Journal — Sympolymathesy, by Chris Krycho

https://v5.chriskrycho.com/journal

5 posts

Life 80% · Tech 20%

Subscribe via RSS

  1. [journal] Impressively Capable, Impressively Foolish

    A little report on the state of contemporary LLM-based tools. Yesterday, I decided to try to use Claude to automate a workflow I do a ton in my composing: Export the mockup audio from my notation software as a WAV file Import it into an audio editor and bump the gain a bit. Export it as an MP3. Add the relevant ID3 tags. Import it into Apple Music. Fix one thing that the ID3 tag editor I use doesn’t handle: setting the track to use “work name” rather than “title” as the primary field — pretty…

    0
  2. [journal] You Can Rightly Judge Things You Have Not Tried

    This actually isn’t that complicated; stop letting people lie to you.Assumed audience: People familiar with and potentially open to (as I am committed to) a Christian view of ethics, informed heavily by the virtue ethics tradition in particular. Many technology-enthused nerds often offer as a truism the idea that you have to have familiarity with a given technology to be able to judge it rightly. This is nonsense. Do you have to try cocaine to judge whether it is good for you? Do you have to…

    0
  3. [journal] 2026 Boston Marathon Race Report

    I did it, and I did it fast! Assumed audience: People interested in endurance running (or endurance sports more generally), or who just like to hear what I am up to as a runner. I ran the Boston Marathon yesterday! It went great! Apparently my body likes sea level and going downhill! The race went fast on multiple levels. The short version It was a blur of cheering and noise and being continually astonished at how fast I was going, the miles ticking away unbelievably quickly from the sheer…

    0
  4. [journal] Humans Are Amazing

    We took a family ski trip and it provoked no little astonishment.

    0
  5. [journal] Noticing, and Solving, Technical Problems in a Codebase

    And a requisite aside about large language models.

    0