Which medical specialties do you think are likely to be replaced by AI's in the next 50 years?

[circlecircledotdot]

Speaking in practical terms, a robot will never be an independent provider. Can you imagine suing one for medical error? Someone has to and always will be responsible for the final call (i.e. a supervising physician). 

[NeuroD]

3 hours ago, circlecircledotdot said:

Speaking in practical terms, a robot will never be an independent provider. Can you imagine suing one for medical error? Someone has to and always will be responsible for the final call (i.e. a supervising physician). 

People often cite this as a reason. Why does a medical provider have to be sue-able? If, 100 years from now, automated providers are 10x more efficient/effective/safer, will we hold back progress just so we have someone or blame/sue? What about when it's 100x better? What's the threshold?

[circlecircledotdot]

.

[NeuroD]

22 hours ago, circlecircledotdot said:

I think you misunderstood what I said. Even if a physician is stripped of all their day to day chores and have it replaced by robots that are 100x more efficient, someone will still need to be held accountable; be it a physician, an engineer, etc. 

If robots start to mimic the function of a physician, or any other job for that matter, they will also mimic their legal liability. But robots are property and they don't have any legal status. So if a robot causes harm like a surgical error, you have to sue the one who has made the final decision to approve of its use during the case.

And correct me if im wrong, but i'm pretty sure the final call will be up to some doctor. 

Long story short, you can probably eliminate most of what a physician does - sure, but that won't eliminate the job - just change the nature of their work. In the context of AI, most of their work will probably be in managing the new technology that is afforded to them. 

Like many have said though, I don't think our generation has to really worry about this during our working lifetime. Pagers are still being used, paper charts are still being used; the latter just starting to get phased out atleast in Ontario

I don't think I misunderstood. I'm asking (not you directly) what the threshold is for how much better robots/AI will have to be, before we as a society are willing to accept not being able to blame "anyone" when it makes a mistake. And if we don't make that leep, will we be holding ourselves back from our progress, just to make sure there's a person behind each decision that can be held liable?

I think we underestimate how soon we will have to deal with these issues. We tend to think linearly, whereas this stuff moves exponentially. That's just not how our brains are set up. I think we'll be dealing with some version of this sooner than people realize. Yes, we still use pagers and paper charts, but some things just take over because the cost vs benefit is too good delay.

[1D7]

Technology doesn't follow a straight line. Some things advance quickly then slow down or stop once we've reached some sort of limit. E.g. while computers and phones have revolutionized the way society works, cars and motorized vehicles have remained largely the same for decades. The most recent AI implementation (IBM Watson) has been a failure. There are still many barriers, including technological ones, that will have to be overcome for AI to cause any displacement at all.

But to answer OP's question, I'll go against the grain and say the most at risk is actually family med/primary care due to continued encroachment by NPs.

Their lobbying associations already push articles about providing equivalent care to family practitioners, and in America they have begun to practice autonomously without physician supervision in some states. It's not too far a stretch to say that in the distant future that family medicine will be replaced by a combination of NPs with AI at their disposal.

 

[la marzocco]

Tbh uptake of AI in Canada will be slower than the US (esp. in a public healthcare system environment). Things we should get done first imo: (1) axe the fax and have interoperability between EMRs across hospitals and practitioners (and preferably across provinces) and (2) empower patients to have better access to information regarding their health. Lots to do before we even consider AI. Baby steps.

Also.. when I read about Phoenix pay system.. and how they keep throwing money to "stabilize" it and now considering to scrap it altogether for something new.. yea.. AI can't come fast enough. I digress  

[rmorelan]

one point I should make here particular with respect to the liability - there is still a lot more wiggle room there than people often give credit - even a robot doesn't have to work perfectly - you have to make sure though that the patient is completely aware of the risks and benefits. To put it another way - just because now you have a bad surgical outcome or infection doesn't mean you get to sue (or at least sue and win). You know up front that infection or a bad outcome are possibilities and agree to continue, and as long as reasonable precautious and care is taken then things are as bad as people make it out to be. It isn't quite as black and white as people make it out to be.