Yes, AI is useful, and it can be programmed for more satisfactory instances, but would you put your trust into someone’s hands who only relies on AI? There’s a reason doctors, engineers, and scientists study for years; the information they obtain should be earned, not just given directly.
Imagine walking into a doctor’s office and announcing the issues you’re facing, suddenly they pull out their phone and ask AI what sickness or injury you have. Imagine in the middle of a life threatening surgery and the anesthesiologist doesn’t know the exact dose measurement, so they ask AI? This is risky for society. AI makes mistakes, constantly, and that’s because it's an online operating system. It does not have the truth, it just obtains common issues.
AI lacks the knowledge doctors have. It lacks real situations because it can’t experience them. Putting trust into an operating system is completely unsafe and will only repay you with guilt of not diagnosing your patients correctly or potentially killing them. Using your connections with actual human beings is much more safer and impactful.