Another Maciej talk: Haunted by Data

I wonder what Jason Scott thinks about all this (I know there's difference between private user data and public web pages, most of the time.)

Most interesting part to me:

But Nature is full of self-modifying, interlocking systems, with interdependent variables you can't isolate. In these vast data spaces, directed iterative search performs better than any amount of data mining.

My contention is that many of you doing data analysis on the real world will run into similar obstacles, hopefully not at the same cost as pharmacology.

Reminds me of something I've heard from multiple people working on data-driven stuff in consumer web stuff: "every variation performed worse than the baseline."

"Directed iterative search" I'm pretty sure means "We'll just try them —and if they don't work — why then we'll just try something else."