AnhrithmonGelasma
Senior Member
The way Tom Waits works with musicians on his songs reminds me of using generative AI text prompts and likes/dislikes:
"He never suggest notes, he never suggests a part, he never sings a line, he never says what to play, but he is what I imagine an actor might experience when working with a great actor's director... he's very in your face, and you play something and he reacts, and his facial expression gets extreme, nah, yeah yeah yeah that's it, could you make it sound a little more poor, oh oh don't get on the Joni Mitchell train now, it's this constant patter and he never actually says what to play [except for the instrument] ... oh, my play my instrument ... he would ... I show up with the banjo, and he would say could you play my banjo ... and he has a 100-year-old banjo with 100-year-old strings"
... but there aren't any (that I'm aware off) that will let you give them textual (or facial) feedback like that yet....
"He never suggest notes, he never suggests a part, he never sings a line, he never says what to play, but he is what I imagine an actor might experience when working with a great actor's director... he's very in your face, and you play something and he reacts, and his facial expression gets extreme, nah, yeah yeah yeah that's it, could you make it sound a little more poor, oh oh don't get on the Joni Mitchell train now, it's this constant patter and he never actually says what to play [except for the instrument] ... oh, my play my instrument ... he would ... I show up with the banjo, and he would say could you play my banjo ... and he has a 100-year-old banjo with 100-year-old strings"
... but there aren't any (that I'm aware off) that will let you give them textual (or facial) feedback like that yet....