This 2007 interview with Steve Albini, engineer of 1000s of records including work by Nirvana, The Stooges, Pixies, PJ Harvey, etc., is full of interesting ideas, probably applicable toward many creative activities. Some of my favs:

On why he still works on so many records at relatively low rates:

“If you think about records that are made over months for superstars, those are horrible, horrible records […] whereas records that you knock out in a weekend, a lot of those are pretty good records, and more importantly, if you work on 50 or 100 records a year you have a much better likelihood of a couple of them being really great experiences than if you work on two records a year. If you’ve worked on two records a year and they both suck, then you’ve blown a year.”

On his production style:

“Most bands are plenty good, they don’t need a lot of production. Most bands, if you allow them to do what they do naturally, you’ll get a pretty good representation of the band, and generally speaking it’ll be a satisfying experience.

When you start deconstruction a band into its component parts and parsing music out into lyrics and choruses and verses and riffs and bridges and turnarounds and fills and modulations and stuff, then you work on all these elements individually and then try to reassemble them into a simulacrum of what the band was doing organically. That makes freakish records that don’t represent the bands.”