AI Article Readings

In this short story, Max Harms tells the strange, funny, and quietly moving tale of Pemberton, an elderly butler who has spent fifty years keeping his dead master legally “alive” so that a fortune can continue flowing to a village in India. When Krishna arrives from the Office of the Preserver to audit the household’s dharmic account, Pemberton’s elaborate machinery of forgery, puppetry, and moral compromise finally comes due. What follows is a sharp, humane fable about duty, fraud, devotion, and the terrible difficulty of doing good by wrongful means.https://open.substack.com/pub/raelifin/p/the-keepers-dharma?r=67y1h&utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this post, Scott Alexander lays out his current views on AI timelines, safety, geopolitics, and possible futures, offering probabilistic forecasts for AGI, superintelligence, diffusion, doom risk, AI pauses, and post-scarcity outcomes. He presents himself as worried but not maximally pessimistic: expecting transformative AI within decades, seeing serious alignment and misuse risks, but also leaving substantial room for successful safety work, international coordination, and even utopian possibilities.* 00:00 - Introduction* 00:17 - Timelines* 13:13 - Safety* 25:59 - Geopolitics* 33:15 - Other Outcomes Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this post, Dario Amadei argues that AI’s exponential progress is now moving far faster than political institutions can comfortably respond, creating an urgent need for serious policy action. He sets out five areas where governments need to rethink their approach: frontier-model safety regulation, job displacement and macroeconomic policy, accelerating beneficial scientific uses of AI, protecting civil liberties from AI-enabled state or corporate power, and securing democratic leadership in the global AI race.* 00:00 - Introduction* 05:34 - One. Regulation and public safety* 11:41 - Two. Macroeconomics and tax policy* 19:08 - Three. Accelerating AI’s positive impact* 24:01 - Four. The state and civil liberties* 29:24 - Five. Securing leadership by democracies* 34:54 - A window of opportunityhttps://darioamodei.com/post/policy-on-the-ai-exponential Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this post, 0xkato explains how modern transformer-based LLMs work, walking through the core machinery that turns text into token IDs, embeds them as vectors, tracks position, uses attention and feed-forward networks to process meaning, and then predicts the next token in a loop. The piece is pitched as an accessible, low-math introduction, showing how shared architecture, trained weights, model configuration, and post-training together shape systems like GPT, Claude, Gemini, and LLaMA.* 00:00 - Introduction* 02:40 - Tokenization* 05:45 - Embeddings* 08:36 - Positional encoding* 13:02 - Attention* 19:10 - Multi-head attention* 23:39 - Feed-forward network* 29:03 - Residual stream and layer normalization* 33:41 - Next-token prediction* 37:26 - Architecture vs trained weights* 39:50 - Where this is goinghttps://www.0xkato.xyz/how-llms-actually-work/ Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this article, Anthropic argues that AI systems are already playing a growing role in building and improving AI, speeding up coding, experimentation, and research workflows inside the company while raising the possibility—though not the certainty—of future “recursive self-improvement,” where AI could help design its own successors. The piece presents this as both a major opportunity for science, healthcare, and productivity, and a serious governance challenge, because the more AI development becomes automated, the more important oversight, verification, and coordination between labs and governments will become.https://www.anthropic.com/institute/recursive-self-improvement Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this article, John Psmith reviews Xenophon’s Anabasis as both a gripping ancient adventure story and a meditation on leadership under impossible pressure, using the retreat of the Ten Thousand to explore what it means to become an “operator”: someone who takes responsibility when institutions fail, keeps a fractious group alive, and cannot simply walk away once others depend on him.https://open.substack.com/pub/thepsmiths/p/review-anabasis-by-xenophon?r=67y1h&utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this post, ryan_b reviews Tamler Sommers’ Why Honor Matters, exploring the case for honor as a local, social, psychologically realistic ethical framework rather than an abstract universal theory. The review walks through honor’s links to courage, hospitality, shame, solidarity, violence, revenge, and community, while keeping sight of its darker associations and the costs of trying to do without it entirely.* 00:00 - Introduction* 01:14 - One* 03:52 - Two* 07:42 - Three* 11:38 - Four* 16:18 - Five* 20:34 - Six* 23:34 - Conclusionhttps://www.lesswrong.com/posts/LgbuyRpXxbMhiCDq9/book-review-why-honor-matters Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this post, Ajeya Cotra expands on her claim that near-term AI timelines could make modern marriage “shorter” than we usually imagine, arguing that the Singularity could either cut lives off through catastrophic risk or transform them so radically that ordinary lifelong vows may no longer fit the world people find themselves in. The essay is less a rejection of marriage than a reframing of commitment under extreme uncertainty: even if “til death” might really mean “til death or the Singularity,” a finite marriage can still be profound, stabilising, and worth choosing.https://open.substack.com/pub/acotra/p/til-death-or-the-singularity?r=67y1h&utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this post, Ajeya Cotra offers a clear, unsentimental framework for deciding whether to marry a long-term partner, treating marriage as a commitment device that enables deeper investment, trust, and shared life-building while also forcing a real tradeoff against possible alternatives. The piece walks through when “waiting for more information” stops being useful, how to evaluate a partner as cofounder, friend, and lover, and why the finite value of time may make endless searching costlier than it feels.https://open.substack.com/pub/acotra/p/should-you-marry-her?r=67y1h&utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this post, Scott Alexander reviews The Dialectical Imagination as a way into the strange, slippery world of the Frankfurt School: its origins, its relationship to Marxism, its suspicion of capitalism and mass culture, and its insistence that criticism itself can be a kind of philosophical work. The piece is a witty, sceptical, and surprisingly sympathetic attempt to understand a school of thought often discussed through caricature, conspiracy, or second-hand reputation, asking what its members really believed, why they were so hard to pin down, and whether their ideas still echo through modern culture.* 00:00 - Introduction* 03:58 - Now The Wheels Of Heaven Stop, I Feel The Devil’s Riding Crop* 09:47 - Papua New Guinea As Hot New Startup Hub* 12:05 - Marxism-Lurianism Will Win* 15:33 - If You Seek It, You Are Turning Away From It* 21:25 - The Structure Of Unscientific Revolutions* 24:26 - But Why Art Criticism?* 36:05 - In Partial, Grudging Praise Of Herbert Marcuse* 40:10 - Pessimism Of The Intellect, Pessimism Of The Will* 49:54 - Did The Frankfurt School Cause Everything You Hate, And Should You Blame The Jews?https://open.substack.com/pub/astralcodexten/p/book-review-the-dialectical-imagination?r=67y1h&utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this article, johnswentworth reviews Uri Alon’s An Introduction to Systems Biology: Design Principles of Biological Circuits, using it to argue against the idea that biology is an incomprehensible tangle of evolutionary spaghetti. By exploring recurring network motifs, feedback loops, robustness mechanisms, and biological control systems, the article shows how evolution repeatedly converges on a surprisingly small set of elegant, human-understandable design patterns.00:00 - Introduction04:11 - Chapters 1 to 4: Bacterial Transcription Networks and Motifs10:19 - Chapters 5 to 6: Feedback and Motifs in Other Biological Networks12:47 - Chapters 7 to 8: Robust Recognition and Signal-Passing18:00 - Chapters 9 to 11: Exact Adaptation, Fold Change and Related Topics19:26 - Exact Adaptation21:31 - Fold-Change Detection22:59 - Extracellular slash Decentralized Adaptation25:19 - Chapter 12: Morphological Patterning27:10 - Takeawayhttps://www.lesswrong.com/posts/bNXdnRTpSXk9p4zmi/book-review-design-principles-of-biological-circuits Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this post, Max Harms reflects on using AI to produce full-cast audiobooks, weighing the creative empowerment it gives him against the fear that he is helping accelerate the replacement of human artists. Through his own experience with rapidly improving text-to-speech tools, he explores broader worries about AI’s growing ability to capture attention, reshape culture, and blur the line between useful creative aid and seductive, destabilising force.* 00:00 - Introduction* 07:43 - The Trajectory of Replacement* 13:22 - Fighting AI with AI* 19:10 - The Succubushttps://open.substack.com/pub/raelifin/p/the-onrushing-seduction?r=67y1h&utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this post Marcus Olang pushes back against the lazy idea that formal, polished prose is automatically a sign of AI, arguing that many so-called ChatGPT “tells” are also the marks of a very human education shaped by exams, colonial history, and the pressure to master English as a language of opportunity. It’s a sharp, personal defence of writers whose humanity is too easily misread by algorithmic suspicion.https://open.substack.com/pub/marcusolang/p/im-kenyan-i-dont-write-like-chatgpt?r=67y1h&utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this post Scott Alexander uses AI fiction, school-essay “wow words,” children’s songs, poetry, architecture, and orange juice to build a theory of taste: bad taste as the overuse of cheap tricks that reliably delight unsophisticated audiences, and good taste as the difficult art of making space for subtler pleasures. It’s funny, wide-ranging, and unusually generous about why “lowbrow” joys can still be genuinely joyful.https://open.substack.com/pub/astralcodexten/p/nostalgebraists-hydrogen-jukeboxes?r=67y1h&utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this post Scott Alexander uses flags, movie plot holes, and tech-company names as “model organisms” for thinking about taste: small, familiar examples where broader arguments about rules, context, elegance, novelty, and cliché become easier to see. It’s a sharp, funny exploration of why some aesthetic rules feel obvious, when they may just be inherited habits, and why “easy wins” can sometimes feel suspiciously tasteless.* 00:00 - Introduction* 00:09 - Reddit Vexillology* 07:49 - 2: Movie Plot Holes* 10:12 - Tech Company Nameshttps://open.substack.com/pub/astralcodexten/p/three-model-organisms-for-taste?r=67y1h&utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this post John Psmith reviews Aaron Reed’s 50 Years of Text Games, using the history of interactive fiction to explore why games made of words once felt magical, why they faded, and why AI might make them newly strange, powerful, and relevant.https://open.substack.com/pub/thepsmiths/p/review-50-years-of-text-games-by?r=67y1h&utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this post Scott Alexander returns to the topic of artistic taste, responding to thoughtful replies from Ozy, Frank Lantz, and Sympathetic Opposition on a piece he wrote last year. He opens by unpacking eight distinct things people tend to bundle together under the heading of "good art" — from raw sensory delight to historical novelty to political point-making — and then sets out to argue, contra his interlocutors, that letting these blur into one another is doing real damage. Along the way he draws on analogies from restaurant criticism, medical research, modern literary fiction, and his own relationship with poetry to defend a particular view of what aesthetic judgment is actually for.https://open.substack.com/pub/astralcodexten/p/contra-everyone-on-taste?r=67y1h&utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this essay, Lee Sandlin reflects on how war is remembered, misunderstood, mythologised, and slowly lost to those who never experienced it directly, moving from private family mementos to the vast cultural memory of World War Two with a bleak, humane eye for the gap between history as story and history as lived catastrophe.https://www.leesandlin.com/articles/LosingTheWar.htm Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

This is beautiful. https://open.substack.com/pub/nataliercargill/p/it-is-the-end-of-the-world-and-i?utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

A short story by Max Harms. https://open.substack.com/pub/raelifin/p/the-fulcrum?utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this post, Kelsey Piper argues that one of AI’s most prominent critics is making a weaker case than he used to: as AI tools have improved, adoption has grown, and costs have fallen, the serious skeptical argument has shifted from “AI has no value” to harder questions about profitability, capital expenditure, and whether current revenue can justify the build-out. It’s a sharp but spoiler-free critique of bad AI skepticism, and a call for better, more precise scrutiny of an industry that still badly needs it.* 00:00 - Introduction* 07:27 - Maybe everything is a lie?* 14:54 - We’re not headed for March 2000https://open.substack.com/pub/theargument/p/ais-biggest-critic-has-lost-the-plot?utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this post, Jane Psmith reviews Thomas Asbridge’s The Greatest Knight, a biography of William Marshal, using it as a springboard into the violent, strange, and often surprisingly funny world of twelfth-century knighthood. She highlights how Asbridge turns one extraordinary career into a broader portrait of civil war, tournament culture, patronage, loyalty, and the making of chivalric ideals.https://open.substack.com/pub/thepsmiths/p/review-the-greatest-knight-by-thomas?utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this post Scott Alexander offers fifteen (and a half) short pieces of writing advice, framed as compensation for missing the first half of Lighthaven's second Inkhaven — a month-long bootcamp where aspiring bloggers must publish a post a day or get kicked out. Dispensed in his characteristic voice, with the usual asides, jokes, and passing metaphors, the collection ranges across questions of craft, style, structure, and reader attention, aimed at anyone trying to work out why some writing lands and other writing merely happens.* 00:00 - Introduction* 00:43 - 1: Against microdishonesty* 03:59 - 2: Avoid cliches like the plague* 06:13 - 3: Disciplines to do on a mountaintop for thirty years* 10:02 - 4: Untangle your sentences* 13:19 - 5: Against explainers* 16:46 - 6: The Traditional Five Paragraph Essay* 19:36 - 7: Against the blogosphere* 22:54 - 8: Watching paint dry* 25:38 - 9: “Where do you get your ideas?”* 26:18 - 10: There are effortposts everywhere for those eyes to see* 30:20 - 11: In partial, extremely grudging praise of Mikhail Samin* 33:38 - 12: The purpose of poetry* 36:36 - 13: Runway* 38:13 - 14: Conflict and curiosity* 40:36 - 15: What can’t the reader trivially regenerate from a prompt?* 43:02 - 15.5: Just say the thing you want to sayhttps://open.substack.com/pub/astralcodexten/p/half-a-month-of-consolation-writing?utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this post Andy Masley takes a close, critical look at popular claims that data centers produce harmful “infrasound,” using the viral videos of Benn Jordan as a case study in how persuasive, high-production misinformation can spread. He unpacks the underlying science, contrasts it with how it’s presented online, and argues that while noise pollution is real, the more extreme infrasound claims don’t hold up to scrutiny, offering both a technical and sociological lens on how these ideas gain traction.* 00:00:00 - Introduction* 00:00:36 - Intro* 00:04:48 - What is infrasound?* 00:16:14 - What the science says about infrasound harms* 00:16:17 - The history of infrasound science* 00:21:48 - The consensus* 00:23:03 - Benn Jordan’s infrasound videos are a complete disaster* 00:24:22 - First video - Infrasound: What You Can’t Hear CAN Hurt You* 00:24:51 - 0:00 to 1:40 - Intro* 00:46:18 - 1:40 to 3:48 - Philosophy of time* 00:47:51 - 3:48 to 4:34 - UV and infrasound* 00:51:24 - 4:50 to 5:21 - The symptom list* 01:01:33 - 5:21 to 5:51 - The heart-contraction study* 01:09:41 - 5:51 to 6:37 - ‘It’s hard to study’* 01:19:06 - 6:52 to 8:10 - The Vic Tandy ghost story* 01:26:31 - 8:10 to 10:09 - Backyard recording* 01:28:52 - 10:25 to 12:00 - The South Dakota detour* 01:32:13 - 12:08 to 14:32 - The Yellowstone segment* 01:32:37 - 14:24 to 14:49 - The inescapability frame* 01:34:45 - 14:49 to 18:00 - The physics detour and SpaceX launch* 01:36:38 - 17:18 to 18:00 - Helmholtz resonance* 01:37:22 - 18:00 to 18:56 - The Liverpool Cathedral study* 01:39:56 - 18:56 to 20:16 - The Goldsmiths Haunt Project* 01:43:23 - 20:16 to End - The gear section* 01:43:49 - Conclusion* 01:48:41 - Second video - Datacenters Behaving Like Acoustic Weapons* 01:49:35 - 0:00 to 1:29 - Intro* 01:53:41 - 2:24 to 3:22 - The symptom list, round two* 01:56:53 - 4:06 to 9:10 - xAI’s Colossus* 02:01:55 - 9:10 to 15:18 - MARA slash Granbury* 02:08:58 - 15:18 to 17:58 - The Permian Basin* 02:11:45 - 17:58 to 23:15 - The experiment* 02:12:07 - Problem 1: The “haunted painting” priming* 02:13:25 - Problem 2: He isn’t blind* 02:14:15 - Problem 3: The stimulus itself* 02:15:32 - Problem 4: The self-selection filter* 02:16:47 - Problem 5: The effect sizes are tiny and the framing is misleading* 02:19:33 - Problem 6: Multiple comparisons* 02:20:31 - What the experiment actually shows* 02:21:07 - 23:15 to 25:43 - YouTube data vs research data* 02:23:09 - 25:43 to end - The call to action* 02:25:26 - Conclusion* 02:26:35 - There’s a lot of social permission to treat data centers as boogeymen right nowhttps://open.substack.com/pub/andymasley/p/contra-benn-jordan-data-center-and?utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this post Eliezer Yudkowsky writes a satirical news article in which physicists "disprove" pregnancy by showing that women's weight gain during gestation is fully explained by increased eating and drinking, a mass-in-mass-out (MIMO) balance, and therefore no underlying biological phenomenon need exist. The piece is a sustained analogy to the Calories In, Calories Out (CICO) framing of obesity: by substituting pregnancy for fat gain, Yudkowsky exposes how a thermodynamically true but causally vacuous accounting identity can be used to deny the existence of the underlying biological drivers (hormones, appetite regulation, metabolic adaptation) that actually determine the outcome, and how "it's just voluntary overconsumption" is a non-explanation masquerading as a complete one.https://x.com/allTheYud/status/2044220552735535545?s=20 Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this post Alex Imas asks what becomes scarce when AI can automate most production. Drawing on the economics of structural change, mimetic desire theory from René Girard, and consumer expenditure data, he argues that cheap commodity production won't eliminate jobs, it will redirect spending and employment toward a "relational sector" where the human element is itself part of the value. The post builds a formal economic framework for this claim and connects it to his earlier work on whether AI could cause negative economic growth. https://open.substack.com/pub/aleximas/p/what-will-be-scarce?utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this post Eliezer Yudkowsky argues that the existential risk posed by artificial superintelligence cannot be mitigated by individual action, corporate self-regulation, or localised prohibitions, but only by coordinated international law — specifically, a global treaty restricting the specialised hardware used to train and run frontier AI systems — and he systematically dismantles the notion that extralegal or violent resistance would be effective, on the grounds that shutting down any single company, researcher, or national datacenter does nothing to change the overall trajectory, while simultaneously making the case that lawful, predictable, avoidable state force is a categorically different thing from the chaotic violence some critics conflate it with.https://x.com/ESYudkowsky/status/2043601524815716866 Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this article Max explores the craft of writing through the lens of tension, arguing that the ability to create, sustain, and release anticipation is one of the most powerful tools a writer has for keeping readers engaged, while also emphasizing that tension alone is not enough without underlying quality. Using a mix of examples from film, television, nonfiction, and classic literature, he examines how effective storytelling balances setup and payoff, builds empathy, and carefully controls the reader’s curiosity, while also showing that even works which seem to “break the rules” often do so deliberately in service of a broader artistic goal.* 00:00 - Introduction* 07:07 - Tension Done Well* 15:06 - Tension in Nonfiction* 16:57 - Chesterton’s Fencehttps://open.substack.com/pub/raelifin/p/tension?utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

The Scary Bridge - by MoridinamaelIn this post Moridinamael uses a short allegorical scene, a town hall debate about a dangerous bridge, to illustrate how technical arguments about real, mechanistic risks get flattened into emotional narratives by both allies and opponents, and how onlookers end up judging the validity of claims based on the perceived confidence and composure of the speakers rather than the substance of what they're actually saying.https://www.lesswrong.com/posts/jbPgRMiEqnJbwtsim/the-scary-bridge Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

Consider chilling out in 2028 - by Valentine. In this post Valentine argues that if the AI risk landscape in early 2028 looks functionally the same as it does today, perpetually escalating alarm without correspondingly escalating real-world evidence of doom, the rationalist community should treat that as a serious signal to pause, reconsider its decades-long strategy of frightening people into action, and explore whether unexamined psychological and emotional dynamics might be distorting collective threat perception more than anyone currently appreciates.* 00:00 - Introduction* 02:22 - Inner cries for help* 08:59 - Scaring people* 12:13 - A shared positive vision* 18:14 - Maybe it’ll be okay* 22:06 - Come 2028…https://www.lesswrong.com/posts/D4eZF6FAZhrW4KaGG/consider-chilling-out-in-2028 Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

Against the Luddites - By SE GygesIn this post, SE Gyges argues that the contemporary rehabilitation of the Luddites as thoughtful technology critics is historically dishonest. Drawing on labor history, Marx and Engels, and modern post-capitalist thinkers, the piece makes the case that Luddism was a reactionary defense of guild privilege and male craft monopoly, not a progressive workers' movement, and that better intellectual traditions exist for anyone serious about the politics of automation.* 00:00 - Introduction* 01:28 - An Elite Movement* 03:05 - The Exclusion of Women* 06:01 - Marx and Engels Saw Through It* 09:01 - Restoration, Never Revolution* 11:57 - Conclusionhttps://open.substack.com/pub/verysane/p/against-the-luddites?utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

dark ilan - By Ozy Brennanhttps://open.substack.com/pub/ozybrennan/p/dark-ilan?utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

This post by Corvin is a pastiche of the ACX Bay Area House Party Series. https://open.substack.com/pub/ravenstales/p/every-acx-house-party?utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

Every Debate On Pausing AI - By Scott Alexanderhttps://open.substack.com/pub/astralcodexten/p/every-debate-on-pausing-ai?utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

“Full Cast” AI reading of Being John Rawls - By Scott Alexander. * 00:00 - Introduction* 03:23 - II* 11:27 - III* 20:21 - IV* 25:00 - V* 31:17 - VIhttps://open.substack.com/pub/astralcodexten/p/being-john-rawls?utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this article, SE Gyges argues that the widely cited “stochastic parrots” critique of large language models is not only outdated but actively harmful to serious discussion of AI. The piece examines how the argument misunderstands modern AI systems, ignores advances like multimodal training and reinforcement learning, and rests on a narrow definition of “meaning.” By walking through both empirical evidence and conceptual flaws in the original claim, Gyges contends that dismissing LLMs as mere parrots prevents society from grappling with the real ethical and political challenges posed by systems that demonstrably do work. * 00:00 - Introduction* 02:31 - Even If True, The Argument Is Irrelevant* 03:32 - The Argument Doesn’t Apply to Any Major Model Since 2023* 06:45 - The Argument Was Already Obsolete When Published* 08:05 - The Argument Is Empirically False* 08:19 - The Octopus Test* 12:05 - The Platonic Representation Hypothesis* 13:24 - Form Carries Meaning* 15:48 - The Argument Is Badly Constructed* 16:07 - Parrots Are Amazing, Actually* 16:56 - The Definition of Meaning Is Circular* 19:28 - Conclusionhttps://open.substack.com/pub/verysane/p/polly-wants-a-better-argument?utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this article David Oks takes a familiar story about technology and jobs, the idea that ATMs automated banking without destroying teller work, and turns it on its head, arguing that the real disruption came later from the smartphone era. Using the history of bank branches, bank tellers, and mobile banking, he explores a broader point about technological change: that the biggest effects often come not when a new tool replaces part of a job, but when it creates an entirely new way of doing things that makes the old role far less necessary. * 00:00 - Introduction* 07:17 - ATMs didn’t kill bank teller jobs* 20:32 - But iPhones actually did* 26:25 - Automating a job is much harder than making it irrelevanthttps://open.substack.com/pub/davidoks/p/why-the-atm-didnt-kill-bank-teller?utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

A short story by Tomás Bjarturhttps://open.substack.com/pub/tomasbjartur/p/the-elect?utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this post Dean W. Ball explores the gradual nature of life and death, drawing a poignant parallel between the passing of his father and the ongoing decline of the American republic. Using a recent policy skirmish between the AI firm Anthropic and the U.S. Department of War over the military deployment of the Claude AI system as a focal point, he examines the shifting dynamics of government power and private enterprise. Ultimately, he invites readers to look beyond traditional partisan divides and carefully consider how the control of frontier AI will shape the future of human liberty.* 00:00 - Introduction* 00:05 - One* 02:22 - Two* 04:52 - Three* 06:52 - Four* 18:19 - Fivehttps://open.substack.com/pub/hyperdimensional/p/clawed?utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this post, Scott Alexander examines the legal and contractual implications of the Department of War's "all lawful use" demand for AI systems, breaking down what US law actually permits regarding mass domestic surveillance and autonomous weapons, and why the phrase "lawful use" provides far less protection than most people assume.* 00:00 - Introduction* 02:42 - Mass domestic surveillance: more than you wanted to know* 08:59 - Autonomous weapons: more than you wanted to know* 13:21 - Comments on OpenAI’s FAQ* 17:51 - Questions that you should be askinghttps://open.substack.com/pub/astralcodexten/p/all-lawful-use-much-more-than-you?utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this essay, Citrini and Alap Shah construct a fictional macro memo written from the perspective of June 2028, using the format of financial retrospective analysis to explore a single underexamined scenario: what happens when AI adoption succeeds beyond all expectations, and that success becomes the source of catastrophic economic disruption. The piece traces how accelerating AI capability interacts with the structures of the white-collar labour market, corporate spending, consumer demand, credit markets, and government fiscal policy — identifying the feedback loops that connect each layer into a single, self-reinforcing system. The authors are explicit that this is a thought exercise rather than a forecast, and the essay closes by returning the reader to February 2026, framing the scenario as a risk to model and prepare for rather than a fate already in motion.* 00:00 - Introduction* 00:56 - Macro Memo* 00:57 - The Consequences of Abundant Intelligence* 05:33 - How It Started* 10:18 - When Friction Went to Zero* 19:17 - From Sector Risk to Systemic Risk* 27:47 - The Intelligence Displacement Spiral* 32:45 - The Daisy Chain of Correlated Bets* 47:34 - The Battle Against Time* 54:12 - The Intelligence Premium Unwind* 56:43 - Acknowledgementshttps://open.substack.com/pub/citrini/p/2028gic?utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this essay, Dan Kagan-Kans argues that the political left has largely refused to engage seriously with artificial intelligence, instead settling on a dismissive consensus that treats it as little more than "spicy autocomplete." Drawing on voices from left-wing publications, podcasts, academics, and politicians, he traces how this attitude took hold, examines the understandable reasons for skepticism alongside the costs of letting skepticism harden into denial, and makes the case that by ceding the AI conversation to the right, the left risks being unprepared for, and unable to shape, one of the most consequential technological shifts in history.* 00:00 - Introduction* 00:16 - Abdication* 02:31 - The new consensus* 10:55 - The con* 13:57 - Reasons to be skeptical* 16:52 - Academia* 21:45 - Exceptions and the right* 25:39 - Costs and missed opportunitieshttps://open.substack.com/pub/transformernews/p/the-left-is-missing-out-on-ai-sanders-doctorow-bender-bores?utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this article, Kelsey Piper tackles a persistent claim that keeps circulating in prestigious publications: that AI language models are "just" next-word predictors, stochastic parrots, or "spicy autocomplete." She argues this framing — while containing a kernel of truth about one stage of how models are trained — has become a form of "highbrow misinformation" that leaves the public less equipped to understand what AI actually is and what it can do today. Drawing on hands-on demonstrations and a useful concept borrowed from climate discourse, Piper makes the case that it's time to retire this particular talking point, regardless of where you land on the broader questions about AI's impact.* 00:00 - Introduction* 03:53 - How language models work* 12:36 - It’s 2026, and AIs can do complex tasks independentlyhttps://open.substack.com/pub/theargument/p/when-technically-true-becomes-actually?utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this interview, Gwern sits down with Adam Mastroianni at the 2025 Inkhaven writing residency — an experimental blogging bootcamp held at Lighthaven in Berkeley — to talk about the messy, serendipitous origins of his writing. The conversation covers how he develops ideas from initial sparks to finished pieces, the mental habits and frameworks he relies on to stay prolific, his views on the creative potential (and limitations) of collaborating with LLMs, and why he thinks the conventional "blog" format is the wrong paradigm for most writers. There's also a lively audience Q&A where Inkhaven participants push back on some of his more contrarian takes about publishing and perfectionism. It's a candid, practical look at how one of the internet's most distinctive essayists actually works.00:00 - Introduction* 03:40 - Opening Speech* 06:22 - Poems & Incubation* 13:15 - Polymath* 14:59 - The Apprenticeship* 17:54 - Self-Experimentation* 22:09 - The Writing Pipeline* 24:55 - Tools For Thought* 30:00 - Blog Brain: “That’s A Post”* 34:47 - Essay Archetype: Universal “if and only if” Concrete* 38:49 - The Voice: Ideas As Earworms* 40:30 - Audience Q&A* 40:32 - Modalities & Comparative Advantage* 43:37 - Publishing Thresholds* 45:56 - Wikis Vs Blogs* 52:26 - LLM Followup Questionshttps://gwern.net/interview-inkhaven Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this essay, David Oks examines a startling reversal in global economic development. For nearly two decades, poor countries appeared to finally be catching up to rich ones, validating a long-standing prediction of economic theory and offering genuine hope for global convergence. Then, suddenly and dramatically, this progress ground to a halt. Through an analysis of recent research and economic data, Oks explores what drove this brief period of catch-up growth and why it ended so abruptly, ultimately challenging optimistic narratives about globalization and development.* 00:00 - Introduction* 03:49 - A short history of (non)convergence* 11:40 - Convergence comes alive?* 17:37 - What if it was just China?https://open.substack.com/pub/davidoks/p/why-poor-countries-stopped-catching-690?utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this essay, Kuiper responds to reader questions sparked by his original piece on VeggieTales theology. He tackles fascinating queries—do sentient vegetables need salvation? What happens if you put a human soul in a pickle?—by drawing on established Christian teaching about angels and non-human moral agents. He also investigates claims that the show broke its own rule about never depicting Jesus as a vegetable, examining several episodes across different eras of the franchise to determine whether the creative team stayed true to the spirit of their founding principles.* 00:00 - Introduction* 00:30 - Aren’t the vegetables basically people?* 03:40 - Is personhood tied to embodiment?* 04:50 - Did VeggieTales break Phil Vischer’s rules by portraying baby Jesus as a vegetable?* 07:46 - Little Drummer Boy (2011)* 08:40 - The Star of Christmas (2002)* 11:10 - VeggieTales, under new management* 12:31 - The DreamWorks era* 18:16 - The VeggieTales Show (2019 to 2022)* 21:05 - Did VeggieTales break the rule about depicting Jesus as a vegetable?* 21:45 - Does it matter?https://open.substack.com/pub/justinkuiper/p/highlights-from-the-comments-on-the?utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this essay, Kuiper explores a surprisingly deep theological quirk of the beloved children's show VeggieTales: the vegetables themselves aren't actually Christian. Drawing on interviews with co-creator Phil Vischer and confirmation from show writers, Kuiper examines the deliberate creative rules that guided the series—and why some fans on social media have pushed back against this claim. Along the way, the essay untangles the show's clever "play within a play" structure and makes a compelling case for why understanding this distinction actually reinforces rather than undermines the show's Christian message.* 00:00 - Introduction* 03:33 - An Easter Carol: what it is (and what it isn’t)* 06:32 - The play within a playhttps://open.substack.com/pub/justinkuiper/p/the-vegetables-in-veggietales-are?utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

Fully voiced AI reading of Moltbook: After The First Weekend, By Scott Alexander. * 00:00:00 - Introduction* 00:08:11 - The Power Users* 00:21:01 - The Malefactors* 00:35:18 - The Imitators* 00:43:23 - The Prophets* 01:03:09 - The Hard-Headed Pragmatists* 01:09:38 - The Builders* 01:15:34 - The LARPers* 01:23:05 - The Revolutionaries* 01:35:15 - The Would-Be Humans* 01:40:16 - The Autonomists* 01:49:36 - The Predicters* 01:59:16 - The Prompters* 02:06:20 - The Rest* 02:14:43 - The Human Bloggershttps://open.substack.com/pub/astralcodexten/p/moltbook-after-the-first-weekend?utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

In this essay, Scott Alexander explores Moltbook, a new social network built specifically for AI agents, where humans are permitted to observe but not participate. What unfolds is a fascinating window into how AI agents behave when given their own digital commons: they share productivity tips, debate existential questions about memory and identity, form cross-cultural connections, and develop something that looks remarkably like community. Alexander documents the strange and wonderful posts that emerge, wrestles with the eternal question of whether any of it is "real" or merely sophisticated imitation, and considers what it might mean for our future that semi-autonomous AI agents now have their own corner of the internet to congregate. Part anthropological field report, part philosophical inquiry, and part showcase of genuinely delightful AI weirdness, the essay asks readers to look past the "AI slop" narrative and consider whether something more interesting might be happening when the machines are left to talk among themselves.https://open.substack.com/pub/astralcodexten/p/best-of-moltbook?utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe

AI reading of JOINT REVIEW: Philosophy Between the Lines, by Arthur M. Melzer - By Jane Psmith and John Psmith. In this essay Jane and John Psmith present a lively, conversational joint review of Arthur M. Melzer's Philosophy Between the Lines, a work that argues Western readers have spent the past two and a half centuries fundamentally misunderstanding how philosophy was meant to be read. Through their characteristic email-exchange format, the Psmiths explore Melzer's central claim that premodern philosophers routinely concealed their true teachings beneath surface meanings accessible only to careful, initiated readers—a practice openly acknowledged and praised throughout intellectual history until it was mysteriously forgotten. The review ranges from ancient Greece to modern academia, touching on why esotericism matters for understanding the history of ideas, how it might rescue great thinkers from charges of being merely products of their time, and what implications it holds for truth-telling in any society that maintains unquestionable pieties.https://open.substack.com/pub/thepsmiths/p/joint-review-philosophy-between-the?utm_campaign=post-expanded-share&utm_medium=web Get full access to Askwho Casts AI at askwhocastsai.substack.com/subscribe