rational
Issue #230: Full List
18 September, 2022 // View curated list# Instrumental
Time is not the bottleneck (on making progress thinking about difficult things) // kman, 1 min
The Ultimate Step-by-Step Hiring Playbook // intellectronica, 4 min
# Epistemic
Many therapy schools work with inner multiplicity (not just IFS) // wallowinmaya, 21 min
Dan Luu on Futurist Predictions // T3t, 6 min
Intelligence failures and a theory of change for forecasting // NathanBarnard, 12 min
Ideas of the Gaps // Q Home, 15 min
In forecasting, how do accuracy, calibration and reliability relate to each other? // Maiwaaro23, 1 min
ε/δ Thinking // Venkatesh Rao, 8 min
# Ai
Takeaways from our robust injury classifier project [Redwood Research] // DMZ, 7 min
[Linkpost] A survey on over 300 works about interpretability in deep networks // scasper, 2 min
Understanding Conjecture: Notes from Connor Leahy interview // akash-wasil, 17 min
AGI safety researchers should focus (only/mostly) on deceptive alignment // marius-hobbhahn, 15 min
AI Safety field-building projects I'd like to see // akash-wasil, 8 min
AI Risk Intro 1: Advanced AI Might Be Very Bad // TheMcDouglas, 37 min
ACT-1: Transformer for Actions // daniel-kokotajlo, 1 min
Katja Grace on Slowing Down AI, AI Expert Surveys And Estimating AI Risk // mtrazzi, 4 min
When does technical work to reduce AGI conflict make a difference?: Introduction // JesseClifton, 8 min
Coordinate-Free Interpretability Theory // johnswentworth, 7 min
Alignment via prosocial brain algorithms // cameron-berg, 7 min
Why are we sure that AI will "want" something? // shminux, 1 min
How should DeepMind's Chinchilla revise our AI forecasts? // strawberry calm, 17 min
The Defender’s Advantage of Interpretability // marius-hobbhahn, 7 min
Deep Q-Networks Explained // Jay Bailey, 26 min
Forecasting thread: How does AI risk level vary based on timelines? // elifland, 1 min
Levels of goals and alignment // zeshen, 7 min
Argument against 20% GDP growth from AI within 10 years [Linkpost] // Aidan O'Gara, 5 min
Are Human Brains Universal? // DragonGod, 5 min
Sparse trinary weighted RNNs as a path to better language model interpretability // Am8ryllis, 3 min
When is intent alignment sufficient or necessary to reduce AGI conflict? // JesseClifton, 12 min
When would AGIs engage in conflict? // JesseClifton, 17 min
Precise P(doom) isn't very important for prioritization or strategy // harsimony, 1 min
ordering capability thresholds // carado-1, 5 min
A market is a neural network // david-hugh-jones, 10 min
Representational Tethers: Tying AI Latents To Human Ones // paulbricman, 19 min
Capability and Agency as Cornerstones of AI risk — My current model // wilm, 13 min
Should AI learn human values, human norms or something else? // Q Home, 3 min
General advice for transitioning into Theoretical AI Safety // martinsq, 12 min
Why do People Think Intelligence Will be "Easy"? // DragonGod, 1 min
Updates on FLI's Value Aligment Map? // rodeo_flagellum, 1 min
Would a Misaligned SSI Really Kill Us All? // DragonGod, 7 min
Are Speed Superintelligences Feasible for Modern ML Techniques? // DragonGod, 1 min
The problem with the media presentation of “believing in AI” // Roman Leventov, 1 min
Trying to find the underlying structure of computational systems // matthias-georg-mayer, 5 min
Ideological Inference Engines: Making Deontology Differentiable* // paulbricman, 16 min
Responding to 'Beyond Hyperanthropomorphism' // ukc10014, 18 min
Black Box Investigation Research Hackathon // esben-kran, 2 min
There is no royal road to alignment // ea-1, 3 min
Some ideas for epistles to the AI ethicists // Charlie Steiner, 5 min
Can you force a neural network to keep generalizing? // Q Home, 5 min
Improving Human Evaluation of Factual Accuracy in Language Models // soren-dunn, 12 min
Risk aversion and GPT-3 // hatta_afiq, 1 min
I Won My Three Year AI Progress Bet In Three Months // Scott Alexander, 5 min
# Longevity
Transhumanism, genetic engineering, and the biological basis of intelligence. // fowlertm, 1 min
Out with the old blood, in with the young blood // Josh Mitteldorf, 8 min
# Anthropic
What's the longest a sentient observer could survive in the Dark Era? // Raemon, 1 min
A Pin and a Balloon: Anthropic Fragility Increases Chances of Runaway Global Warming // avturchin, 64 min
Simple proofs of the age of the universe (or other things) // Astynax, 1 min
# Decision theory
Logical counterfactuals are brittle; use conditionals // Sylvester Kollin, 24 min
I'm taking a course on game theory and am faced with this question. What's the rational decision? // dalton-mabery, 1 min
# Math and cs
Most sensible abstraction & feature set for a systems language // jasen-q, 12 min
Is there an Ultimate text editor? // johannes-c-mayer, 1 min
# Ea
Effective altruism in the garden of ends // tyler-alterman, 31 min
Introducing School of Thinking // Luca Parodi, 11 min
# Fun
D&D.Sci September 2022: The Allocation Helm // abstractapplic, 1 min
Simple 5x5 Go // jkaufman, 1 min
# Misc
Why Do People Think Humans Are Stupid? // DragonGod, 3 min
Freeloading? // jkaufman, 2 min
Emergency Residential Solar Jury-Rigging // jkaufman, 3 min
What would happen if we abolished the FDA tomorrow? // yair-halberstadt, 5 min
Seeing the Schema // vitaliya, 1 min
Guidelines for Mad Entrepreneurs // David Udell, 14 min
Affordable Housing Investment Fund // jkaufman, 1 min
Towards a philosophy of safety // jasoncrawford, 9 min
The Value of Not Being an Imposter // sudo, 1 min
Guesstimate Algorithm for Medical Research // pktechgirl, 9 min
Sequencing Intro II: Adapters // jkaufman, 2 min
Unpredictable Reward, Predictable Happiness // Scott Alexander, 13 min
Contemporary Linguistics: A Perspective on Research and Information Sharing // Miniman, 3 min
# Rational fiction
I’ve written a Fantasy Novel to Promote Effective Altruism // timothy-underwood-1, 16 min
I wrote a fantasy novel to promote EA: More Chapters // timothy-underwood-1, 56 min