rationalnewsletter

Issue #226: Full List

When it comes to the carbon footprint of our diet, what you eat matters much more than where it comes from and how it is packaged. // twitter.com

# Instrumental

How do you get a job as a software developer? // lsusr, 1 min

Skills Plateau Because Of Decay And Interference // Scott Alexander, 6 min

# Epistemic

The Parable of the Boy Who Cried 5% Chance of Wolf // ea247, 2 min

Novelty Generation - The Art of Good Ideas // DarkSym, 10 min

Vague concepts, family resemblance and cluster properties // Q Home, 8 min

No One-Size-Fit-All Epistemic Strategy // adamShimi, 2 min

"What Mistakes Are You Making Right Now?" // David Udell, 1 min

Are there practical exercises for developing the Scout mindset? // ChristianKl, 1 min

# Ai

A Mechanistic Interpretability Analysis of Grokking // neel-nanda-1, 48 min

What's General-Purpose Search, And Why Might We Expect To See It In Trained ML Systems? // johnswentworth, 11 min

How to do theoretical research, a personal perspective // mark-xu, 17 min

What's the Least Impressive Thing GPT-4 Won't be Able to Do // Algon, 1 min

Human Mimicry Mainly Works When We’re Already Close // johnswentworth, 6 min

Seeking Interns/RAs for Mechanistic Interpretability Projects // neel-nanda-1, 2 min

The Core of the Alignment Problem is... // thomas-larsen, 10 min

Brain-like AGI project "aintelope" // Gunnar_Zarncke, 1 min

PreDCA: vanessa kosoy's alignment protocol // carado-1, 7 min

Interpretability Tools Are an Attack Channel // Thane Ruthenis, 1 min

Announcing the Distillation for Alignment Practicum (DAP) // Jonas Hallgren, 3 min

AI Safety bounty for practical homomorphic encryption // acylhalide, 5 min

Matt Yglesias on AI Policy // grant-demaree, 1 min

Autonomy as taking responsibility for reference maintenance // ramana-kumar, 5 min

What if we approach AI safety like a technical engineering safety problem // zeshen, 8 min

Epistemic Artefacts of (conceptual) AI alignment research // Nora_Ammann, 6 min

Thoughts on 'List of Lethalities' // alex-lszn, 12 min

Benchmarking Proposals on Risk Scenarios // paulbricman, 16 min

Concrete Advice for Forming Inside Views on AI Safety // neel-nanda-1, 10 min

What Makes an Idea Understandable? On Architecturally and Culturally Natural Ideas. // Nicky, 19 min

Conditioning, Prompts, and Fine-Tuning // adam-jermyn, 4 min

Broad Picture of Human Values // Thane Ruthenis, 11 min

Alignment's phlogiston // ea-1, 2 min

A brief note on Simplicity Bias // Spencer Becker-Kahn, 5 min

Are language models close to the superhuman level in philosophy? // Roman Leventov, 2 min

What is the probability that a superintelligent, sentient AGI is actually infeasible? // Nathan1123, 1 min

Mesa-optimization for goals defined only within a training environment is dangerous // Rubi, 4 min

Deception as the optimal: mesa-optimizers and inner alignment // ea-1, 6 min

Understanding differences between humans and intelligence-in-general to build safe AGI // Florian_Dietz, 1 min

Intuitions Around General-Purpose Search // johnswentworth, 11 min

AI Transparency: Why it’s critical and how to obtain it. // zohar-jackson, 5 min

Limits of Asking ELK if Models are Deceptive // Oam Patel, 4 min

On Preference Manipulation in Reward Learning Processes // Felix Hofstätter, 4 min

No shortcuts to knowledge: Why AI needs to ease up on scaling and learn how to code // Hybris, 1 min

AGI Is Sacred // Robin Hanson, 5 min

# Meta-ethics

What does moral progress consist of? // jasoncrawford, 2 min

Against population ethics // jasoncrawford, 3 min

# Anthropic

Could the simulation argument also apply to dreams? // Nathan1123, 4 min

A First Attempt to Dissolve "Is Consciousness Reducible?" // DragonGod, 5 min

# Decision theory

Discovering Agents // zkenton, 7 min

And the Revenues Are So Small // Zvi, 13 min

Capital and inequality // NathanBarnard, 6 min

# Math and cs

What are the Limits on Computability? // DragonGod, 2 min

# Books

Review: Amusing Ourselves to Death // LRudL, 18 min

The Open Society and Its Enemies: Summary and Thoughts // matto, 21 min

Your Book Review: 1587, A Year Of No Significance // Scott Alexander, 35 min

MacAskill on Value Lock-In // Robin Hanson, 6 min

# Community

Paper is published! 100,000 lumens to treat seasonal affective disorder // Fabienne, 1 min

Oops It's Time To Overthrow the Organizer Day! // Screwtape, 4 min

[About Me] Cinera's Home Page // DragonGod, 9 min

Introducing the Existential Risks Introductory Course (ERIC) // TheMcDouglas, 7 min

Are you allocated optimally in your own estimation? // Emrik North, 1 min

Guided Consumption Theory: A Virtuous Dance between Altruistic Agents, Economic Discriminators, and Opportunistic Helpers // Brad West , 13 min

# Culture war

The Loire Is Not Dry // jkaufman, 1 min

# Fun

What Games These Days? // jkaufman, 3 min

Spoiler-Free Review: Across the Obelisk // Zvi, 7 min

# Misc

In Defense Of Making Money // George3d6, 9 min

What's up with the bad Meta projects? // yitz, 1 min

I’m mildly skeptical that blindness prevents schizophrenia // steve2152, 4 min

If a nuke is coming towards SF Bay can people bunker in BART tunnels? // DonyChristie

Playing Without Affordances // Alex Hollow, 1 min

Dealing With Delusions // adrusi, 1 min

So, I Want to Be a "Thinkfluencer" // DragonGod, 7 min

A conversation about progress and safety // jasoncrawford, 9 min

Spoons and Myofascial Trigger Points // vitaliya, 1 min

Is Nothing Sacred? // Robin Hanson, 3 min

# Podcasts

#136 – Will MacAskill on what we owe the future // , 174 min

EP 162 Max Borders on Decentralism // The Jim Rutt Show, 116 min

# Videos of the week

Magatte Wade: Africa, Capitalism, Communism, and the Future of Humanity | Lex Fridman Podcast #311 // Lex Fridman, 213 min