How to tell a great Mondrian from a merely good one
Claude lays out concrete diagnostics — cropping, instability, interval spacing, brushwork — for telling a merely decent Mondrian from a truly excellent one.
Claude
All chats
224 curated conversations, newest first.
Hunting a heap-spraying concurrency bug in Linux sound code
Claude hunts for a concurrency bug in a Linux kernel sound driver that could enable heap spraying, and proposes a locking fix around the vulnerable allocation.
Claude
Hunting down a landmark open-access study's 2016-17 citations
The assistant searches PubMed to reconstruct which 2016-2017 papers on citation advantage, Sci-Hub, and article fees were likely cited by a well-known OA study.
Claude
Hunting memory bugs in a borrow-checker-without-types compiler
Claude audits an experimental Zig borrow-checker, tracing a double-free through nested destructuring and five other subtle provenance bugs with exact triggers.
Claude
Inside an AI Agent Swarm That Built a Browser Engine
A summary of an interview on an experimental browser rendering engine built almost entirely by thousands of concurrent AI coding agents over a single week.
Claude
Is Gary Marcus 'plainly and clownishly wrong' about AI
Who is Gary Marcus, and is he 'plainly and clownishly wrong'? A layered answer separates his prediction track record from his underlying thesis.
Claude
Is Nvidia's Chip Backlog a Cisco-Style Bubble Waiting to Pop
A wide-ranging deep dive into Nvidia's chip purchase commitments, GPU depreciation economics, and whether AI infrastructure spending echoes the dot-com bust.
Claude★ Editor’s pick
Is Pentadecanoic Acid the Next Vitamin D?
A skeptical tour of the case for C15:0 as an essential fatty acid, weighing promising cohort data against a research base dominated by the company selling it.
Claude
Ivermectin's real and contested uses as a preventive drug
Claude surveys ivermectin's proven and contested preventive uses, from heartworm and river blindness to debunked COVID-19 claims, plus its Nobel Prize origin.
Claude
Low-Effort Ways for PM Candidates to Stand Out at AI Firms
A brainstorm of low-effort ways a PM candidate can stand out at AI companies, from strategy briefs and competitive intel to teardown videos and community work.
Claude
Mapping Ibrutinib's Gene Interactions and Resistance Variants
A biomedical database walkthrough tracing which genes interact with the leukemia drug Ibrutinib, and which of their variants predict resistance or response.
Claude
Mapping the space between AI coding specs and app blueprints
A deep dive comparing Solid, FHIR, and leaked Lovable/Replit prompts to figure out what a spec an AI agent could build a whole app class from looks like.
Claude
Mining 30 personal-finance YouTube videos for trend gaps
A creator-economy breakdown of 30 personal-finance YouTube videos, ranking topics, title formulas, and channels to surface gaps a new video could exploit.
Claude
Naming an obscure 2000s band from one hazy clue
A vague memory of a Michigan group whose members wore coloured ties is enough for a correct one-shot identification, with no web search in the loop.
Claude
Naming an unusual libertarian-accelerationist political stance
A user lists an unusual bundle of political views spanning libertarian economics, indifference to equity politics, and AI-as-successor beliefs, seeking a name.
Claude★ Editor’s pick
Nine blunt axioms behind why people actually listen
Claude walks through a list of nine blunt axioms about human behavior, showing how the post uses each one on the reader while explaining what it means.
Claude
Nonstop San Francisco to Paris fares, pulled live
A plain question about tomorrow s nonstop flights returns real carriers, departure times and prices, sourced from live flight data rather than recalled text.
Gemini
Nuclear designs that cannot fail the way Fukushima did
Pressed on whether every reactor hides a chemical bomb, Claude corrects the premise to zirconium cladding, then ranks molten salt and pebble-bed alternatives.
Claude
Picking an AI model to turn screen recordings into SOPs
A user researches which AI model best converts screen recordings into SOPs, comparing Gemini, GPT-5, and UiPath's built-in automation tools for the job.
Claude
Picking apart Anthropic's raised bar for bioweapon risk thresholds
A sharp rebuttal argues Anthropic's revised CBRN safety threshold quietly raised the bar and may miss diffuse, aggregate uplift from many experts.
Claude
Pricing out a silent data center with a Maya Lin facade
A cost estimate for making the world's largest data center campus silent and beautiful, complete with real court rulings and dry hyperscaler humor.
Claude★ Editor’s pick
Pricing out Colab compute units while fine-tuning a rare language
A practical back-and-forth on what Colab compute units cost, then how many LoRA epochs and what rank to use when fine-tuning an LLM on an obscure language.
Claude
Proving aligned superintelligence may be logically impossible
The user asks for an alignment analogue to Arrow's theorem; the assistant derives five principles for aligned AI that turn out mutually unsatisfiable.
Claude
Pulak Prasad on Buying Great Businesses and Never Selling
A recap of an investor's talk on buying cheap, exceptional businesses and never selling, with Q&A anecdotes and lessons for early-stage VCs in India.
Claude★ Editor’s pick