done

Despite really annoying cold symptoms that have kept me home from work, partially in bed and partially half-dozing in the recliner, I’ve finished mastering the rest of the tracks — and recorded one more short one, because I felt like something with a particular character was called for.

The process mostly got easier as I went. There’s a sort of iterative logic to it that is fairly similar from one track to the next, though specifics vary. Fixing my stereo phase faults took me hours to figure out the first time, but seconds now that I know what generally works. And I was more cautious in setting up my effects on the new track and didn’t create trouble for myself in the first place.

There’s one note I have to check into with fresh ears; other than that I’m just waiting on cover art and need to write up some text. And then on to the next project!

Other than that, I’ve been reading a lot and TRYING to sleep.

I finished Last Dance, a novel about an Earth-to-Mars transport ship, its ornery captain and his rebellion. Pretty good!

I read a couple of short books on mixing and mastering, which were not awful but less helpful than the advice and practical knowledge I’d already gained. Meh.

I also read You Look Like a Thing and I Love You, Janelle Shae’s (of AI Weirdness) humor/non-fiction book on the current and likely future state of AI. She says “the danger of AI is not that it’s smart, but that it’s not smart enough” and that certainly seems true. It feels like there have been amazing breakthroughs in machine learning in the past few years, but they’re all extremely narrow in scope, easily fooled, have no sense of value judgement other than what they’re programmed to seek or avoid, are highly susceptible to bias and extraneous information in training data, tend to have very small memories, it can be difficult to understand why they made their specific decisions, and approximately as smart as a worm. On the other hand, there are some problems they can solve… as long as a human verifies them.