> b) the nicely looking graphs they saw are all pure bullshit, and the things that looked correct only seemed so, and only by accident. They’re not “bullshit”

> b) the nicely looking graphs they saw are all pure bullshit, and the things that looked correct only seemed so, and only by accident. They’re not “bullshit” per se, they’re just “prototypes” for further refinement :) The road to success is trodden in stumbling half-drunken steps on the pebbles of failures.