Current AI is also very confidently incorrect.
I have to second that.
A.I can be useful to _get an idea_ about something but, it is _crucial_ to keep in mind that it is _very often_ wrong, sometimes in obvious ways and sometimes in very subtle ways (which is a lot worse.)
Yes, A.I can help but, it cannot and must not be trusted to provide correct answers. Sometimes it does but, more often than not, there are problems in the answer that range from obvious (which is great) to subtle (which is difficult to catch even for someone who knows the area because, something subtle is easy to overlook.)
Bottom line: A.I is a know-it-all that _never_ says it doesn't know, will answer no matter how preposterous and ill-conceived the answer is. Conclusion: use A.I with _extreme_ care.
BTW: no matter what the answer is, it stole it from someone out there, reworded it and it is now assertively passing it as its own.