actually good music are blocked out by money burning singers whose voices around like a dolphin trying to speak.
I think there's a degree of irony you aren't consciously aware of in lodging this complaint (that vocals in modern music sound too chaotic and visceral) while also complaining about the use of computers in music.
In music, there are those who believe that there's one objectively correct way to, say, play Chopin's Raindrop Prelude. And I disagree.
Now, at first glance this sounds nonsensical, since of course there's sheet music for this piece and if you just did your own thing it'd no longer be Chopin's Raindrop Prelude. But I name this piece because it's the perfect candidate for a small degree of controlled chaos. If there's a slight delay in striking certain notes, it just sounds richer (and more suitable to the theme of rain) than if you were to play it strictly per the notation.
Those who wanted the perfectly rendered Raindrop Prelude would have no reason to listen to a world class pianist play it live, since any computer can execute the same maneuver flawlessly. But what do you get? A completely predictable experience. Every note in a given segment is the exact same volume, and the exact specified length. Nothing at all unexpected happens. It's perfect to the point of being a tad boring and unspirited.
I would suggest that in the future (heck, maybe in 2 or 3 years), AI will be capable of mimicking technically perfect human vocals as well. When this AI vocalist is on the note G per the notation, they sing exactly the note G. There is never an improvised "trill" effect for lack of a better word.
When this day comes, I suggest it will be the very fact that a human voice is unlike computerized vocals that sells human music. And I don't mean this in the sense of people buying subpar human music out of a feeling of obligation to support human artists, but rather as a matter of authentic preference by many listeners.