Lukasz Stafiniak

First letters found in new scroll PHerc. 172 (Scroll 5) is here.

x.com/natfriedman/...
via Nat Friedman:
scrollprize.substack.com/p/first-lett...

The paper on data-race-free OCaml is out! If you want to learn more about the zoo of new mode axes (contention, uniqueness, portability, ...) and cool new abstractions like capsules, it's all there!

iris-project.org/pdfs/2025-po...

7. Consciousness and the Social Brain by Michael Graziano (2013)
8. The Centered Mind: What the Science of Working Memory Shows Us About the Nature of Human Thought by Peter Carruthers (2015)
9. Human and Animal Minds: The Consciousness Questions Laid to Rest by Peter Carruthers (2020)

4. The Architecture of the Mind: Massive Modularity and the Flexibility of Thought by Peter Carruthers (2006)
5. How Can the Human Mind Occur in the Physical Universe? by John R. Anderson (2007)
6. The Conscious Brain: How Attention Engenders Experience by Jesse J. Prinz (2012)

Philosophy of Mind adjacent books I like: (3 of 9)
1. Phenomenal Consciousness: A Naturalistic Theory by Peter Carruthers (1996)
2. Breakdown of Will by George Ainslie (2001)
3. Being No One. The Self-Model Theory of Subjectivity by Thomas Metzinger (2003)

(BTW, "lowerer", while a bit awkward, is a good name for "converts into a low-level representation".)

Thanks, that makes sense! I carried into this my definition of scheduling as ordering and/or synchronizing the computation of kernels, or materialized buffers, based on a topo-sort of inputs-outputs.

But it might well be worth it! I'm warming up to things as I get more familiar with tinygrad. Maybe the steeper learning curve pays off and my original comment was unwarranted.

(2) Before PR 7065, engine/schedule.py is preparing indexing aka. movement ops, is integrating indexing with computation. Why is the file called "schedule" when it doesn't do scheduling? I probably mis-generalized from this, to the feeling that it's surprising where things happen in tinygrad.

(1) Axis representation combines user-facing buffer-specific semantics (index into a tensor), with operation-specific semantics (whether axis is reduced), with tunable computation details (global mem vs. local mem, whether computation is upcasted, when the reduction happens: local vs late).

A steeper learning curve because often a single representation interacts with various stages of processing. Two examples below.

Much as I like tinygrad, it's primary sin is not code density as some claim, it is non-local design.

Synphaera Synphaera is an independent record label for electronic and ambient music. For booking and license inquiries: info@synphaera.com

My favorite coding music label:
synphaera.bandcamp.com

So how to square this with Tesla's published statistics: FSD users have half the accidents of non FSD users? Until a few months ago -- certainly during the reporting here -- FSD was only granted to people with good safety score. That would cut inferring that FSD is good, but still a defeater.

12:45 Restate my assumptions. (a) The appropriate tool for the representation of scientific theories is mathematics. (b) The relationships between successive theories, and theories at different scales whether spatio-temporal or energetic, are often limiting relations and similarities of mathematical structure (formally captured by structure-preserving maps or morphisms of various kinds), rather than logical relations between propositions. (c) Theories, like Newtonian mechanics, can be literally false as fundamental physics, but still capture important modal structure and relations. (*) Therefore, there are patterns, everywhere in nature. (*) quote from Ladyman & Ross

What if I told you your Goedel number Matrix meme

Memes I made.

metta-lang.dev is a homoiconic, symbolic (partial evaluation), functional (immutability, first class functions, pattern matching), logic (first-class variables and multiple results) programming language.

#1000 Karl Friston: The Free Energy Principle and Active Inference: From Physics to Mind YouTube video by The Dissenter

I think The Dissenter deserves more support for his prolific and wide-ranging interviews. One example:

www.youtube.com/watch?v=2Bzm...

Mind mapping: one page of pithy notes organized into a graph.

Coding music sprint: put on music, start a timer, completed with a good line of code or an idea that moves me forward.

My current daily stack:
* Exercise
* Mind mapping
* Coding music sprint
* Chores or a nap
* Work outside for a while (a park or a cafe)
* Meditation
* Stretching

I decided to lean into atomic habits. An atomic habit is where you are supposed to get some satisfaction within 2 minutes, and call it complete in a few minutes if you're under time pressure, to let them stick.

AI Interpretability, Safety, and Meaning - Nora Belrose YouTube video by Machine Learning Street Talk

@thedavidsj.bsky.social you probably know, but you have a shout-out in the latest Machine Learning Street Talk episode! Congrats!
www.youtube.com/watch?v=VgPr...