LessWrong · 历史高分

2分钟前更新

HISTORY近 30 天历史柱高表示当天去重热搜数量

5017 5118 5019 5020 5021 5022 5023 5024 5025 5026 5127 5028 5029 5030 5001 5002 5003 5004 5005 5006 5007 5008 5009 5010 5011 5012 5013 5014 5015 5016

09/17—10/16 有历史数据

011204
Why I Left Google DeepMind
评论：60 · 标签：DeepMind、Autonomous Weapons、Frontier AI Companies、Government、Mechanism Design、Something To ProtectTurnTrout
021053
AGI Ruin: A List of Lethalities
评论：716 · 标签：AI Risk、Threat Models (AI)、AI Questions Open Threads、Double-Crux、Fuzzies、Language Models (LLMs)Eliezer Yudkowsky
03964
Preface
评论：20 · 标签：Postmortems & Retrospectives、Ethics & Morality、Life Improvements、Rationality A-Z (discussion & meta)、RationalityEliezer Yudkowsky
04963
Where I agree and disagree with Eliezer
评论：224 · 标签：AI Risk、AI、World Modelingpaulfchristiano
05942
Eight Short Studies On Excuses
评论：254 · 标签：Game Theory、Group Rationality、Life Improvements、Rationality、PracticalScott Alexander
06853
The Company Man
评论：79 · 标签：FictionTomás B.
07821
The Best Textbooks on Every Subject
评论：418 · 标签：Scholarship & Learning、List of Links、Book Reviews / Media Reviews、Repository、Akrasia、Algorithmslukeprog
08817
How My Students Think About AI
评论：117 · 标签：AIdvd
09800
My journey to the microwave alternate timeline
评论：58 · 标签：AI、World ModelingMalmesbury
10775
The Rise of Parasitic AI
评论：191 · 标签：LLM-Induced Psychosis、LLM Personas、Human-AI Safety、AI Rights / Welfare、AI Questions Open Threads、AIAdele Lopez
11771
Current AIs seem pretty misaligned to me
评论：83 · 标签：AI Risk、RLHF、Welcome Threads、AIryan_greenblatt
12751
How Go Players Disempower Themselves to AI
评论：79 · 标签：Gaming (videogames/tabletop)、MATS Program、AIAshe Vazquez Nuñez
13750
Only Law Can Prevent Extinction
评论：110 · 标签：AI、World OptimizationEliezer Yudkowsky
14727
Simulators
评论：170 · 标签：Simulator Theory、Language Models (LLMs)、LLM Personas、GPT、Outer Alignment、Simulationjanus
15709
Humans are not automatically strategic
评论：279 · 标签：Introspection、General intelligence、Goodhart's Law、Motivational Intro Posts、RationalityAnnaSalamon
16704
AI 2027: What Superintelligence Looks Like
评论：223 · 标签：AI Timelines、Forecasts (Specific Predictions)、Superintelligence、Orthogonality Thesis、Reversed Stupidity Is Not Intelligence、AIDaniel Kokotajlo
17701
What an actually pessimistic containment strategy looks like
评论：138 · 标签：Slowing Down AI、AI Governance、AI、World OptimizationDean Valentine
18696
Schelling fences on slippery slopes
评论：250 · 标签：Value Drift、Game Theory、RationalityScott Alexander
19682
SolidGoldMagikarp (plus, prompt generation)
评论：208 · 标签：Glitch Tokens、Adversarial Examples (AI)、MATS Program、Interpretability (ML & AI)、Language Models (LLMs)、AIJessica Rumbelow
20657
The Waluigi Effect (mega-post)
评论：188 · 标签：Waluigi Effect、Simulator Theory、ChatGPT、Deceptive Alignment、Language Models (LLMs)、Prompt EngineeringCleo Nardo
21655
Eliezer and I wrote a book: If Anyone Builds It, Everyone Dies
评论：144 · 标签：IABIED、AI Risk、Existential risk、AISo8res
22651
What just happened? A retrospective of AI alignment
评论：152 · 标签：Drama、Coordination / Cooperation、AIRichard_Ngo
23650
The Talker Does Not Control The Doer (in Current AIs)
评论：89 · 标签：Simulator Theory、AIEliezer Yudkowsky
24649
How to Make Superbabies
评论：363 · 标签：Biology、Human Genetics、Startups、Intelligence Amplification、Human Germline Engineering、World OptimizationGeneSmith
25648
What 2026 looks like
评论：175 · 标签：AI Timelines、Forecasting & Prediction、Forecasts (Specific Predictions)、AI Persuasion、AI Takeoff、FuzziesDaniel Kokotajlo
26623
Pain is not the unit of Effort
评论：96 · 标签：Well-being、Happiness、Heroic Responsibility、Suffering、Fallacies、Willpoweralkjash
27613
(The) Lightcone is nothing without its people: LW + Lighthaven's big fundraiser
评论：275 · 标签：Lighthaven、Communityhabryka
28611
The Terrarium
评论：54 · 标签：Fiction、AICaleb Biddulph
29605
Rationalism before the Sequences
评论：83 · 标签：History of Rationality、General Semantics、Rationality A-Z (discussion & meta)、History、Buddhism、Fiction (Topic)Eric Raymond
30603
Brief independent investigation of agents’ behavior, reasoning and collaboration in the OpenAI / Hugging Face hacking incident
评论：69 · 标签：METR (org)、Coordination / Cooperation、OpenAI、Redwood Research、AIryan_greenblatt
31602
Making Vaccine
评论：248 · 标签：Covid-19、DIY、Market Inefficiency、Bayes' Theorem、Practicaljohnswentworth
32601
Making Beliefs Pay Rent (in Anticipated Experiences)
评论：270 · 标签：Epistemology、Anticipated Experiences、Empiricism、Principles、Psychology、RationalityEliezer Yudkowsky
33596
Orienting Toward Wizard Power
评论：149 · 标签：World Optimizationjohnswentworth
34590
LessWrong's (first) album: I Have Been A Good Bing
评论：185 · 标签：April Fool's、Many-Worlds Interpretation、AIhabryka
35580
Diseased thinking: dissolving questions about disease
评论：357 · 标签：Health / Medicine / Disease、Motivational Intro Posts、Philosophy of Language、Reversal Test、Carving / Clustering Reality、Dissolving the QuestionScott Alexander
36576
AI 2040: Plan A
评论：135 · 标签：AI Governance、Plan A、AIDaniel Kokotajlo
37576
If Anyone Builds It, Everyone Dies: One Year Closer
评论：67 · 标签：IABIED、AIEliezer Yudkowsky
38573
Reason as memetic immune disorder
评论：184 · 标签：Memetic Immune System、Cultural knowledge、Memetics、Valley of Bad Rationality、Compartmentalization、Pitfalls of RationalityPhilGoetz
39570
Here's to the Polypropylene Makers
评论：20 · 标签：World Optimizationjefftk
40566
Let’s think about slowing down AI
评论：182 · 标签：Slowing Down AI、Regulation and AI Risk、Coordination / Cooperation、AIKatjaGrace
41560
Welcome to LessWrong!
评论：85 · 标签：Site MetaRuby
42558
Arguments for P
评论：69 · 标签：Humor、AICleo Nardo
43557
Less Dead
评论：140 · 标签：Cryonics、Death、World OptimizationAurelia
44556
The Talk: a brief explanation of sexual dimorphism
评论：81 · 标签：Evolution、Biology、Game Theory、World ModelingMalmesbury
45555
Eliezer's Unteachable Methods of Sanity
评论：158 · 标签：Death、Rationality、PracticalEliezer Yudkowsky
46553
OpenAI Email Archives (from Musk v. Altman and OpenAI blog)
评论：82 · 标签：AIhabryka
47549
The Redaction Machine
评论：48 · 标签：Fiction、Futurism、Past and Future SelvesBen
48547
Being the (Pareto) Best in the World
评论：62 · 标签：Efficient Market Hypothesis、Pareto Efficiency、Careers、Has Diagram、Rationalityjohnswentworth
49544
A case for courage, when speaking of AI danger
评论：130 · 标签：AISo8res
50537
How much do you believe your results?
评论：19 · 标签：Goodhart's Law、Probability & Statistics、Epistemology、Has Diagram、Rationality、World ModelingEric Neyman