I wish you researched this more guys. Suno was the wrong tool to focus on for starters - anyone semi-serious in music considers suno a plaything. Udio, on the other hand… i can give you plenty of examples where its impossible for any outsider to accurately determine whether it's human-made or not. But the biggest issue i have is this. The current set of tools is the worse it will ever been from this point onwards. The next generation of this stuff, where the intelligence is in the application (and enhancement) of music theory rather than just generative - that is what is going to blow people's minds. Let's revisit your position in 6 months
Firstly, thanks for watching! You raise some valid points, and we’re actually probably closer in seeing it your way than what you might think. Udio does seem to offer more variety at the moment, however in general we’re still struggling to find any generated music that we would listen to on repeat. It would be awesome to be proven wrong (esp. in six months) - and the technology is crazy cool, but the high bar HAS TO BE for this to generate the best albums of all time on demand with minimal human involvement, otherwise what is the point? It feels like reaching this outcome might be tied to achieving AGI, which is not a given. In the meantime, these tools will certainly help semi-talented musicians be better, which is actually a good thing!
I wish you researched this more guys. Suno was the wrong tool to focus on for starters - anyone semi-serious in music considers suno a plaything. Udio, on the other hand… i can give you plenty of examples where its impossible for any outsider to accurately determine whether it's human-made or not.
But the biggest issue i have is this. The current set of tools is the worse it will ever been from this point onwards. The next generation of this stuff, where the intelligence is in the application (and enhancement) of music theory rather than just generative - that is what is going to blow people's minds.
Let's revisit your position in 6 months
Firstly, thanks for watching! You raise some valid points, and we’re actually probably closer in seeing it your way than what you might think. Udio does seem to offer more variety at the moment, however in general we’re still struggling to find any generated music that we would listen to on repeat.
It would be awesome to be proven wrong (esp. in six months) - and the technology is crazy cool, but the high bar HAS TO BE for this to generate the best albums of all time on demand with minimal human involvement, otherwise what is the point? It feels like reaching this outcome might be tied to achieving AGI, which is not a given.
In the meantime, these tools will certainly help semi-talented musicians be better, which is actually a good thing!