Sample-based music isn’t stealing. It’s valuable and important. It shows the way toward a future for recorded music that’s more in continuity with music’s past. Recordings are cool and everything, but they encourage passivity. If I buy a recording, I can listen to it or dance to it, both fine activities, but what if I want to go further? What if I want to engage with it, converse with it, customize it or adapt it to my own needs? According to the law, I can’t. This flies in the face of the uncountable centuries of music practice that predate the invention of recordings. Before recordings, if you wanted to hear music, someone needed to play or sing it. To learn how to play or sing, you have to learn and interpret a ton of music by other people. The normal method for passing music along for nearly all of human history was by oral tradition, and a lot of adaptation and reinterpretation was an inevitable part of this transmission process.
In the modern world, most of the music you encounter is in recorded form. Adapting or customizing music is going to continue as it has for uncountable centuries. To adapt or customize a recording usually requires sampling. As it stands, the law is in the way. We need open-source music like we need open-source software.