Not when there are a lot more possible wrong answers than right answers.

That doesn't follow. 2+2 has more wrong answers than right answers, and yet, if you gave me the wrong answer to arithmetic sums half the time I'd think you were a lousy calculator. Worse still if those answers had the potential to kill me.

Yes, but we are not comparing half time right to nothing, we are comparing how often AI right to how often your local human right.

(not the smartest expert human, the human you have access to)

And you believe your local human aka the doctor is right only half the time or even worse?

I think the parent believes the opposite.

It depends, half the time right is very useful in scenarios where you can quickly verify the correctness - i.e. with troubleshooting suggestions or code which you can run and see if it does what you needed.