most trajectories are not orbits
most functions are not differentiable
most programs do nothing
the universe is a desolate wilderness. we just live in the interesting/useful/meaningful parts of it.
Suppose AI can read sheet music & MIDI/Ableton files. Further suppose it can compose/produce music in such a digital/notational format, at a level such that a typical person would judge it competent (even if they don't like it).
Does the AI understand music?
This is Mary's Room in real life with music instead of color. I'm endlessly amused by the thought experiments from philosophy which have become real-life questions for AI.
My personal vibecoded tools now mostly take the form of read-only dashboards. For the write/mutate path, I don’t even bother with UI, I just directly tell Claude to make some update to the data.
My doctor: “What you have is no big deal, I’m not worried about it, it’s called X, you can ask ChatGPT about it if you care.”
That was a new one for me (doctor delegating to ChatGPT)
My personal vibecoded tools now mostly take the form of read-only dashboards. For the write/mutate path, I don’t even bother with UI, I just directly tell Claude to make some update to the data.
Anyone who’s worked in a large organization knows that, at the boundaries of different teams’ responsibilities, stuff falls through the cracks. The same phenomenon occurs where multiple SFPD precincts’ literal borders meet: sfchronicle.com/sf/article/h…
Fable and Astra are happy to make up jargon (too eager, in Fable's case) but neither seem good at constructing new vocabularies a la Rorty, or Bloom's poets. I want to say that's interesting but maybe it's just a consequence of the ML paradigm.
Are there alien types of math that are more akin to biology than the beautiful math that we tend to study? Math that is unruly and inscrutable but has layers of depth and structure and is ultimately useful to understand and try to tame?
Linguistic prescriptivism should be subdivided into
Traditional Prescriptivism: follow these rules handed down from our ancestors
vs Progressive Prescriptivism: speak in this new fashion for a better future
New wacky theory: the idea of qualia only occurs when you have an idea of interiorality which comes from silent reading, and not before
That’s why we get Cartesian Dualism after the printing press
Also why one of the rare pre-Cartesian accounts of internal thought appears in Augustine
a lot of the surprise of ai x math advances comes from people not internalizing test time compute scaling properly.
inference compute per problem has gone up sharply this year (consumer gpt in jan, $2k for unit distance, $15m for ns) partly bc ttc scaling has improved a lot
In personal news, I had a baby three months ago. Since many of my friends don't know what a baby is like, I wrote a little about it here: noahtye.com/i-have-met-the-b…