{
  "blyg": "0.3",
  "id": "7j3fqzt3nn3qzd9e9tbdydx4dw",
  "kind": "fragment",
  "origin": "https://www.msweet.net/notes/",
  "page": "092-jade-geese",
  "author": {
    "name": "Matthew McDowell-Sweet",
    "url": "https://www.msweet.net/"
  },
  "created": "2026-09-26T16:55:45Z",
  "updated": "2026-09-26T16:55:45Z",
  "version": 1,
  "content_md": "I’m a big fan of the ethos behind [Simon Willison’s tiny tools](https://tools.simonwillison.net/). The idea is that we’re now empowered to solve our own hyper-niche problems and when we do it’s helpful to share the solutions with others.\n\nI also like these stances from [Thomas Ptacek](https://sockpuppet.org/blog/2026/08/20/stop-making-tuis):\n\n- “I’m not packaging this application up. If you want it, just screenshot this section of the post and give it to Claude. It’ll build something useful.”\n- “The truth is, I barely think about these things as ‘apps’ (I have no intention to distribute them). They’re artifacts of me making my computer do stuff for me, the way I want it to.”\n\nFinally, the concept of arbitrariness in the hard sciences is relevant: sometimes a decision or choice is made out of pure, unadulterated convenience because what really matters is the [fulfilling of intent](https://aneeshsathe.com/the-lightening-of-intent) via an outcome.\n\nThese three taken together form what I’m now thinking of as “arbs”: a sequence of incantations given to a machine once to make something happen reliably many times.\n\nArbs are cousins of tiny tools. Both exist for a narrow purpose. The difference is that a tiny tool’s form is *part of what’s shared*, while an arb’s form is *arbitrary in the physicist’s sense*. The most convenient choices are made to get to the outcome quickly. So there’s no repository to expose. What gets shared is the intent.",
  "content_html": "<p>I’m a big fan of the ethos behind <a href=\"https://tools.simonwillison.net/\">Simon Willison’s tiny tools</a>. The idea is that we’re now empowered to solve our own hyper-niche problems and when we do it’s helpful to share the solutions with others.</p>\n<p>I also like these stances from <a href=\"https://sockpuppet.org/blog/2026/08/20/stop-making-tuis\">Thomas Ptacek</a>:</p>\n<ul>\n<li>“I’m not packaging this application up. If you want it, just screenshot this section of the post and give it to Claude. It’ll build something useful.”</li>\n<li>“The truth is, I barely think about these things as ‘apps’ (I have no intention to distribute them). They’re artifacts of me making my computer do stuff for me, the way I want it to.”</li>\n</ul>\n<p>Finally, the concept of arbitrariness in the hard sciences is relevant: sometimes a decision or choice is made out of pure, unadulterated convenience because what really matters is the <a href=\"https://aneeshsathe.com/the-lightening-of-intent\">fulfilling of intent</a> via an outcome.</p>\n<p>These three taken together form what I’m now thinking of as “arbs”: a sequence of incantations given to a machine once to make something happen reliably many times.</p>\n<p>Arbs are cousins of tiny tools. Both exist for a narrow purpose. The difference is that a tiny tool’s form is <em>part of what’s shared</em>, while an arb’s form is <em>arbitrary in the physicist’s sense</em>. The most convenient choices are made to get to the outcome quickly. So there’s no repository to expose. What gets shared is the intent.</p>",
  "content_hash": "sha256:dc16b48e2b76063e6cec1c7f522b4c80a7ca88f0af9837e6cceb0c276107eade",
  "media": [],
  "changelog": [
    {
      "version": 1,
      "at": "2026-09-26T16:55:45Z",
      "note": null
    }
  ]
}