I have been thinking a lot about this concept of bullshit (or some similar term) jobs, even before I came across Graeber's book. Although I would argue that it's more the level of bullshit within a job that is more important. I guess there might be jobs that are 100% bullshit, however most jobs will be on a scale where for example 80% of the tasks are bullshit, but there is also some level of "useful" output.