2025-11-23
malencolia
Derek Thompson hosts a podcast roundtable including Kelsey Piper, discussing her recent articles on education reform. In general Kelsey doesn’t need any commentary, but this episode was particularly good in that it particularly focused on standardized testing, and how making them optional or even banning them is the primarily reason why grade inflation goes completely unchecked. There is a section where Derek asks Josh Goodman about equity concerns, but if I recall correctly another major argument against standardized testing was that it encouraged “teaching to the test”, and although that at least ensure that students are learning something (which is more than nothing), this is probably the standard defense that people are going to make now that equity appeals are somewhat less in vogue.
Ben Goldhaber reviews of How Big Things Get Done, which he argues fails to address the central problem of people from siphoning out the budget. Somewhat related, Sabrdance continues his series on political machines, describing how GooGoo reforms forcing federal funding to go through diffuse sources favored Democratic political machines over Republican ones.
Byrne Hobart reviews The Age of Extraction and it’s thesis that sufficiently large and inscrutable businesses are making money not through value creation but through monopolistic rent extraction. Amanda from Bethlehem has two related articles: a review of Kiki’s Delivery Service, which describes two possible conclusions of the idea that succeeding under capitalism should be easy, depending on whether one believes this is actually possible; also, and article on social coercion which applies quite well to economic “coercion” as well1.
Econboi hosts a debate between Bryan Caplan and Matt Bruenig on individual property rights versus goods held in common, and what those different viewpoints imply about how welfare programs should be structured. Related, Camille Berger has a post about how left and right political views are really about conflicts between people’s beliefs about how relationships should be structured, which intuitively isn’t something which should be imposed. That being said, it’s hard to avoid doing so as long as people feel tied to their land or community, in which case “a change of surroundings” doesn’t imply movement, but control.
Justis Mills describes preference set points, the phenomena where one feels that, for example, that people neater than you are terribly fussy while those more casual are inveterate slobs. Related, Vortex Goddess describes her set point for required contact frequency to prevent friendship decay2.
David Chapman on the transformative power of relationships, specifically how he learned to view people non-transactionally, and also how to maintain his own self and independence despite that.
Extelligence on panpsychism as a framing to reason about the behavior of things, and what it implies for LLM consciousness. This seems to be related to what Andrej Karpathy is getting at with his description that LLMs are intelligent but not in an animal-manner; in that things like intelligence or consciousness probably do not exist within a strict hierarchy or binary.
Rohit Krishnan presents evidence that Gemini is superintelligent3.
Charli XCX writes about the experience of being a pop star.
Jenneral post-mortem for her father in his role as her dad.
Eneasz Brodski with a story of applied sex-positivity.
Étienne Fortier-Dubois linkthread.
Emergent Ventures cohort linkthread (includes Justin Kuiper, who has a review of Downtown Owl today).
What’s interesting is that many don’t seem to see it that way, like some socialists who seek to free us from rent-seeking monopolies through online cancellation mobs, or conversely, various free-market economists who nevertheless strenuously object to euthanasia as a subject. Edit: While on this topic, Andrew Heaton interviews Jeremiah Johnson on the End Kidney Deaths Act.
Interestingly, all the authors’ set points described in these articles seem to match my own.
I am Sam Kriss. You know, even though I’m enjoying reading the Inkhaven blogs I’ll also be glad when it’s over, because I feel myself starting to overfit; my mental phrasing is beginning to produce things like “recursive self-improvement is bounded because there is a tradeoff between becoming a better optimizer and improved meta-optimization”. This is why I don’t read LessWrong.
Wow, standardized tests! What if they were adaptive AI?