US doctors want neural network developers to be held accountable for errors in medical AI, not them.
The American Medical Association (AMA) decided that if a doctor used a neural network and he made a mistake, the manufacturer of the neural network should be held accountable. The logic is clear, but the developers will be extreme. This could lead to developers banning their AI from discussing medicine with users at all.
I really like this joke:
- Doctor, how can I cure it?
- I'll google it now.
- Or maybe I myself?
Let's not self-medicate. There is a paradox in this story. AI can give the doctor an explanation, advice, but not an order to act in a certain way. The doctor can evaluate it, ask a colleague or another neural network, or bring the question to a consultation, but he still makes the decision. And he is responsible for this decision.
If a doctor, without his own opinion, gives the patient the answer of the neural network as his failure and at the same time does not want to be responsible for the result, then it is not clear why such a doctor is needed at all. It is easier to immediately send the patient to consult an AI.
It's funny, but the use of AI by doctors is already a reality. According to the AMA survey, 81% of American doctors use artificial intelligence in their work, compared to 38% in 2023. It is often used to refer to research and standards of care.
If neural networks stop giving advice about medicine, the medical care will become less quality. There are not enough doctors: in Russia, according to the data of the Ministry of Health, there is a shortage of more than 23 thousand doctors, and in the world by 2030 there will be a shortage of 11 million health workers. Everyone's skills are different. In medicine, as in all professions, there are smart experts, and there are not so smart ones. The doctor, like the patient, also needs to consult someone.
We must understand that the use of artificial intelligence in medicine is an objective process, and it cannot be prevented.
"Oleg Tserev".