I tried the Aleph0's with ML Quests, and it might have been a bad match or underpowered, but the results were interesting. When friends came to listen, they would say the Aleph sounded great, but then they would talk while the music was playing. That is, the system had lost its ability to captivate. I always thought a similar thing might happen in A/B tests where the mind arrives at one conclusion if it is tasked to judge, but arrive at another conclusion if the measurement is pleasure.