What do you plan on doing, when intelligence is to cheap to monitor?
The Library is my personal archive of essays, books, tweets, quotes, podcast transcripts etc. that I have not yet made public—but I plan to!
This Library is a living archival project. It is organized to unobjectively draw connections, inclined toward my point of view of the world. It is somewhere between a wiki, an encylopedia, and libgen. I believe it is rhizomic—though I don’t claim to have read D&G. Some other living archives I am inspired by are
- wisdomlib
- gwern’s blog posts
- art and popular culture
- tufts’ perseus
- online etymology dictionary
- wikipedia and other wikis
While the first two examples are curated by a single person, most such sites require a tremendous amount of crowdsourced effort. When intelligence is to cheap to monitor we have a chance at a much more ambitous project. But it cannot be done carelessly. There is zero-tolerance on slop, and so the use of AI must be highly specialized. Namely: it should not be the one writing, summarizing, reviewing, or teaching. It is there to transcribe, translate, organize, and draw connections.
In the last few centuries, highly trained philologists and classicists could ascertain connections between the thought of Nagarjuna and Pyrrho; or could read Plato in the Septuagint. This is no easy task—I do not expect results of this magnitude from an LLM yet. But who is to say that a certain passage of Machiavelli might not find eery correspondence with a yet-untranslated Zhou dynasty text? Or matchings in observations of celestial correspondences Islamic and Mayan recordings…
It is impossible to know the scope of such connections a priori. I believe that even a few well-read Greek texts—works of Pindar, Aristophanes, Plato—are still near endless in their connections within each other. There is a reason Lord Bryon, Nietzsche, Jefferson, BAP—scholars of Greek—find themselves to be such radicals of thought outside their time.
The perennialists have long speculated about the universiality of religions… of secret agreements in the fate of the human soul and the nature of the Godhead. I have no reason to doubt this, but do we have reason to suspect that a true perenneliast project has been attempted? One that treats the remarkable similarities in Egyptian rites and Tibetan Bon with the same respect as theological Catholicism?
In fact, this issue of importance, or weighing of attention might be genuinely solveable. If one can draw connections between texts in a corpus, one can also measure novelty and succession of thought. What will be able to do, when we have mapped out the manifold of human thought and assigned a metric of similarities and novelties? Can Nietzsche be assigned a tuple, a score of similarity and a score of novelty? Who are the peaks, who are the valleys? Maybe we can prove Rufi’s poetry is truly preeminent of human thought… or perhaps that the Bard had surpassed him…