I'm a doctor currently working in a startup that is very likely going to replace doctors in the coming decade. It won't be a full replacement, but it's pretty clear that an ai will be able to understand/chart/diagnose/provide treatment with much better patient outcomes than a human.
Right now nuance is being implemented in some hospitals (microsoft's ai charting scribe), and most people that have used it are in awe. Having a system that understand natural language, is able to categorize information in an chart, and the be able to provide differential diagnoses and treatment based on what's available given the patients insurance is pretty insane. And this is version 1.
Other startups are also taking action and investing in this fairly low hanging apple problem.
The systems are relatively simple and it'll probably affect the industry in ways that most people won't even comprehend. You have excellent voice recognition systems, you have LLM's that understand context and can be trained on medical data (diagnoses are just statistics with some demographics or context inference).
My guess is most legacy doctors are thinking this is years/decades away because of regulation and because how can an AI take over your job?
I think there will be a period of increased productivity but eventually, as studies funded by ai companies show that patient outcomes actually have improved, then the public/market will naturally devalue docs.
Robotics will probably be the next frontier, but it'll take some time. That's why I'm recommending anyone doing med to 1) understand that the future will not be anything like the past. 2) consider procedure-rich specialties
[link] [comments]