rationalnewsletter

Issue #71: Full List

According to self-reported studies, altruism predicts mating success in humans, supporting the hypothesis that altruism is a sexually selected costly signal of difficult-to-observe qualities. // drive.google.com

# Instrumental

How to Ignore Your Emotions (while also thinking you're awesome at emotions) // Hazard, 5 min Favorite

Does improved introspection cause rationalisation to become less noticeable? // jacobjacob, 1 min

What supplements do you use? // ioannes_shade, 1 min

# Epistemic

Keeping Beliefs Cruxy // Raemon, 4 min

Conversation on forecasting with Vaniver and Ozzie Gooen // jacobjacob, Vaniver, ozziegooen, 78 min

Proposed algorithm to fight anchoring bias // John_Maxwell_IV, 1 min

Arguments for the existence of qualia // Chris_Leong, 5 min

# Ai

Does it become easier, or harder, for the world to coordinate around not building AGI as time goes on? // elityre, 3 min

What does Optimization Mean, Again? (Optimizing and Goodhart Effects - Clarifying Thoughts, Part 2) // Davidmanheim, 5 min

Practical consequences of impossibility of value learning // Stuart_Armstrong, 3 min

Applying Overoptimization to Selection vs. Control (Optimizing and Goodhart Effects - Clarifying Thoughts, Part 3) // Davidmanheim, 3 min

Toy model piece #1: partial preferences revisited // Stuart_Armstrong, 4 min

Very different, very adequate outcomes // Stuart_Armstrong, 1 min

# Meta-ethics

End War Or Mosquitoes? // Robin Hanson, 1 min

# Anthropic

Sleeping Beauty: Is Disagreement While Sharing All Information Possible in Anthropic Problems? // dadadarren, 12 min

# Decision theory

Power Buys You Distance From The Crime // Elizabeth, 8 min Favorite

Why Subagents? // johnswentworth, 8 min

Permissions in Governance // sarahconstantin, 10 min

# Math and cs

Walkthrough: The Transformer Architecture [Part 1/2] // Matthew Barnett, 7 min

Rethinking Batch Normalization // Matthew Barnett, 9 min

Understanding Batch Normalization // Matthew Barnett, 7 min

Compilers/PLs book recommendation? // johnswentworth, 1 min

Walkthrough: The Transformer Architecture [Part 2/2] // Matthew Barnett, 14 min

# Ea

Against Against Billionaire Philanthropy // Scott Alexander, 23 min Favorite

Mistake Versus Conflict Theory of Against Billionaire Philanthropy // Zvi, 4 min

'Longtermism' definitional discussion on EA Forum // ESRogs, 1 min

How to evaluate neglectedness and tractability of aging research // emanuele ascani, 10 min

# Community

The Importance of Those Who Aren't Here // Davis_Kingsley, 2 min

Drive-By Low-Effort Criticism // lionhearted, 3 min

Gathering thoughts on Distillation // Raemon, 1 min

What woo to read? // calebo, 1 min

Adversarial Collaboration Update // Scott Alexander, 1 min

# Culture war

On Framing Political Opinions to Quickly Assess the Crux of Disagreement // Tuk, 2 min

Squareallworthy On UBI Plans // Scott Alexander, 4 min

# Misc

Forum participation as a research strategy // Wei_Dai, 3 min

When Having Friends is More Alluring than Being Right (by Ferrett Steinmetz) // CronoDAS, 1 min

Open problems in human rationality: guesses // romeostevensit, 1 min

Information empathy // Evan Rysdam, 1 min

Do you use twitter for intellectual engagement? Do you like it? // Hazard, 1 min

Will autonomous cars be more economical/efficient as shared urban transit than busses or trains, and by how much? What's some good research on this? // MakoYass, 2 min

Another case of "common sense" not being common? // jmh, 1 min

Mediocratopia: 7 // Venkatesh Rao, 4 min

Advice Wiki // Robin Hanson, 2 min

# Podcasts

90 – Secret to Our Success // The Bayesian Conspiracy, 116 min

# Videos of the week

How to ask for help -- and get a "yes" | Heidi Grant // TED, 11 min