"The difficulty is that machine learning is a fundamentally hard debugging problem. Debugging for machine learning happens in two cases: 1) your algorithm doesn't work or 2) your algorithm doesn't work well enough."
Read it!"Machine Learning can give you an edge, but probably not an edge so huge that it allows you to ignore other factors. You still need to build good infrastructure, collect good data, and have low latencies."
Read it!A critique of dynamite plots and suggestions for better alternatives.
Read it!