Coby Kassner

on the atlas — 68

• Neural Networks Are Already Clean39 savers
• By default, Chinese AI falls very behind — LessWrong3 savers
• Becoming a magician – Autotranslucence43 savers
• Three AI Safety Related Ideas — LessWrong1 savers
• Two Neglected Problems in Human-AI Safety — LessWrong1 savers
• Plan R: AI Safety by ASICs — LessWrong1 savers
• We need a better theory of polarization, because it's failing to predict the AI debate — LessWrong1 savers
• How to shape policy from outside government3 savers
• An opinionated primer on Causal Decision Theory (CDT) and Evidential Decision Theory (EDT) — LessWrong1 savers
• Swarm Scaling — LessWrong2 savers
• Nate Soares' Life Advice - LessWrong5 savers
• How to have executive function – scribbles in the margins7 savers
• on doing Real work - by jessica dai7 savers
• [2607.24653] Kimi K3: Open Frontier Intelligence1 savers
• Rest in motion24 savers
• Data Bottlenecks Won't Stop an Intelligence Explosion2 savers
• Dario Amodei — We Must Pace the Frontier36 savers
• Defining alignment research — LessWrong2 savers
• The ML ontology and the alignment ontology — LessWrong1 savers
• Post 47: How I Formed My Own Views About AI Safety — Neel Nanda4 savers
• Dr. R. W. Hamming's Advice on Research14 savers
• How To Be Successful96 savers
• Research acceleration: The view inside OpenAI | OpenAI14 savers
• Superintelligent surveillance to prevent galactic anarchy3 savers
• Risk-Averse AIs4 savers
• Salad days — LessWrong5 savers
• The Dynamics of Intelligence Explosions — LessWrong1 savers
• Should We Lock in Post-AGI Agreements Under Uncertainty?1 savers
• Obvious advice10 savers
• What is the likelihood that civilizational collapse would directly lead to human extinction (within decades)? — EA Forum1 savers
• The Four Pillars: A Hypothesis for Countering Catastrophic Biological Risk3 savers
• How I think about catastrophic biological risk (part I): risk breakdown by type of response3 savers
• 21 Facts About Throwing Good Parties12 savers
• Letter Four – The Rainer Maria Rilke Archive1 savers
• Productivity: Working towards a summary of what we know3 savers
• Will reward-seekers respond to distant incentives? — LessWrong1 savers
• Privacy-Preserving AI Audit Tools — OpenMined2 savers
• Notes on "Managing to Change the World" — EA Forum3 savers
• Fitness-Seekers: Generalizing the Reward-Seeking Threat Model — LessWrong2 savers
• Many arguments for AI x-risk are wrong — AI Alignment Forum2 savers
• Reward is not the optimization target — LessWrong9 savers
• Some Ways I Think About Evaluating Grant Applications — LessWrong1 savers
• Strong Evidence is Common — LessWrong1 savers
• Be less scared of overconfidence | benkuhn.net14 savers
• My weekly review habit | benkuhn.net17 savers
• Impact, agency, and taste | benkuhn.net31 savers
• Do Things that Don't Scale13 savers
• How I select alignment research projects — LessWrong1 savers
• Learning By Writing43 savers
• The Moat of Low Status - by Sasha Chapin5 savers
• How to be More Agentic - by Cate Hall - Useful Fictions24 savers
• Be impatient | benkuhn.net24 savers
• Don't Build Mindreading — LessWrong7 savers
• Why do models task game? — LessWrong2 savers
• models may behave differently in graded episodes (a tirade) — LessWrong8 savers
• philpapers.org/archive/GOLLFC-2.pdf2 savers
• Stickiness in AI Behavioral Design1 savers
• What Should Go In A Model Spec?1 savers
• AI-Enabled Coups: How a Small Group Could Use AI to Seize Power11 savers
• What makes a strategy great2 savers
• Strategy as Problem-Solving > US Army War College - Publications > Display2 savers
• The impact merge - Joe Carlsmith2 savers
• A playbook for field strategy - by Dewi Erwan5 savers
• Good Strategy/Bad Strategy | Summary & Notes - Will Patrick1 savers
• The AI Industrial Explosion — Part 1: Maximum growth rates with current production methods4 savers
• Three Types of Intelligence Explosion8 savers
• Persistent Path-Dependence: Why Our Actions Matter Long-Term4 savers
• Curius / Bookmarks for the extremely curious184 savers