There are, of course, concerns about the Music Genome Project from some in the music world. First, since the Genome is proprietary, there's no possibility for independent review. For all we know, Pandora Media's "experts" don't know the difference between syncopation and vamping. Also, in a much broader way, the Music Genome Project assumes that music's traits can be objectively analyzed at all -- that the mind of the listener can be left out of the equation. Some experts doubt that music can be quantified in this way.
Regarding the player itself, an issue arises when you create a station using an artist like The Beatles as the seed. Some artists have such a varied collection of styles that there are endless ways the Genome algorithm can go when determining matches. In such a case, Pandora may return music you don't like at all. For instance, if you love the Beatles' later stuff like "Across the Universe" or "I am the Walrus," you'll probably be disappointed if Pandora returns music that's similar to "I Want to Hold Your Hand." For this reason, it's often wise to use a song instead of an artist as a seed.
With all of Pandora's attractive features and novel approaches to personalized radio, it does tend to wow people when they discover it. But wows don't pay the bills, and Pandora will have to bring in real cash if it's going to survive. In the next section, we'll find out how Pandora plans to turn the academic Music Genome Project into a commercial commodity.