Use all the free resources available to you from SDN: articles, guides, expert advising, forums discussions, and school research.
Good question. Looks like it’s been a research topic since at least 2023, but nothing at scale has come out yet to my knowledge. My guess is that it is either marginally better than physicians and hospitals would rather just have a actual human writing their signature their (probably for liability purposes) or it is marginally better but much more expensive.
Now what if the automated EKG was recognized as an independent provider, with their own NPI?
Who carries the liability when something goes wrong? Until this is answered I think this will continue to be my biggest question regarding any kind of AI administered healthcare.
Yes and no. It completely depends on how you're defining "deviation from the community standard", and any lawyer worth anything could easily punch a hole through your argument if an AI misses something major that leads to harm. Misreading an EKG would be a "deviation from the community standard" if it is something that should be caught. If your argument is that AI will be so perfect that a misread is impossible then I'd argue AI should be reading every EKG, but that's not what I'm talking about.
The preferred argument is that AI should be federally recognized as an independent provider, but that there is no liability because the AI standard of care is something else.
This is exactly my sense. Psychiatry is far, far from being replaced compared to other fields. Radiology has such "hard data" and almost no soft data - with predominantly zero human touch. This is a field ripe for a true 1 for 1 replacement in the long run. Rads can pivot to IR more and keep their jobs. Pathology is going to keep working because AI isn't slicing brains or actually doing the biopsy, staining, running labs and prepping slides. Most other fields with human touch will have a harder time being replaced since AI isn't performing physical exams for primary data. Some fields, like surgery, also have such a high barrier for AI to get primary data that they will likely be replaced last. By the time surgery is replaced - no one will have a job on the planet except politicians, artists, philosophers, and priests.
Agreed. Maybe individual practitioners will get hit with lawsuits for over-relying on AI, but AI as a whole is not going to get derailed because of malpractice fear.
This does not match the data I have seen at all. They ran OpenAI, Claude, and Gemini (and I believe a 4th) through adult and child psych board exams and the best performances were just barely passing adult boards and below passing for child boards (where they do significantly worse). It's a far cry from "blowing physicians out of the water". I am sure the scores will improve, but the limits of having good data to enter and refine the models is a bigger hurdle than $37265861723 data center. This is supported by all the models making about 50% more errors on child psych than adult psych with less information to train on.
Idk that we are, at least not from the liability standpoint. AI as a tool, the one using the tool is liable. AI as a provider...the AI is liable? The hospital or office the AI is based out of is liable? The company that created or maintains the AI is liable? If AI is practicing medicine, then there's already a standard of care. I would argue that this doesn't just disappear because it's AI vs a real human, but who knows?
Maybe? Independent practice of NPs has slowed significantly where I'm at, and I'm in a state with FPA where I've noticed the slowing has taken place AFTER gaining FPA. Many nurses have realized being an independent doctor isn't the land of milk and honey they thought it would be. Admittedly, it's not the legal aspect that slowed things down, it was the practical one here, so you may be right there.
So are self-help books. And dogs. And a rock with a face painted on it. The bar for accomplishing this is pathetically low. I could descend into hell and I'd still have to go down another few levels to find it. Considering many therapists don't actually practice therapy, this isn't really something I'm concerned about.
How long ago was it that we were promised self-driving cars? I remember reading about how the trucking industry and taxi/Uber industry would be wiped away. But that hasn't happened yet.
Are you just referring to Waymo? If so they're still pilot projects in only 5 cities and there are still some very major barriers (like inclement weather cities, which they are in zero). They've also already had to recall their fleets in several cities last year (Austin and Atlanta at least) because the software wasn't recognizing school buses and were passing them while the stop signs were out and kids were exiting. As of this month, this still hasn't been fixed. Georgia politicians are already looking at filing more restrictive legislation and and liability law proactively due to incidents in Atlanta. If they're analogous, we're going to see similar problems with medical expansion of AI, which I would expect anyway.
I am 1000% confident that we will not be replaced in the next 10 years and imo thinking otherwise is borderline delusional, but I think this also depends on how you're using the word "replace". Do I think it's possible that some doctors (likely PCPs) will get forced out of their current employment due to AI making things more efficient or even treating basic patients? Sure. Do I think doctors are going to be replaced en masse by AI in the next 5-10 years, absolutely not.
I would argue that you are downplaying the negative effects midlevels have had. Salaries might be up over the past 10 years, but from what I have seen they have very much plateaued or even gone down slightly over the past 2-3 years. Degree mills are pumping out more and more NPs each year. I know of at least 3 or 4 psych NPs around here who cannot find jobs as providers and are working RN jobs. There are a number of provider positions being filled with midlevels now that would have been filled by a psychiatrist 10+ years ago. Not only does this impact salaries, but it impacts how we are treated in the work environment, as employers know they can replace us fairly easily, even if it is with lower quality.
That's for neurology. Psychiatry was 80% for adults, 70% for CAP (roughly), this was just presented at AACAP 3 months ago.
I have heard others on here also refer to what we are discussing as "fear mongering". Why do you use this term? AI taking over is not something to fear or be upset about. Quite to the contrary, it is very exciting and could potentially lead to a much better quality of life. Perhaps "euphoria mongering"?
Maybe they're big in SF/LA, but my understanding is that in the other cities (Austin, Atlanta, Miami), it's still a pilot phase that's hitting some major hurdles (though the way people in Miami drive anything to replace them is probably an improvement). I'm also not saying that AI is worse than people driving. I've actually talked to friends about how even the early Tesla self-driving cars were much safer. The problems are optics and social inertia. I remember when it looked like Tesla was going to get the next stages approved but a single major highway accident completely derailed the project and set it back years because of the media response. One kid runs out from in front of the stopped school bus and there goes Waymo. It's how Ryan Haight got implemented, it's how residency hour limits got implemented, it's how Tesla's self-driving program got set back. Spook the wrong person and the whole movement will come to a halt.They are in no sense a pilot project at this point in CA, although I take your point about inclement weather. They are expanding extremely rapidly:
It definitely isn't a perfect technology, but the statistics are pretty clear. They are dramatically safer and less likely to be involved in accidents, fatal or otherwise, than the average human driver. If we were deciding things purely as a safety issue, we'd be trying to figure out how to make all cars self-driving as fast as possible.
Edit:
For those who have access to the NYTimes, a neurosurgeon makes the same point.
Opinion | The Data on Self-Driving Cars Is Clear. We Have to Change Course.
If we could make this magnitude of improvements across the entire American automotive fleet we would basically eliminate motor vehicle accidents as a major cause of death in the US.
So I'll quote my questions to you from another thread because you seem to have very different opinions on what a future run by AI looks like:
The US is growing faster than we can train PCPs despite and explosion in FM residencies and a decent increase in IM ones. Even if AI makes me more efficient, it would have to basically triple my productivity before jobs start going away.
I enjoyed that same premise in Herbert Spencer's writings about the industrial revolution. We're all drowning in leisure time since the industrial revolution happened.
This is a great point that I had not thought much about before. Basically if the easier work is handled for you and you are "freed up" to do only the most difficult work, the day would feel a lot less balanced and burnout would be more likely. Imagine a day of only borderline PD patients in crisis, clozapine patients who are decompensating, etc.
Why wouldn’t AI be able to handle these patients?
Agree. 5 years to build that really nice nest egg and hit true escape velocity, if not already financially free.
Which is probably the majority of active people on these forums IMO. Not to mention most of their real social networks etc. No one has seen a real downturn yet. Wonder how many people will still feel comfortable going 100% equities or even stock picking, when "the crash" finally hits.
🤯⤵️ Sounds easy on paper "YEAHH I'M GETTING CHEAP STOCKS SO I WILL BUY MORE!" but I don't think anyone really knows how they will behave until they're actually facing it.
Did it though? I still know quite a few independent bookstores that are open, but like I said they do a lot of selling online now. Same thing for anyone in any field based on collectibles. Much to my chagrin, eBay has helped my dad build a massive card and comic collection that would not have been possible if he had to keep buying from physical locations.
