HN user

folkstack

2 karma

contact me: info at folkstack dot com

Posts2
Comments7
View on HN

Extempore was almost my jam when I got started with code-music in 2014, but javascript was just good enough, and the web audio API soon had a script processor, so I started building my own stack in the browser and never looked back.

But I was impressed, and envied the processing power. However, digital musical interfaces are really what inspired me to code music in the first place, the idea of building them (which I do now).

I actually installed and tried extempore, and that almost got me into emacs, b/c the author had integrated it. Smdh. Luckily, I abandoned both, and have used vim ever since. I had looked at all the weird code music editor systems, deemed extempore the best, but ultimately I had to go raw and work my way up.

PhDs are over-valued/hyped for ML, especially as it is currently the practice of brute forcing undergraduate maths via CS fundamentals.

ML is a creative field of algorithm composition, and the hardest part is the actual programming. These are not the typical skills of a PhD student.

The current pace, and variegate provenance, of ML research suggests that a PhD student is going to be, at best, a scholar of the web wasting too much time on academics.

At least one of two things will happen in ML: there will be a singular advance in theory that opens up a world of practical applications (like QM), xnor the field of ML will be open to endless synthesis (like programming and music).

All that said, no doubt a PhD will pay for itself...