全部/知识/短内容

LessWrong · 短内容

4分钟前更新

HISTORY近 30 天历史柱高表示当天去重热搜数量

7517 8118 6419 6520 6221 8222 6523 7124 6725 6326 5927 7428 8029 6930 7301 6502 7103 6104 5905 7006 6207 6908 7909 6710 6211 10912 6913 7014 6615 6216

09/17—10/16 有历史数据

0114
Bad Guy Tokens
评论：0 · 标签：AIaxel_sdq
0215
Mental Health Navigator site for EA and AI safety
评论：0juliawise
0322
Staying sane while trying to save the world: Highlights from Spencer Greenberg and Luisa Rodriguez
评论：1 · 标签：Practicaljuliawise
0429
AI #190: Scion of the Times
评论：1 · 标签：AIZvi
0522
Headwinds Everywhere (in the Chinese AI Safety Field)
评论：1 · 标签：AI Governance、China、AICMLKevin
064
Which AI safety organization does what? Puzzle/party games
评论：0 · 标签：AIdavid reinstein
075
Anthropic's unreleased model submits visa applications
评论：0 · 标签：Anthropic (org)、News、AIbecausecurious
08187
Linkpost: Neural Networks Are Already Clean
评论：21 · 标签：Interpretability (ML & AI)、AIYair Halberstadt
093
Children of the Cloud: Noah’s Ark as an Alignment Song
评论：0pennyfan87
1059
Preface to the Computational Agent Theory and ASI Alignment Sequence
评论：0 · 标签：Agent Foundations、Research Agendas、AIVanessa Kosoy
117
Predictor and picker – convergent architectural features of intelligent agents
评论：0 · 标签：Agent Foundations、Instrumental convergence、AIJáchym Fibír
124
ML4Good Impact Assessment
评论：0 · 标签：AIJack_S
139
Analysis of a tossed Coin: An Introduction to Phase Space
评论：0 · 标签：World ModelingSaltAndMetal
1417
Robustness: Backfire Risks in AI Safety and Governance
评论：0 · 标签：AIDawn Drescher
1599
A short review of "The Zizians: Inside a Trans Death Cult"
评论：10 · 标签：Dramajessicata
1645
The basic argument for making AI better at moral philosophy
评论：1 · 标签：AICaspar Oesterheld crossposts
1710
AI Marketplaces Can Contribute to, and Profit From, the AI 2040 Plan: Research Directions and Strategy
评论：0 · 标签：AIMatthew Khoriaty
182
Rogue AI developers as hostis humani generis: under an international AI treaty, should they be subject to international concurrent jurisdiction?
评论：2 · 标签：AIConnor Williams
1933
The user in the LLM might be a moral patient too (a strange implication of the PSM and simulators)
评论：7 · 标签：Interpretability (ML & AI)、AI、World Modelingagastyasridharan
205
Why are some explanations bad?
评论：1 · 标签：Rationality、World ModelingElijah
2118
someone's agents are still "hacking"
评论：1jonathanbreitg
224
When bureaucracy gives you something it shouldn't
评论：1 · 标签：Government、Law and Legal systems、World Modelingbecausecurious
2311
You Should Specify Your Code Unambiguously (and Prove That You Did)
评论：0Simon Henniger
2446
The Safety Culture at the Labs is Broken
评论：2 · 标签：AIZvi
2581
Some interesting internal documents from Anthropic litigation
评论：10 · 标签：AIloops
269
Agent memory the human can still audit: primary sources, epistemic status markers, and user sign-off
评论：0 · 标签：Epistemic Hygiene、Language Models (LLMs)、Productivity、AIFlorian_Dietz
276
Should We Edit the Human Genome?
评论：0 · 标签：World OptimizationWalter Veit
284
(Why) we’re launching the next cohort of Generator
评论：0 · 标签：AIAris
29122
Your prompt… plus everything you've ever (publicly) written
评论：13 · 标签：AIkromem
3082
A Chinese Perspective on “On Social Reality in China”
评论：2 · 标签：China、World ModelingDusty
3139
(Why) we’re launching the next cohort of Generator
评论：0 · 标签：AI Alignment Fieldbuilding、AIAris
3225
Hindsight: a trivia + calibration game
评论：0 · 标签：Games (posts describing)Arjun Panickssery
337
South Korea, ACX Meetups Everywhere Fall 2026
评论：0Dalcy
341
Inventing a New Poetry Format: Superposition Poetry
评论：0 · 标签：Art、Bayes' Theorem、Poetry、Probability & Statisticsarkendor
3515
Roko Mijic talks AI Risk on The Peter McCormack Show
评论：0 · 标签：AIRoko
3616
OpenAI finds recording device in one of their offices
评论：1 · 标签：AI Safety、AIdefun
378
Every Bay Area House Party house party in London 2
评论：0Mikhail Samin
3815
Today I Killed; Tomorrow I Will Kill Again
评论：2Lance Spencer
393
AI Horizons Forum (Dec 12–13, SF)
评论：0 · 标签：AIOllieRodriguez
4010
Below the Frontier: The Battle for Good-Enough AI
评论：1 · 标签：AIykevinzhang
4173
Arguments About AI Consciousness Remain Highly Motivated and At Best Overconfident
评论：3 · 标签：AIZvi
427
A Spine within the Swarm
评论：0 · 标签：AI Safety、Aligned AI Proposals、AINolan Coffey
4324
Agreement not to defect (alone) can mitigate loss of control risks
评论：2 · 标签：AIDavidmanheim
4415
LLMs Knowingly Hide Their Mistakes
评论：1 · 标签：AILucas Florin
45133
Personas in a Pre-Internet Language Model
评论：6 · 标签：AIDavid Africa
468
Launching the Hong Kong AI Safety Hub (HASH), a new home for AI safety work in Hong Kong
评论：0 · 标签：AISchizoid Rentoid
4715
Book Review: The Receipt Horizon by Joel Borgen
评论：2 · 标签：Book Reviews / Media Reviews、AISolenoid_Entity
4815
Chess without opening theory
评论：2 · 标签：World ModelingABishop
497
Too Many Yellow Flags
评论：6Dr. Diction
5047
OpenAI Fires Three Safety Researchers
评论：0 · 标签：AIZvi