I have to look for a while before finding any non-AI posts. Seems LW is mainly an AI / alignment discussion forum at this point.
1It seems more informative to just look at top (inflation adjusted) karma for 2022 (similar to what habryka noted in the sibling). AI posts in bold.
- AGI Ruin: A List of LethalitiesΩ
- Where I agree and disagree with EliezerΩ
- SimulatorsΩ
- What an actually pessimistic containment strategy looks like
- Let’s think about slowing down AIΩ
- Luck based medicine: my resentful story of becoming a medical miracle
- Counter-theses on Sleep
- Losing the root for the tree
- The Redaction Machine
- It Looks Like You're Trying To Take Over The WorldΩ
- (My understanding of) What Everyone in Technical Alignment is Doing and WhyΩ
- Counterarguments to the basic AI x-risk caseΩ
- It’s Probably Not Lithium
- Reflections on six months of fatherhood
- chinchilla's wild implicationsΩ
- You Are Not Measuring What You Think You Are Measuring [AI related]
- Lies Told To Children
- What DALL-E 2 can and cannot do
- Staring into the abyss as a core life skill
- DeepMind alignment team opinions on AGI ruin argumentsΩ
- Accounting For College Costs
- A Mechanistic Interpretability Analysis of GrokkingΩ
- Models Don't "Get Reward"Ω
- Why I think strong general AI is coming soon
- Without specific countermeasures, the easiest path to transformative AI likely leads to AI takeoverΩ
- Why Agent Foundations? An Overly Abstract ExplanationΩ
- MIRI announces new "Death With Dignity" strategy
- Beware boasting about non-existent forecasting track records [AI related]
- A challenge for AGI organizations, and a challenge for readersΩ
I count 18/29 about AI. A few AI posts are technically more general. A few non-AI posts seem to indirectly be about AI.
5
1I think the AI posts are definitely substantially more interlinked than the non-AI posts, so I think specific metric oversamples AI posts.
I just updated the spreadsheet to include All Time posts. Lists of Lethalities is still the winner by Total Pingback Karma, although not by Pingback Count (and this seems at least partially explained by karma inflation)
1The list would look pretty different if self-cites were excluded. E.g. my posts would probably all be gone 😂
1Yeah if I have time today I'll make an "exclude self-cites" column, although fwiw I think the "total pingback karma" is fairly legit even if including self-cites. If your followup work got a lot of karma, I think that's a useful signal about your original post even if you quote yourself liberally.
For the past couple years I've wished LessWrong had a "sort posts by number of pingbacks, or, ideally, by total karma of pingbacks". I particularly wished for this during the Annual Review, where "which posts got cited the most?" seemed like a useful thing to track for potential hidden gems.
We still haven't built a full-fledged feature for this, but I just ran a query against the database, and made it into a spreadsheet, which you can view here:
LessWrong 2022 Posts by Pingbacks
Here are the top 100 posts, sorted by Total Pingback Karma
Title/LinkPost KarmaPingback CountTotal Pingback KarmaAvg Pingback KarmaAGI Ruin: A List of LethalitiesMIRI announces new "Death With Dignity" strategyA central AI alignment problem: capabilities generalization, and the sharp left turnSimulatorsWithout specific countermeasures, the easiest path to transformative AI likely leads to AI takeoverReward is not the optimization targetA Mechanistic Interpretability Analysis of GrokkingHow To Go From Interpretability To Alignment: Just Retarget The SearchOn how various plans miss the hard bits of the alignment challenge[Intro to brain-like-AGI safety] 3. Two subsystems: Learning & SteeringHow likely is deceptive alignment?The shard theory of human valuesMysteries of mode collapse[Intro to brain-like-AGI safety] 2. “Learning from scratch” in the brainWhy Agent Foundations? An Overly Abstract ExplanationA Longlist of Theories of Impact for InterpretabilityHow might we align transformative AI if it’s developed very soon?A transparency and interpretability tech treeDiscovering Language Model Behaviors with Model-Written EvaluationsA note about differential technological developmentCausal Scrubbing: a method for rigorously testing interpretability hypotheses [Redwood Research]Supervise Process, not OutcomesShard Theory: An OverviewEpistemological Vigilance for AlignmentA shot at the diamond-alignment problemWhere I agree and disagree with EliezerBrain Efficiency: Much More than You Wanted to KnowRefine: An Incubator for Conceptual Alignment Research BetsExternalized reasoning oversight: a research direction for language model alignmentHumans provide an untapped wealth of evidence about alignmentSix Dimensions of Operational Adequacy in AGI ProjectsHow "Discovering Latent Knowledge in Language Models Without Supervision" Fits Into a Broader Alignment SchemeGodzilla Strategies(My understanding of) What Everyone in Technical Alignment is Doing and WhyTwo-year update on my personal AI timelines[Intro to brain-like-AGI safety] 15. Conclusion: Open problems, how to help, AMA[Intro to brain-like-AGI safety] 6. Big picture of motivation, decision-making, and RLHuman values & biases are inaccessible to the genomeYou Are Not Measuring What You Think You Are MeasuringOpen Problems in AI X-Risk [PAIS #5][Intro to brain-like-AGI safety] 1. What's the problem & Why work on it now?Conditioning Generative ModelsConjecture: Internal Infohazard PolicyA challenge for AGI organizations, and a challenge for readersSuperintelligent AI is necessary for an amazing future, but far from sufficientOptimality is the tiger, and agents are its teethLet’s think about slowing down AINiceness is unnaturalAnnouncing the Alignment of Complex Systems Research Group[Intro to brain-like-AGI safety] 13. Symbol grounding & human social instinctsELK prize resultsAbstractions as Redundant Information[Link] A minimal viable product for alignmentAcceptability Verification: A Research AgendaWhat an actually pessimistic containment strategy looks likeLet's See You Write That Corrigibility Tagchinchilla's wild implicationsWorlds Where Iterative Design Failswhy assume AGIs will optimize for fixed goals?Gradient hacking: definitions and examplesContra shard theory, in the context of the diamond maximizer problemWe Are Conjecture, A New Alignment Research StartupCircumventing interpretability: How to defeat mind-readersEvolution is a bad analogy for AGI: inner alignmentRefining the Sharp Left Turn threat model, part 1: claims and mechanismsMATS ModelsCommon misconceptions about OpenAIPrizes for ELK proposalsCurrent themes in mechanistic interpretability researchDiscovering Agents[Intro to brain-like-AGI safety] 12. Two paths forward: “Controlled AGI” and “Social-instinct AGI”What's General-Purpose Search, And Why Might We Expect To See It In Trained ML Systems?Inner and outer alignment decompose one hard problem into two extremely hard problemsThreat Model Literature ReviewLanguage models seem to be much better than humans at next-token predictionWill Capabilities Generalise More?Pivotal outcomes and pivotal processesConditioning Generative Models for AlignmentTraining goals for large language modelsIt’s Probably Not LithiumLatent Adversarial Training“Pivotal Act” Intentions: Negative Consequences and Fallacious ArgumentsConditioning Generative Models with RestrictionsThe alignment problem from a deep learning perspectiveInstead of technical research, more people should focus on buying timeBy Default, GPTs Think In Plain Sight[Intro to brain-like-AGI safety] 4. The “short-term predictor”Don't leave your fingerprints on the futureStrategy For Conditioning Generative ModelsCall For DistillersThoughts on AGI organizations and capabilities workOptimization at a Distance[Intro to brain-like-AGI safety] 5. The “long-term predictor”, and TD learningWhat does it take to defend the world against out-of-control AGIs?Monitoring for deceptive alignmentLate 2021 MIRI Conversations: AMA / DiscussionHow to Diversify Conceptual Alignment: the Model Behind Refinewrapper-minds are the enemyBut is it really in Rome? An investigation of the ROME model editing techniqueAn Open Agency Architecture for Safe Transformative AI