“Dynamite saxophone solos in rock n’ roll songs” kinda died, per 30 Rock, back in the 80’s, but thanks to my colleagues in machine learning (who are truly doing God’s work 🙃), perhaps they’ll soon be back:

I’m told that the model requires ~10-15 minutes of audio training data, and once it’s ready you can apply a transfer in around 2-3 seconds.

Perhaps now we’ll all be Jack Blacks (Jacks Black?) 🎷

[Via Sehmon Burnam]