcome their fear in a way that improves their customer service experience overall. And sometimes despite your best effort they persist in their bad habits but that's also their right to do so.
>I just can't see a machine handle people. I can't see how an AI helps someone get through their purchase, how it helps someone deal with the fact that they need groceries, years after years. How it helps a person get out of the store quickly.

>> No.10746174

>>10746132
even surgeons don't do a 100% of procedures. You missed the point of my post anyway, it was about the human component in the relationship. People are very different and asks for a personalized approach to their problem. Just because you see the same dermatologist with the same condition doesn't necessarily every person wants the same care/treatment.

>>10746133
you're absolutely correct about unions, I didn't mention it but it's one of the other reasons that I'm not even remotely afraid. For the records, I'm not even that attached to the profession, I'm lucky to do that job but if I'm laid off tomorrow I'm at peace with doing something else, I'm not sure I'd try to go back to it.
Doctors might use more algorithms in the future, it's not like they don't already exists and they're helpful for trickier conditions like pulmonary embolism. It could also help lower medical burden to handle mostly conditions like flu especially during the peak epidemic, though even for that I'm pretty sure a lot people will reject it if they feel like their concerns are not adressed properly, especially in terms of amount of days of rest given. After all, it will come down to that, if you give that power to the AI, then the government will decide how much time you get when you're sick, and that time will most likely be much less than what a doctor might give you.

>> No.10746184

>>10746109
This, people will never rely on some kind of internet search algorithm, or something like "wikipedia." What people are used to is a travelling encyclopedia salesman knocking on their door by chance, buying an encyclopedia that will never be updated, and looking things up when needed. Just because some internet thing can have a vast amount of instant, up to date information, means nothing, people need the human touch.

>> No.10746191

>>10746174
>even surgeons don't do a 100% of procedures.

but AutoDocs DO 100% of the procedures and they NEVER get tired, or slow, or discouraged.

Africa in less than 25 years will have more new people in it than ALL of European population combined.

There are not enough doctors. AutoDocs solve this problem.

>> No.10746197

>>10746184
why do you guys think diagnostic is the only part of a doctor's job. It's already there, the internet, you can look up your symptoms and find out what you have for the most frequent conditions. Why is there still obesity in the world, just look up a diet and exercise regimen online. And yet it's considered one of the biggest health issue in the world right now. What's your AI going to do about that ? it's the future of medicine, should be able to handle something like that

>> No.10746200
File: 1009 KB, 640x360, fatty_wash.webm [View same] [iqdb] [saucenao] [google]
10746200

>>10746197
>Why is there still obesity in the world,

Obesity is an issue for a health oriented social worker, it is a HUGE waste of resources to have a medical doctor discussing diet to a fatty.

>> No.10746208

>>10746197
>Why is there still obesity in the world, just look up a diet and exercise regimen online. And yet it's considered one of the biggest health issue in the world right now.
Guess human doctors have proven ineffective. What kind of retarded argument are you trying to make here?
>AI can't do something about obesity! Neither can human doctors, who were the standard option for healthcare during the obesity crisis, but uhhh AI will never work!

>> No.10746229

>>10746200
>it is a HUGE waste of resources to have a medical doctor discussing diet to a fatty.
well I can't speak about your country but in mine, the price for a visit at your family doctor or a dietician is about the same (25 euros for the doctor, 20 to 60 euros depending on the amount of time for the dietician).

>>10746208
calm down, we're having a conversation, I say what I think you say what you think, some of it goes over your head, some of it goes over mine. No need to be insulting or patronizing because I won't respond to it.

>> No.10746242

>>10746229
>price for a visit at your family doctor or a dietician

price is not the issue. An MD giving diet advise to a fatty is a waste of time, that is a dieticians job.

Like having an IT department sending over a top developer to show you how to change the font size in your browser. Yeah he can do it, but it is a huge waste of his time, a lesser trained individual should be solving the problem.

>> No.10746293

>>10746242
but that's exactly what a family doctor's job is, to have basic knowledge in every field so that they can either treat or refer patient where they will get the help they need. What you propose instead is that a patient should go to the dietician for his weight problem, go to the physical health social worker for his back problems, to the psychologist for his difficulties to adapt to his new job, to the gynecologic health social worker when her period is late or something.
Every single time, the patient will have to test a new relationship which in my experience people hate to do. It's a daily occurence that people don't go to the dietician or psychologist, many times they go once and say it didn't help them at all or they didn't like the person, very few of them actually go through the therapy. Also, these people would have a very superficial understanding of medicine so they wouldn't always see what the person might actually need if it goes outside of their field of expertise.
Which drives us back to the AI, which in order to work you'd have to fill out a pretty extensive medical form. Living in a poor district in my country, 20-30% of the people I see can barely answer the simple questions I ask them, I also have syrian refugees who can't even write, even in arabic.

Well that's it for me, nice talking to you.

>> No.10746666

>>10745656
Surgery is not only what goes on in the operating theatre. There always needs to be a man on top coordinating the team. Did the robot also prepare the patient?

>> No.10746694

>>10746666
>Surgery is not only what goes on in the operating theatre.
Is there any part of what goes on in the operating theater that transcends physical reality? Also, if there's a person doing something other than surgery in a future autonomous robotic surgery scenario that doesn't necessarily mean this person is a surgeon.
>There always needs to be a man on top coordinating the team.
It wasn't just the physical execution the autonomous robot provided. The plan itself for how to go about performing the procedure was generated by its program.
>We demonstrate in vivo supervised autonomous soft tissue surgery in an open surgical setting, enabled by a plenoptic three-dimensional and near-infrared fluorescent (NIRF) imaging system and an autonomous suturing algorithm. Inspired by the best human surgical practices, a computer program generates a plan to complete complex surgical tasks on deformable soft tissue, such as suturing and intestinal anastomosis.

>> No.10746721

there are so many doctors that are fucking retarded and their actions result in people dying. oh sorry I had a few too many drinks last night, your grandma is dead

welcome the robot overlords

>> No.10746733

>>10746694
Listen nigger. I'm saying that the surgeon is the boss for surgery and the boss is never replaced by robots.
Look at any major company, the work the CEO does is absolutely useless but he always keeps his job. Why? Because he decides. The same applies to medicine in general, doctors will never lose their jobs because they will always be the ones to use/oversee robots in case of mistakes, because we will never let machinery operate without supervision. Especially in life threatening situations.

The exception is radiologists those guys aren't gonna be around in 30 years

>> No.10746738

>concern
Only if you're a Doctor. A Doctor is expected to sit there for 8+ years studying in general or a specific niche, then has to have years working in real environments, then has to keep up with technology and advancements in a time we can tease out tiny effects. A human being cannot know this information accurately. Then multiply this by every single patient being different.

It's remarkable we even let Doctor's be humans. There's no reason whatsoever test results shouldn't be cross-correlated and inferences applied in every situation, to determine the likelihood of various disorders and optimal treatments based on "approved" results from studies and real-world applications.

Doctor's should be like engineers, you take a theory, find an application, and use references and technology to properly apply it. There's literally no reason Doctor's exist except to create a separate social caste.

>> No.10746755

>>10746738
It's not possible, we don't know enough about medicine in the first place.
Here's the post that summarises it the best yet was conveniently ignored by everyone in the thread >>10744541

>> No.10746893

A question to the people who think AI will completely replace doctors: how would an AI be able to know when somebody is a hypochondriac?

>> No.10746905

>>10746109
You don't need 8 years of medical school to be a social worker and patient support. All you need is four years of college and charisma. If autodocs or automated diagnosis covers more facets of healthcare (they are already used for some), the role of doctor could easily be replaced by other medical professionals, and with equal or better quality healthcare.

>> No.10746910

>>10746893
This is trivially easy for an AI with access to a database of millions of medical records including the patients, and difficult for a doctor. An AI could look at their past complaints, compare them to a larger population in a NN, and know that it's not realistic for someone to have 25 unrelated complaints. A doctor seeing a patient for the first time will look at their record manually, interview them, the person doesn't like the answer and goes to another GP, process repeats, which is the norm for hypochondriacs now and how they continue to exist.

>> No.10746921

>>10746893
>how would an AI be able to know when somebody is a hypochondriac?
Same way as how AI learned to interpret MRIs. It could run through historical case data sets with known outcomes and learn to make decisions that minimize its error function (difference between its prediction outputs and the known case outcomes) in attempting to label serious bodily medical problems vs. psychological problems.
More generally, how would a human be able to? If there's an answer for how a human would be able to then there's an answer for how an AI would be able to.

>> No.10746929

>>10744541
Us biomedical researchers are doing the best to create convenient, fast, accurate, and less invasive diagnostic tests of all times. Part of that is reducing the amount of work doctors have to do. It's my sincere dream that we can treat people with less doctors, since doctors are highly educated and expensive, and still prone to be unreliable compared to a tested, highly characterized medical product that passed FDA trials.

>> No.10746947
File: 4 KB, 394x338, v8n23a07-f01.gif [View same] [iqdb] [saucenao] [google]
10746947

>>10744574
How do you even compare those two? The progress in medicine in the last 50 years is incredible, and continues at an ever faster rate as well. We've cured Hep C, nearly cured AIDS, begun work on curing genetic diseases, have far cheaper genetic testing for medical purposes, and new oncology drugs come out every year.

>> No.10746959

>>10746947
That cure for Hep C though, could you really imagine a better breakthrough? Something like 92% success rate average, 8-12 weeks of medication, and few side effects. There's some complaints because it costs like $100k, but really, many other (albeit unpreventable) diseases cost way more than that. I know someone who had a stroke and got charged $100k-200k to end up nearly a vegetable.

>> No.10747006

>>10746026
it's interesting, blue only gets up when the two greens glue together

>> No.10747017

>>10746910
>look at their past complaints
Don't know what country you live in but here you need to consent for that information to be accessible by anybody, which most people dont.
It is much easier to lie to a computer than to lie to a person, even lie detectors aren't even any good and people need to be strapped into those for them to work.

>>10746921
>how would a human be able to?
A tutor of mine told me when patients showed up with a list of textbook symptoms and test results were negative he would have a talk to the person. This is the part that a robot can't do, since there needs to be an interpersonal relationship to help the hypochondriac understand they are not in need of medical attention.
>If there's an answer for how a human would be able to then there's an answer for how an AI would be able to
Does an AI understand emotions?

>> No.10747024

>>10742593
This is so fucking true. Can't let all that shit we went through go to waste.

>> No.10747036

>>10747024
I don't understand, you aren't happy about getting to prepare my hamburgers in a few years? t. AI expert

>> No.10747038

>>10742417
Too fucking kek
>legally protected
Until it isnt.

>> No.10747064

>>10747036
Ah don't worry. You don't have to understand it because it's not gonna happen anyway~

>> No.10747072

i can already program a machine to randomly prescribe either opiates, antibiotics, or exercise.

>> No.10747094

>>10747072
If patient = fat Then
Prescription = exercise
Else
If patient = sick Then
Prescription = antibiotics
Else
Tell patient to fuck off

Wew that was hard

>> No.10747106

>>10747094
That's not random. There goes your interview.

>> No.10747116

>>10747106
Shit

>> No.10747159

>>10746063
It would be shut down for targeting the impoverished

>> No.10747200

>>10747159
>implying government cares about poor people
We're talking about cities in the only country that matters.

>> No.10747933
File: 247 KB, 705x527, 1523490452247.jpg [View same] [iqdb] [saucenao] [google]
10747933

>>10742395
>mfw Australian internet is too shit for a robot takeover

t-thanks Malcolm

>> No.10747975
File: 546 KB, 1031x1880, circumcision blood money.jpg [View same] [iqdb] [saucenao] [google]
10747975

>>10742395
The sooner doctors are put out of business, the better. No human doctors means no more kikes chopping off baby foreskins.

>> No.10747994

>>10742395
I am doing my master's thesis in machine learning and certain tasks such as spotting cancer from x-rays are such things that I sure hope they replace human error ASAP. Those systems are already better than humans at it. Anything involving repeating pattern recognition is done better by machines now. As for human interaction, there have been trials where bot screens self-given symptom descriptions matching them to age, race, gender and personal disease history to see if there is reason for alarm, also doing it better than human doctor over phone and most importantly instantly instead of calling back in a few days.

>> No.10748000

I hope so then i have an excuse to stop
Fuck you mum

>> No.10748053

>>10745170
>implying a doctor would know how to service and repair a complex medical examining machine
If anything it would be maintained by a team of codemonkeys and engineering technicians.

>> No.10748157

I have suffered a lot and have irreparable damage (which is negated by drugs i have to take every day, I'm completely normal otherwise) from false diagnoses and shitty surgeons.

I am now getting into my MSc i. data science specialized in machine learning and have made it my life's goal to help accelerate the removal of stupid doctors and the institution of automatic systems that help whatever medical doctors are left. Not even larping. Pharmaceuticals decide your curriculum and we know where their interests lie.

>> No.10748186
File: 340 KB, 702x598, emotions.png [View same] [iqdb] [saucenao] [google]
10748186

>>10747017
>Does an AI understand emotions?
Yes. Emotions aren't magic. They're observable behaviors and observable physiological reactions.
>It is much easier to lie to a computer than to lie to a person
Lie detected.
https://www.indiatimes.com/technology/news/this-ai-can-detect-online-liars-better-than-humans-and-could-be-your-perfect-tinder-filter-364091.html
>As it turns out, a human can spot lies in online messages about 50 percent of the time. Ho's machine learning system however, can do it between 85 to 100 percent of the time.
https://www.cnbc.com/2018/05/15/lie-detectors-with-artificial-intelligence-are-future-of-border-security.html
>A virtual border agent kiosk was developed to interview travelers at airports and border crossings and it can detect deception to flag human security agents. The U.S., Canada and European Union have tested the technology, and one researcher says it has a deception detection success rate of up to 80 percent — better than human agents. The technology relies on sensors and biometrics, and its lie-detection capabilities are based on eye movements or changes in voice, posture and facial gestures.
>There needs to be an interpersonal relationship to help the hypochondriac understand they are not in need of medical attention.
No there fucking doesn't.

>> No.10748442

>>10743656
>diagnose problems and order the same tests
no, he can't, he's not as good. the point that anon was making is that the AI is superior. There's no tradeoff, it's just better

>> No.10748449

>>10743969
same, we'll have a party when it happens

>> No.10748452

Why does people here hate Doctors so much?

>> No.10748457

>>10748452
They're robots who regurgitate everything fed to them by their pharma corporate overseers on unsuspecting patients. They're highly overrated in their use to society and people go into it pretty much only because they want the money.

>> No.10748551

>>10748452
in America, they are often are arrogant pricks who don't listen to their patients, spend as little time as possible with you, were autists or Machiavellians or both in school in order to get ahead, and get things wrong. I have met a few doctors who fit the old Normal Rockwell stereotype and aren't like the above description at all, as a matter of fact I'd say these ones were definitely men of exceptional character and IMO well deserved the wealth, prestige, and respect that are afforded to doctors, and more besides that. But the good ones are less than 10%, and the assholes run free.

>> No.10748555

>>10748157
>made it my life's goal to help accelerate the removal of stupid doctors and the institution of automatic systems that help whatever medical doctors are left. Not even larping
unironically, based. This is the kind of thing a great story is made of. Hope you get your dream anon

>> No.10748559

>>10748000
if you don't stand up to your parents now things'll get worse and worse. You won't be happy once you finish med school.

>> No.10748575

>>10748157
Good luck, anon. It'll be a tough fight given whose profits are on the line, but it's an important one.

>> No.10749286
File: 12 KB, 236x228, e2581bd67d93d899abc73ca5cb528c83--fruits-and-vegetables-veggies.jpg [View same] [iqdb] [saucenao] [google]
10749286

>>10748452
There are many reasons for people to hate doctors, both intelligent and stupid:

>Hate doctors because you're a conspiracytard
>Hate doctors because you think they're arrogant and stupid
>Hate doctors because you realize they are one of the greatest expenses in healthcare, with some of the lowest productivity/impact on patient wellbeing, compared to pharmaceuticals, preventative care, and medical devices themselves
>Hate doctors because they purposely form cartels like the board of medicine, which creates an artificial scarcity of doctors to drive up their wages, resulting in a net loss of health and wellbeing for all Americans, while also being worshipped by normies for being so "smart and kind"
>Hate doctors because they are paid much more than the people inventing the devices, therapies, and medicines that are actually saving people (biologists, biomedical engineers, and chemists)
>Hate doctors because you have negative associations with them or negative experiences on a personal level.

>> No.10749320

>>10748452
>Why does people here hate Doctors so much?
They're mostly smug assholes who forced themselves through a few extra years of school so they can have official authority and financial advantage over other people. It'd be neat to see them get reduced back to normal person status because of biomedical research anons replacing them with impersonal programs that just do the work without having any feelings or ego driving their results.

>> No.10749347

>>10742395
Doctors will.be just like pilots

Only there to reassure the client. And to correct grevious errors. Though with complacency they will panic under the pressure to actually perform and eventually more power will go to the robots

>> No.10749392

>>10744306
>implying mathematicians, physicists and statiscians couldn''t do the workm cs-niggers do a million times better

>> No.10749484
File: 120 KB, 710x516, Untitled.jpg [View same] [iqdb] [saucenao] [google]
10749484

>>10749392
Not him, but programming really is its own skill set. I've seen subject matter experts who were brilliant (*in their own field*) try to program and the results are always horrendously bloated trash built with some domain specific pseudo-language that could be accomplished with a tiny fraction of the effort they expended and with much, much greater efficiency by almost anyone having actual dev experience under their belt.
Not to mention almost none of them do proper data management, meaning instead of a relational database they dump all their work into Microsoft fucking Excel and end up causing disasters like this:
https://www.popularmechanics.com/science/a22577/genetics-papers-excel-errors/
Where thousands of researchers were found to be saving gene names in Excel in a way that auto-converted them to dates and floating point numbers e.g. SEPT2 converted to September 2nd. And when Excel does this is doesn't just change how the data appears. It changes the actual value e.g. MAR1 in Excel would get auto-formatted to 1-MAR, and if you try to undo it only the formatting part would be undone to reveal the value is now 42430 (serial number Excel uses to represent first of March).
And this happened with ONE IN FIVE scientific papers on genes out there, so I'm not just talking about some one-off researcher who doesn't know his stuff.

>> No.10749510

>>10749484
Can confirm this, worked in an ecology lab doing data management. The head professor was some boomer chick who had a major distrust of anything digital, her access database was trash tier in it's operation. It took one worker 2+ hours to upload simple GPS data, you had to upload it to excel first, do a bunch of stupid operations, then append it to access. I ended up writing an R script and some sql which condensed it down to 5 minutes for her, but the bitch kept insisting her method worked. All I know is, her database is a disaster waiting to happen.

>> No.10750160

>>10748186
Sorry you had shit doctors anon

>> No.10750344

>>10748452
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5885453/
Following a checklist has been shown to reduce surgical complications by a third. no more sponges left by people, operating on the wrong limb, forgetting to get everything out that you're meant to. Surgeons don't like using them because basically they feel the lead nurse is in charge, or the checklist is in charge, and goddamn it they went through all that stuff to be God, not follow a damn list. Imagine people trusting you with their lives and you being unable to overcome your ego enough to give them every chance for health. That's the mindset of, I'd say, the majority of doctors. Not all their fault, the US medical training and maybe other countries' as well, promotes tunnel vision and narcissism in order to survive the slog

>> No.10750355

>>10749510
I almost worked with a biology lab that was making "revolutionary" image processing software. One of their demos was "turning a 2d picture into a 3d model", which was done by making the Z axis brightness. They could not be convinced that it wasn't amazing, because it worked "really well" on their test images of white people with dark backgrounds. Boomers should not be allowed to use computers.

>> No.10750766

>>10742471
why won't this ai do the research aswell?

>> No.10750783

>>10750344
>that 5-year period where research showed doctors washing their hands more frequently would prevent some 60% of hospital-acquired illness, but the doctors refused

why tho, you already have driving gloves for your fleet of mercedes

>> No.10750850

>>10749484

Are there any online or part-time courses/bundles/books/etc. that an aspiring physicist/biologist/inventor/etc. would be wise to go through, in order to manage data and programming better?

>> No.10750892

>>10743173
So they are exactly like human docs in usa

>> No.10751015

>>10750766
Original research is a creative and somewhat random process, often requiring fine motor skills and many many original movements. Most of it would not be feasible for an ai, unless it was near human intelligence. (Which we don't even know if that is possible)

>> No.10751243

>>10742419
Legally he means. A machine gives the result, the doctor then confirms/denies it. You dont walk up the MRI, let it scan you and then it prints 'congrats you have cancer, heres the list of the procedures that i advise you to do'