TFX: A TensorFlow-based production scale machine learning platform

Brilliant!

  > We want the user to treat data errors with the
  > same rigor and care that they deal with bugs in code.
  > To promote this practice, we allow anomalies to be
  > filed just like any software bug where they are 
  > documented, tracked, and eventually resolved. 

If a human screws something up we expect that person to fix it. If an A.I. screws up, we're like "we're within 2% of the best score on Kaggle so we're doing pretty good".

A.I. is not going to replace humans until we have a disciplined process to get it to work right.