As DevOps matures, and the "shift-everything-left" philosophy gains ascendancy, there's a movement to automate all—or most—phases of testing. However, there are some critical functions that may resist automation. In fact, contrary to the "automate-everything" impetus, human testers won't ever go away. Indeed, human testers need to be more involved, and earlier in the development process. In this podcast, Mike Kavis and guest, Angie Jones, discuss the human aspect of testing and how humans add significant value by assessing the system as a whole, helping developers design better code, and determining the level of testing automation that should occur. Angie also shares the testing automation matrix she has developed. Finally, they cover testing of machine learning algorithms—ways to help prevent or reduce bias and make algorithms more effective, and the emerging field of visual testing, which uses humans to ensure that graphics-heavy apps function and appear as designed.