What Will Medicine Be Like In 2123?

A doctor makes 10 predictions about the state of medicine in 100 years by extrapolating from recent trends. This is mostly for fun and not meant to be taken too seriously!

@0:00 Introduction
@1:31 Artificial Intelligence
@5:33 Diabetes
@7:02 Cancer
@7:56 Partitioning of medical specialties
@9:58 Longevity
@11:58 Nutrition
@13:44 Wearables
@15:54 Emerging infectious diseases
@17:10 Medical education
@20:34 Universal health care (in the US)

[StrongMed]

Since I don’t want to upset anyone, here is a little more on the specific prediction that radiologists and pathologists will be the first specialties to be *partially* replaced by AI by the mid-21st century. As I mentioned, radiologists and pathologists won’t be the only specialties to be affected by AI, but rather they will just be the first – so I’m not in denial that this will impact internal medicine too! There has been a lot written about whether or not this phenomenon will happen, for example:




It is my impression that some of the disagreement about this possibility arises from people considering different time scales. Folks who think AI will not replace many radiologists and pathologists seem to be thinking 20-25 years in the future (evidenced by straw man arguments against anyone who advocates for stopping to train residents in these specialties now – which virtually no one is actually advocating for). Whereas those who believe they will replace many (or most) radiologists and pathologists are looking farther ahead (i.e. more than a single physician’s career in the future).

One of the earlier steps will be AI suggesting diagnoses that radiologists and pathologists will need to individually review. In other words, AI will initially help improve speed and accuracy of these fields while the numbers of physicians in them remains the same. Over time however, as AI diagnoses become incrementally more accurate and detailed, and as the time needed to review each study shortens, employers will realize that there isn’t a need for as many humans. Less time per study --> less need for physicians. For common diagnoses, it’s possible that AI will even surpass human accuracy (and it will definitely far surpass human speed). There will still be a need for radiologists and pathologists to review the more complex cases (at least for a time), and a need for them to tweak AI algorithms, and to study newly described diseases. But it is inevitable that eventually – and possibly within my lifetime – AI will be making routine “bread and butter” radiographic and pathological diagnoses with minimal to no human intervention

[chubbyemu]

Something that always interests me is seeing old ideas from hundreds of years ago come back. In re: cancer immunotherapy, there were Coley’s Toxins back in the 1800s that fell out of favor but the idea of the immune system attacking a cancer came back in the 1980s with IL2 and interfereon, and then in the 90s with the animal experiments on CTLA4 and PD1 which were approved 20 years later giving way to what we have now. The tech that benefits medicine the most, in my opinion, is the tech from a new fundamental understanding of nature like genetic engineering making insulin, X-ray, radiation therapy, MRI, etc. Benefits from tech of computation, and I would lump AI into this category, are more nebulous and sometimes counterproductive. Medicine and this tech can be incompatible because we are dealing with a matter of life or death, not a financial transaction or sending an instant message. In context of old ideas coming back, it would be funny if AI somehow makes a connection that brings back a modern day version of the 4 humours.

[hemmojito]

It seems unbelievable for the current generations of nations outside the US, let alone future generations, that a person with heart attack may face bankruptcy soon. Speaking as someone living in Germany.

[BenjaminKuruga]

Hey, 28 year old Physical Therapist here! I still come back to review this video, and i share it with many friends of mine.

[Bunicutaintelectuala]

Dr. Eric Strong, thank you for all your work and efforts. You are the mentor we've never had in real life.

[UnfilteredMedic]

A.I. will never be able to duplicate human intuition, instinct, and that "gut feeling".

[farwakhan]

My little hope is that bidets get world-wide used.

[DeepakKumar-tlpk]

Request a lecture on altered sensorium and it’s management

[peybak]

Fun list of things to imagine. But I bet AMA would have something to say about PA's doing any of those procedures.

[DrHameedMD]

Very interesting points. One prediction that I found a little far-fetched is the one about fast foods becoming " rare indulgence" instead of a daily life thing. If we were to extrapolate the future eating behaviour from the current one, I believe the junk food culture will only become more commonplace.
I wonder what will the case be with the psychiatric disorders.

[tomharrysson]

Never been this early, always a great day when you upload!!

[ethannorman]

Very soon surgery will have a remarkable transformation. complex procedures like artificial heart transplants, where just one nurse and one doctor oversee the operation with advanced robotic systems, will be handling the majority of the surgery, reducing the need for human intervention.

Also the pursuit of a cure for aging is gaining momentum. Genetic engineering and sophisticated computer modeling will undoubtedly extend human lifespans significantly by slowing or completly stopping biological aging.

These groundbreaking developments are completely possible and the potential for precision surgeries and the promise of longer, healthier lives. Brace yourself, because the future of healthcare is nothing short of unbelievable but still fully achievable within this century.

[an]

I think there is another caveat - these are predictions for medicine in developed countries. Some countries still dont have (wide) access to technology used in other everyday (for example robotic surgery) so there the changes will happen later and slower if ever.

[blackwhite]

Will, in your imagination, editing out susceptible genes - HLA genes for example - and replacing them with resistance conferring variants to make human embryos that have a higher chance at survival be a normal medical practice?

[christiansoto]

human intuition is a gamble, but I like its odds

[judypeng]

Do you know where have good basic radiology class for FNP student?

[twistedtea]

I predict in 100 years the VA will still use CPRS

[ammsenosy]

What about emergency and critical care medicine?
In these areas, there are a lot of machines.. Could they be able to do the work themselves in that far future?

[farwakhan]

Love Love Love these predictions. So astute. Especially the last one. I'm optimistically hopeful that we'll have Universal Healthcare or atleast some form of it.

[slumpig]

Here in Australia medical schools are transitioning to the American model of post-graduate medicine.