text stringlengths 300 320k | source stringlengths 52 154 |
|---|---|
# Pitfalls with Proofs
This post distills [some](https://arxiv.org/abs/1902.09469) [excellent](https://www.lesswrong.com/posts/hpAbfXtqYC2BrpeiC/troll-bridge-5) [writing](https://www.lesswrong.com/posts/yXfka98pZXAmXiyDp/my-current-take-on-counterfactuals) on decision theory from Abram Demski about spurious proofs and... | https://www.lesswrong.com/posts/ofKg5dKkqqJZF3bgX/pitfalls-with-proofs |
# What are the simplest questions in applied rationality where you don't know the answer to?
In the podcast between Spencer Greenberg and Buck Shlegeris, [Taking pleasure in being wrong](https://open.spotify.com/episode/4Q54moES6YvjCEyOORRgX3), Buck says:
I think that when you are learning subjects something that yo... | https://www.lesswrong.com/posts/caB6ApqG7rSFayoaq/what-are-the-simplest-questions-in-applied-rationality-where |
# How to Diversify Conceptual Alignment: the Model Behind Refine
*This post is part of the work done at* [*Conjecture*](https://conjecture.dev)*.*
Tl;dr: We need far more conceptual AI alignment research approaches than we have now if we want to increase our chances to solve the alignment problem. However, the concep... | https://www.lesswrong.com/posts/5uiQkyKdejX3aEHLM/how-to-diversify-conceptual-alignment-the-model-behind |
# Enlightenment Values in a Vulnerable World
Introduction:
==============
*The Vulnerable World Hypothesis: If technological development continues then a set of capabilities will at some point be attained that make the devastation of civilization extremely likely, unless civilization sufficiently exits the semianarc... | https://www.lesswrong.com/posts/yHsWbgvQ78gzBrCPt/enlightenment-values-in-a-vulnerable-world |
# Cognitive Risks of Adolescent Binge Drinking
The takeaway
============
Our goal was to quantify the cognitive risks of heavy but not abusive alcohol consumption. This is an inhernetly difficult task: the world is noisy, humans are highly variable, and institutional review boards won’t let us do challenge trials of ... | https://www.lesswrong.com/posts/rwvN3x5fDbaiLBbde/cognitive-risks-of-adolescent-binge-drinking |
# Personal forecasting retrospective: 2020-2022
Overview
========
I’ve been forecasting with some level of activity for over 2 years now, so I’m overdue for a retrospective.[^nf2o6fne29e]
I discuss:
1. [My overall track record on each platform](#Platform_track_records): My track record is generally strong, though ... | https://www.lesswrong.com/posts/XAkhqkNQByEaT8MED/personal-forecasting-retrospective-2020-2022 |
# Don't take the organizational chart literally
> "Do you know how much power I'd have to give up to be President?" - Lex Luthor
The CIA, like all intelligence agencies, spends much of its time managing and creating spies. Doing that is its primary job, and though it's also the clandestine operations arm of the U.S. ... | https://www.lesswrong.com/posts/LyywLDkw3Am9gbQXd/don-t-take-the-organizational-chart-literally |
# [AN #173] Recent language model results from DeepMind
HIGHLIGHTS
[Scaling Language Models: Methods, Analysis & Insights from Training Gopher](https://storage.googleapis.com/deepmind-media/research/language-research/Training%20Gopher.pdf) _(Jack W. Rae et al)_ (summarized by Rohin): This paper details the training o... | https://www.lesswrong.com/posts/HXDkCtk9tae5wFmjG/an-173-recent-language-model-results-from-deepmind |
# Why are politicians polarized?
There has been a lot of attention dedicated to increased political polarization. And a lot of it has focused on the factors causing *voters* to be more polarized. I want to discuss a different, but related question. *Given* that voters have become more polarized, why are *politicians* ... | https://www.lesswrong.com/posts/EdoCrSGM8FzJhez3C/why-are-politicians-polarized |
# Is Gas Green?
The EU decided to declare gas along with nuclear to be[ green](https://www.dw.com/en/european-commission-declares-nuclear-and-gas-to-be-green/a-60614990). It's easy to see why one might see nuclear energy as a green energy source, it's harder to see the case for gas.
As I[ previously pointed out](htt... | https://www.lesswrong.com/posts/wmKn6wLx87eBPgepY/is-gas-green |
# How much to optimize for the short-timelines scenario?
Some have argued that one should tend to act as if timelines are short since in that scenario it's possible to have more expected impact. But I haven't seen a thorough analysis of this argument.
**Question:** Is this argument valid and if yes how strong is it?
... | https://www.lesswrong.com/posts/iJYzREEsuG8g95pvC/how-much-to-optimize-for-the-short-timelines-scenario |
# Covid 7/21/22: Featuring ASPR
The slow transition from ‘almost all about Covid’ to ‘only somewhat about Covid’ continues, with the non-Covid various-things-to-note portion this week reaching something like 25%. My presumption is that this portion will continue rising over time.
What’s happening with Covid? More of ... | https://www.lesswrong.com/posts/4yHkjpnctHvJYWmnZ/covid-7-21-22-featuring-aspr |
# YouTubeTV and Spoilers
Author’s Note: The purpose of this is a combination of fun rant, illustration of a general failure mode, that I am going to eventually use this for something in at least some worlds, and the fact that a decent number of workers at Google read my blog and it would be nice if they actually went ... | https://www.lesswrong.com/posts/bjKJPC4Ex49tdmRcG/youtubetv-and-spoilers |
# Which personalities do we find intolerable?
**TLDR**: I suspect that we will generally be intolerant of other people’s personalities when:
1. We are on the culturally decided “right” side of a particular trait.
2. And the other person is on the “wrong” side of that same trait.
I have no real evidence beyond anec... | https://www.lesswrong.com/posts/a9SaYAkh3ir8Si9wg/which-personalities-do-we-find-intolerable |
# Changing the world through slack & hobbies
*This post has been recorded as part of the LessWrong Curated Podcast, and can be listened to on* [*Spotify*](https://open.spotify.com/episode/18HoIZed8RIrsjtE5iN5oi?si=9c815bb3a6fd4bc6)*,* [*Apple Podcasts*](https://podcasts.apple.com/us/podcast/changing-the-world-through-... | https://www.lesswrong.com/posts/DdDt5NXkfuxAnAvGJ/changing-the-world-through-slack-and-hobbies |
# Our Existing Solutions to AGI Alignment (semi-safe)
This post addresses the best methods we have of creating (relatively) aligned AGI. The safety of these methods IS NOT guaranteed. AI Alignment is not a solved issue.
These methods *appear* to offer better solutions than existing methods, such as simple maximizers... | https://www.lesswrong.com/posts/ubQDcDxjNJ2Exp3ni/our-existing-solutions-to-agi-alignment-semi-safe |
# Conditioning Generative Models with Restrictions
*This is a followup to* [*Conditioning Generative Models*](https://www.alignmentforum.org/posts/nXeLPcT9uhfG3TMPS/conditioning-generative-models) *based on further discussions with Evan Hubinger, Nicholas Schiefer, Abram Demski, Curtis Huebner, Hoagy Cunningham, Derek... | https://www.lesswrong.com/posts/adiszfnFgPEnRsGSr/conditioning-generative-models-with-restrictions |
# Internal Double Crux
*Author's note: While most CFAR workshops taught double crux first and then internal double crux afterward, it's long been my opinion that both concepts would be better served in the other order. Recent experimentation has shown results consistent with this hypothesis; practicing collaborative t... | https://www.lesswrong.com/posts/x2KrcscqgKDk6pMeD/internal-double-crux-1 |
# One-day applied rationality workshop in Berlin Aug 29 (after LWCW)
Hi, all, former CFAR employee here,
I'll be giving a one-day, crash-course introduction to some of the core concepts from CFAR's applied rationality workshops in Berlin on the Monday of the [LessWrong Community Weekend](https://www.lesswrong.com/eve... | https://www.lesswrong.com/posts/pEFzqeTCBfo4nrN47/one-day-applied-rationality-workshop-in-berlin-aug-29-after |
# Making DALL-E Count
Can DALL-E count? Does DALL-E's counting ability depend on *what* it is counting? Let's find out!
I only generate the images once and paste them in along with the prompt in quotations.
**Numbers**
===========
"Zero"
------
, common misconceptions about EleutherAI and his new AI Alignment company [Conjec... | https://www.lesswrong.com/posts/zk6RK3xFaDeJHsoym/connor-leahy-on-dying-with-dignity-eleutherai-and-conjecture |
# Trying out Prompt Engineering on TruthfulQA
*I try out "let's gather the relevant facts" as a zero-shot question answering aid on TruthfulQA. It doesn't help more than other helpful prompts. Possibly it might work better on more typical factual questions.*
*This post could potentially be useful to people interested... | https://www.lesswrong.com/posts/9goJrmQDT96eGAgit/trying-out-prompt-engineering-on-truthfulqa |
# Which values are stable under ontology shifts?
Here’s a rough argument which I’ve been thinking about lately:
We have coherence theorems which say that, if you’re not acting like you’re maximizing expected utility over outcomes, you’d make payments which predictably lose you money. But in general I don't see any a ... | https://www.lesswrong.com/posts/Cuig4qe8m2aqBCJtZ/which-values-are-stable-under-ontology-shifts |
# On Akrasia, Habits and Reward Maximization
Content Warning: Discussion of Free Will. Please do not read further if you believe in free will and suspect that this belief is important to your mental well-being.
For most of my life, I've struggled with terrible akrasia, like many here. While it wasn't bad enough to... | https://www.lesswrong.com/posts/BvGSDfBR6KmPkaafg/on-akrasia-habits-and-reward-maximization |
# Eating Boogers
Should you ~~eat boogers~~ practice automucophagy?
The science on this is surprisingly lacking. Maybe you’re one of the normal people that find it gross and refuses to engage in the behavior for fear of breaking social norms.
But what about us, bold thinkers, outspoken intelligentsia willing to thru... | https://www.lesswrong.com/posts/LSbpCoQBC3PG4zSJj/eating-boogers |
# Robustness to Scaling Down: More Important Than I Thought
*(Edit: I added text between "...was really about reductions." and "To use the mental move of robustness...", because comments showed me I hadn't made my meaning clear enough.)*
*This post is part of the work done at* [*Conjecture*](https://conjecture.dev)*.... | https://www.lesswrong.com/posts/pA3F9oejzvGg6Kf3a/robustness-to-scaling-down-more-important-than-i-thought |
# What Environment Properties Select Agents For World-Modeling?
*Thanks to John Wentworth for helpful critique.*
0\. Introduction
----------------
Agency/goal-driven behavior is a complex thing. By definition, it implies [a system that is deliberately choosing to take actions that will achieve the goal it wants](htt... | https://www.lesswrong.com/posts/hpBtyssp49gP6tgWN/what-environment-properties-select-agents-for-world-modeling |
# A Bias Against Altruism
I've noticed a phenomenon in our culture whereby altruistic actions face much more scrutiny than selfish actions. I think we should be aware of this bias effect, especially when discussing incentivizing people to e.g. work on AI alignment instead of ML research.
When I say 'our culture', I m... | https://www.lesswrong.com/posts/SdPJqJ7Gz4G63kveq/a-bias-against-altruism |
# Curating "The Epistemic Sequences" (list v.0.1)
***Summary for regular readers:** The epistemic content of The Sequences — i.e., the advice on finding true beliefs — has a different epistemic status than the instrumental content — i.e., the advice on how to behave. Specifically, the epistemic content is based upon ... | https://www.lesswrong.com/posts/kS2A5wvMEuu5SjXj9/curating-the-epistemic-sequences-list-v-0-1 |
# Easy guide for running a local Rationality meetup
This weekend I've participated in a retreat in Berkeley for organizers of prior ACX meetups, with around 50 meetup organizers attending. This post serves to write down my learnings from the sessions on the basic logistics of meetups and hopefully provide a reference ... | https://www.lesswrong.com/posts/HoxwcmD8SNEFWECnP/easy-guide-for-running-a-local-rationality-meetup |
# What's next for instrumental rationality?
Preceded by: [Curating "The Epistemic Sequences" (list v.0.1)](https://www.lesswrong.com/posts/kS2A5wvMEuu5SjXj9/curating-the-epistemic-sequences-list-v-0-1)
*Epistemic status: speculations and ideations from me about the potential for further progress on broadly accessible... | https://www.lesswrong.com/posts/irKEJpkWDuNTsxJtp/what-s-next-for-instrumental-rationality |
# Eavesdropping on Aliens: A Data Decoding Challenge
This is my reply to faul_sname's claim in the "[I No Longer Believe Intelligence To Be Magical](https://www.lesswrong.com/posts/rqAvheoRHwSDsXrLw/i-no-longer-believe-intelligence-to-be-magical?commentId=yHafNW3DMadoyMYuK)" thread:
> In terms of concrete predictions... | https://www.lesswrong.com/posts/BMDfYGWcsjAKzNXGz/eavesdropping-on-aliens-a-data-decoding-challenge |
# Double Crux
*Author's note: There is a preexisting standalone essay on double crux (also written by Duncan Sabien) available* [*here*](https://www.lesswrong.com/posts/exa5kmvopeRyfJgCy/double-crux-a-strategy-for-mutual-understanding)*. The version in the handbook is similar, but has enough disoverlap that it seemed... | https://www.lesswrong.com/posts/WLQspe83ZkiwBc2SR/double-crux |
# Hiring Programmers in Academia
A professor I know through the EA community has been trying to hire a software engineer for their research, and they explained some privately about how this is tricky. The following are points I took away that might be useful to programmers considering job postings, people in academia ... | https://www.lesswrong.com/posts/9GweYgHABZAjH6T6f/hiring-programmers-in-academia |
# The Most Important Century: The Animation
This is a linkpost for the Rational Animations' video based on [The Most Important Century](https://forum.effectivealtruism.org/s/isENJuPdB3fhjWYHd) sequence of posts by Holden Karnofsky.
Below, the whole script of the video.
Matthew Barnett has written most of it.... | https://www.lesswrong.com/posts/yhCkaocWYkdX3i9AG/the-most-important-century-the-animation |
# Impact of " 'Let's think step by step' is all you need"?
*Title from:* [*https://twitter.com/arankomatsuzaki/status/1529278581884432385*](https://twitter.com/arankomatsuzaki/status/1529278581884432385)*.*
This is not quite a linkpost for [this](https://arxiv.org/abs/2205.11916) paper.
Nonetheless, the abstract is:... | https://www.lesswrong.com/posts/DaPi3ZrEkydycwHzb/impact-of-let-s-think-step-by-step-is-all-you-need |
# Finding Skeletons on Rashomon Ridge
*A product of a* [*SERI MATS*](https://www.serimats.org/) *research sprint (taking 1.5 weeks).*
> Cannot yet assign positively to animal or vegetable kingdom, but odds now favour animal. Probably represents incredibly advanced evolution of radiata without loss of certain primitiv... | https://www.lesswrong.com/posts/PJhvcTkwpGr9Ysmcd/finding-skeletons-on-rashomon-ridge |
# Brainstorm of things that could force an AI team to burn their lead
| **Comments:** The following is a list (very lightly edited with help from Rob Bensinger) I wrote in July 2017, at Nick Beckstead’s request, as part of a conversation we were having at the time. From my current vantage point, it strikes me as narro... | https://www.lesswrong.com/posts/p3s8RvkcyTwzu27ps/brainstorm-of-things-that-could-force-an-ai-team-to-burn |
# Reward is not the optimization target
*This insight was made possible by many conversations with Quintin Pope, where he challenged my implicit assumptions about alignment. I’m not sure who came up with this particular idea.*
In this essay, I call an agent a “reward optimizer” if it not only gets lots of reward, but... | https://www.lesswrong.com/posts/pdaGN6pQyQarFHXF4/reward-is-not-the-optimization-target |
# Unifying Bargaining Notions (1/2)
This is a two-part sequence of posts, in the ancient LessWrong tradition of decision-theory-posting. This first part will introduce various concepts of bargaining solutions and dividing gains from trade, which the reader may or may not already be familiar with.
[The upcoming part](... | https://www.lesswrong.com/posts/rYDas2DDGGDRc8gGB/unifying-bargaining-notions-1-2 |
# Opening Session Tips & Advice
**Meta**
--------
CFAR ran many, many workshops.
After each workshop, there would be feedback from the participants, and debrief discussions among the staff. We would talk about what had worked and what hadn't, what we wish had been said or done, what we would try differently in the ... | https://www.lesswrong.com/posts/gKeHcikcXA3bApyoM/opening-session-tips-and-advice |
# AGI Safety Needs People With All Skillsets!
*(Cross-posted to* [*EA Forum*](https://forum.effectivealtruism.org/posts/NJtC8xzD8BgF3TmEp/agi-safety-needs-people-with-all-skillsets)*)*
*, where the videos are posted which I can't upload here - take a look at those* [*here*](https://www.pchvykov.com/post/mistakes-as-agency)*. *
We’re all somehow familiar with the concept that a person's actions can be “t... | https://www.lesswrong.com/posts/aMrhJbvEbXiX2zjJg/mistakes-as-agency |
# Alignment being impossible might be better than it being really difficult
*Epistemic status: Thinking out loud.*
***TL;DR:** If alignment is just really difficult (or impossible for humanity), we might end up with an unaligned superintelligence which itself solves the alignment problem, gaining exponentially more p... | https://www.lesswrong.com/posts/CBpatGkdEqr4hCCAW/alignment-being-impossible-might-be-better-than-it-being |
# Utility functions and probabilities are entangled
*Originally posted as an* [*EA forum comment*](https://forum.effectivealtruism.org/posts/9pJ9CwsNmJQ8BmgSt/the-ea-movement-s-values-are-drifting-you-re-allowed-to-stay?commentId=kuh2ZSoDxmkh3eejE)*.*
Suppose that there are two effective altruist billionaires, April ... | https://www.lesswrong.com/posts/zsqvW3voHjadEhZzx/utility-functions-and-probabilities-are-entangled |
# NeurIPS ML Safety Workshop 2022
We're excited to announce the NeurIPS [ML Safety workshop](https://neurips2022.mlsafety.org/)! To our knowledge it is the first workshop at a top ML conference to emphasize and explicitly discuss x-risks.
X-Risk Analysis Prizes
----------------------
$100K in paper prizes will be aw... | https://www.lesswrong.com/posts/pY4J2qNaHgKp2nbEd/neurips-ml-safety-workshop-2022 |
# Active Inference as a formalisation of instrumental convergence
The goal of this post is mainly to increase the exposure of the AI alignment community to Active Inference theory, which seems to be highly relevant to the problem but is seldom mentioned on the forum.
This post links to a freely available book about A... | https://www.lesswrong.com/posts/ostLZyhnBPndno2zP/active-inference-as-a-formalisation-of-instrumental |
# «Boundaries» Sequence (Index Post)
In an upcoming short sequence of posts, I'm going to try to describe a causal pathway from
1. a key missing idea in the utility-theoretic foundations of game theory, leading to
2. some problems I think I see in effective altruism discourse, leading further to
3. gaps in some a... | https://www.lesswrong.com/posts/MNntt22m7jutNyJqW/boundaries-sequence-index-post |
# «Boundaries», Part 1: a key missing concept from utility theory
*This post has been recorded as part of the LessWrong Curated Podcast, and can be listened to on* [*Spotify*](https://open.spotify.com/episode/73rKuCaxCaAbLqNIvRVi99)*,* [*Apple Podcasts*](https://podcasts.apple.com/us/podcast/boundaries-part-1-a-key-mi... | https://www.lesswrong.com/posts/8oMF8Lv5jiGaQSFvo/boundaries-part-1-a-key-missing-concept-from-utility-theory |
# Technocracy and the Space Age
Earlier I [proposed a hypothesis](https://rootsofprogress.org/technocracy-hypothesis) that the 1930s–60s in the US were characterized by the attempt to achieve progress through top-down control by a technical elite.
The 1997 preface to Walter McDougall’s Pulitzer-winning [*The Heavens ... | https://www.lesswrong.com/posts/m9j9XfHHmXGNqooML/technocracy-and-the-space-age |
# Norfolk Rationalists Weekly Wednesday Meetup @ Fair Grounds
Hello!
Come on out to Fair Grounds for good conversation, good food & drink, and a lovely atmosphere :)
When: We meet each Wednesday from 17:00-19:30 EDT.
Where: Fair Grounds, an excellent coffee shop located at 806 Baldwin Ave # 2, Norfolk, VA 23517. Do... | https://www.lesswrong.com/events/LACAGY6H8wRcdEk2a/norfolk-rationalists-weekly-wednesday-meetup-fair-grounds-8 |
# AGI ruin scenarios are likely (and disjunctive)
*Note: As usual, Rob Bensinger helped me with editing. I recently discussed this model with Alex Lintz, who might soon post his own take on it (edit:* [*here*](https://forum.effectivealtruism.org/posts/eggdG27y75ot8dNn7/three-pillars-for-avoiding-agi-catastrophe-techni... | https://www.lesswrong.com/posts/ervaGwJ2ZcwqfCcLx/agi-ruin-scenarios-are-likely-and-disjunctive |
# Unifying Bargaining Notions (2/2)
Alright, time for the payoff, unifying everything discussed in the [previous post](https://www.lesswrong.com/posts/rYDas2DDGGDRc8gGB/unifying-bargaining-notions-1-2). This post is a lot more mathematically dense, you might want to digest it in more than one sitting.
**Imaginary Pr... | https://www.lesswrong.com/posts/RZNmNwc9SxdKayeQh/unifying-bargaining-notions-2-2 |
# Levels of Pluralism
*This post is part of the work done at* [*Conjecture*](https://conjecture.dev)*.*
Preparing the Pluralism Question
================================
When do we want all our research eggs in the same paradigm basket?
Although most people don't go as far as the extreme paradigmatism of [Thomas Ku... | https://www.lesswrong.com/posts/wi3upQibefMcFs5to/levels-of-pluralism |
# Progress links and tweets, 2022-07-27
**Opportunities**
-----------------
* [Arcadia Science launches an Entrepreneur-in-Residence (EIR) program](https://jobs.lever.co/arcadiascience/36090acd-7dfa-4d80-ab21-fdc37d940b9c) (thread from [@ArcadiaScience](https://twitter.com/ArcadiaScience/status/1549439425150537728)... | https://www.lesswrong.com/posts/WsnFKuAH6DEXScimg/progress-links-and-tweets-2022-07-27 |
# Moral strategies at different capability levels
Let’s consider three ways you can be altruistic towards another agent:
* You care about their welfare: some metric of how good their life is (as defined by you). I’ll call this care-morality - it endorses things like promoting their happiness, reducing their sufferi... | https://www.lesswrong.com/posts/jDQm7YJxLnMnSNHFu/moral-strategies-at-different-capability-levels |
# Principles of Privacy for Alignment Research
The hard/useful parts of alignment research are largely about understanding agency/intelligence/etc. That sort of understanding naturally yields capabilities-relevant insights. So, alignment researchers naturally run into decisions about how private to keep their work.
T... | https://www.lesswrong.com/posts/SsuqYoyBnheSj7jLw/principles-of-privacy-for-alignment-research |
# Seeking beta readers who are ignorant of biology but knowledgeable about AI safety
*Seeking the biology-ignorant! Come and do your part!*
I am working on a post with Beth Barnes that will explain meiotic drive to AI safety people so they can determine if there are any useful analogies to gradient hacking. If yo... | https://www.lesswrong.com/posts/wBP7pzQwwhNGspwRJ/seeking-beta-readers-who-are-ignorant-of-biology-but |
# For Better Commenting, Stop Out Loud
"Man stands face to face with the irrational. He feels within him his longing for happiness and for reason. The absurd is born of this confrontation between the hum... | https://www.lesswrong.com/posts/yyunbQTRKfWZHw2o3/for-better-commenting-stop-out-loud |
# Monkeypox Post #2
Enough distinct things happened this week in Monkeypox that it made sense to split off the information on it into a distinct post. This is probably _slightly_ over the threshold required for doing that.
### Executive Summary
[WHO declares Monkeypox a public health emergency of international conce... | https://www.lesswrong.com/posts/fp9fcyK8weYcyqjNK/monkeypox-post-2 |
# Covid 7/28/22: Ruining It For Everyone
There are lots of ways to try and ruin it for everyone. Covid restrictions that make no sense are a proven successful strategy for this, demanding the Doing of More. Monkeypox offers exciting additional ruining potential.
The surprise winner, however, is an old Alzheimer’s pap... | https://www.lesswrong.com/posts/gH9NLAyqLKzfsFBrh/covid-7-28-22-ruining-it-for-everyone |
# The Conversations We Make Space For
*(Cross-posted to* [*EA Forum*](https://forum.effectivealtruism.org/posts/LFmvDhxBpA8C5jsK8/the-conversations-we-make-space-for)*)*
While developing the necessary skills and models for making [The Berlin Hub](https://forum.effectivealtruism.org/posts/Yt8jv8Fby6R9DrEg7/the-berlin-... | https://www.lesswrong.com/posts/hg4aamsniibSYbd9F/the-conversations-we-make-space-for |
# The generalized Sierpinski-Mazurkiewicz theorem.
Suppose any uncountable set $S$ and any countable set of bijective transformations $\mathbb T=\{T_1,T_2,...\}~\forall_i~~T_i:S\to S~\text{bijective}$.
Then assuming that for all pairs of distinct finite sequences of transformations $[p_1,p_2,...p_m]$ and $[q_1, q_2 ... | https://www.lesswrong.com/posts/iiqhik4sZkyAf4mny/the-generalized-sierpinski-mazurkiewicz-theorem |
# Distillation Contest - Results and Recap
This post:
* Announces the winning submissions to the Distillation Contest
* Gives further insight into the scoring process for the contest
* Examines the effectiveness of our advertising strategies
* Gives a brief impact estimate of the contest
* Shares my advice... | https://www.lesswrong.com/posts/RdPdzWkEKH2K5Y8eg/distillation-contest-results-and-recap |
# Bucket Errors
*Author's note: There is a preexisting standalone essay on bucket errors by CFAR cofounder Anna Salamon available* [*here*](https://www.lesswrong.com/posts/EEv9JeuY5xfuDDSgF/flinching-away-from-truth-is-often-about-protecting-the)*. The version in the handbook is similar, but has enough disoverlap that... | https://www.lesswrong.com/posts/WFEt8QHGrnSQzPa9W/bucket-errors |
# Abstracting The Hardness of Alignment: Unbounded Atomic Optimization
*This post is part of the work done at* [*Conjecture*](https://conjecture.dev)*.*
Disagree to Agree
=================
([Practically-A-Book Review: Yudkowsky Contra Ngo On Agents](https://astralcodexten.substack.com/p/practically-a-book-review-yud... | https://www.lesswrong.com/posts/uhxpJyGYQ5FQRvdjY/abstracting-the-hardness-of-alignment-unbounded-atomic |
# Conjecture: Internal Infohazard Policy
*This post benefited from feedback and comments from the whole Conjecture team, as well as others including Steve Byrnes, Paul Christiano, Leo Gao, Evan Hubinger, Daniel Kokotajlo, Vanessa Kosoy, John Wentworth, Eliezer Yudkowsky. Many others also kindly shared their feedback a... | https://www.lesswrong.com/posts/Gs29k3beHiqWFZqnn/conjecture-internal-infohazard-policy |
# Focusing
**Epistemic status**: Firm
*The Focusing technique was developed by Eugene Gendlin as an attempt to answer the question of why some therapeutic patients make significant progress while others do not. Gendlin studied a large number of cases while teasing out the dynamics that became Focusing, and then spent... | https://www.lesswrong.com/posts/f3o9ydY7iPjFF2fyk/focusing-1 |
# Comparing Four Approaches to Inner Alignment
*Early work on this was supported by *[*CEEALAR*](https://ceealar.org/) *and was finished during an internship at*[*Conjecture*](https://www.conjecture.dev/) *under the mentorship of Adam Shimi.*
*Thank you to Evan Hubinger and Rohin Shah for answering some questions rel... | https://www.lesswrong.com/posts/KWmrz9WbGntMGMb73/comparing-four-approaches-to-inner-alignment |
# Humans Reflecting on HRH
**TL;DR**: HRH places a theoretical bound on the best reflection process achievable by us. Notably, this is not necessarily guaranteed to converge to human values, nor is it something that is actually implementable in practice (analogously to HCH). In particular, this is intended to argue ag... | https://www.lesswrong.com/posts/XsYAL4jvzztomSjKy/humans-reflecting-on-hrh |
# Drexler’s Nanotech Forecast
In 1986, Drexler predicted (in [Engines of
Creation](https://en.wikipedia.org/wiki/Engines_of_Creation)) that we\'d
have molecular assemblers in 30 years. They would roughly act as fast,
atomically precise 3-d printers. That was the standard meaning of
nanotech for the next decade, un... | https://www.lesswrong.com/posts/oYks6LXzNfNc7ugyB/drexler-s-nanotech-forecast |
# Translating between Latent Spaces
### *Produced as part of the SERIMATS Program 2022 Research Sprint under John Wentworth*
Introduction
============
The gold-standard of interpretability in ML systems looks like finding embeddings of human-identifiable concepts in neural net architectures, and being able to modif... | https://www.lesswrong.com/posts/JCCktdeHwTosrSgyu/translating-between-latent-spaces |
# How transparency changed over time
*This post is written as part of the SERI MATS program.*
This post explores how model interpretability has changed over time, and if the existing approaches to interpretability are likely to remain viable as the models increase in size and capability. I examine several approaches ... | https://www.lesswrong.com/posts/ngwNHAy5TjStZnJzQ/how-transparency-changed-over-time |
# AGI-level reasoner will appear sooner than an agent; what the humanity will do with this reasoner is critical
The recent advancements in language models, [Gato](https://www.deepmind.com/publications/a-generalist-agent), [Minerva](https://ai.googleblog.com/2022/06/minerva-solving-quantitative-reasoning.html), etc. su... | https://www.lesswrong.com/posts/qhurqnyFbxZY8WC4t/agi-level-reasoner-will-appear-sooner-than-an-agent-what-the |
# chinchilla's wild implications
([Colab notebook](https://colab.research.google.com/drive/1qv2-hUR5hPqw3OcfmLEhq6Tg15bf06Mj?usp=sharing) here.)
$$This post is about language model scaling laws, specifically the laws derived in the DeepMind paper that introduced Chinchilla.[^1trot6ka6e2]
The paper came out a few mon... | https://www.lesswrong.com/posts/6Fpvch8RR29qLEWNH/chinchilla-s-wild-implications |
# Anatomy of a Dating Document
I’ve noticed a trend of writing out dating documents (or web pages, blog posts, etc) as a means of having everything you’d normally put on a dating profile in one place. There’s a particular way many of these are written, such that they’re more straightforward and practical about dating ... | https://www.lesswrong.com/posts/6yiayg5QWtWme4JN8/anatomy-of-a-dating-document |
# Wanted: Notation for credal resilience
_Meta: I've spent 30-60 minutes thinking about this, and asking people who I'd expect to know about existing notation. I don't have scientific training, and I'm not active in the forecasting community._
# Problem
I want a clear shorthand notation for communicating [credal res... | https://www.lesswrong.com/posts/uGqdJCrqznzLBDXcr/wanted-notation-for-credal-resilience |
# Perverse Independence Incentives
In writing about parenting, I don't write enough about what didn't work: this might give the impression that everything I try works, which is not the case! Here's an example of something that didn't, and which I might have caught before trying if I'd thought a bit more.
Brief recap:... | https://www.lesswrong.com/posts/hfik2KmSwdfDegvSH/perverse-independence-incentives |
# On akrasia: starting at the bottom
*\[ Why theorizing about your addictions fails to help you combat them: an addict-thinker's version of "just do it." \]*
* * *
For a few years, I've struggled to overcome several addictions. I'll list them here briefly: (nicotine, nail-biting, overeating, physical inactivity, and... | https://www.lesswrong.com/posts/T3Hii22niGNTKPBpt/on-akrasia-starting-at-the-bottom |
# Astral Codex Ten Berlin Meetup
This meetup is an open get-together for readers of Scott Alexander's blog [Astral Codex Ten](https://astralcodexten.substack.com/). We meet in Tiergarten and sit either on comfy grass or on blankets that you bring with you. If you want to make sure you have a blanket, take agency and ... | https://www.lesswrong.com/events/LbmZmGRsPf3nzeAzC/astral-codex-ten-berlin-meetup-1 |
# Polaris, Five-Second Versions, and Thought Lengths
*Author's note: During CFAR's 4.5d workshops, concepts that had been formalized as "techniques," and which could be described as algorithms and practiced in isolation, generally received 60+ minute sessions. Important concepts which did not have direct practical ap... | https://www.lesswrong.com/posts/afEoXzGB4yt4Hk8zy/polaris-five-second-versions-and-thought-lengths |
# Meditation course claims 65% enlightenment rate: my review
TL;DR
=====
* I eliminated my impostor syndrome and dramatically reduced my work-related anxiety. I did this in a way that I think can be replicated.
* Different techniques work for different people. If you want to get the benefits of meditation, you s... | https://www.lesswrong.com/posts/PGv9THs68ArPur7yP/meditation-course-claims-65-enlightenment-rate-my-review |
# Is there any writing about prompt engineering for humans?
In much the same way that we need to prompt engineer language models to get them to handle our questions correctly, we know that humans respond differently to "This intervention will save 200 lives" and "This intervention will result in 400 deaths" when talki... | https://www.lesswrong.com/posts/YKygo9kHHv42LmRph/is-there-any-writing-about-prompt-engineering-for-humans |
# One-day applied rationality workshop with Duncan Sabien
I invited Duncan Sabien to give a workshop the day after LWCW (and I'm doing the organizational background work).
If you want to read more and sign up, the description is at https://www.lesswrong.com/posts/pEFzqeTCBfo4nrN47/one-day-applied-rationality-workshop... | https://www.lesswrong.com/events/mh72WDEpGenehHEYv/one-day-applied-rationality-workshop-with-duncan-sabien |
# Letter from leading Soviet Academicians to party and government leaders of the Soviet Union regarding signs of decline and structural problems of the economic-political system (1970)
Crossposted from my blog: [mzuo.ca](https://mzuo.ca/2022/08/01/letter-from-leading-soviet-academicians-to-party-and-government-leaders... | https://www.lesswrong.com/posts/9rbTvbdg8HZadAqoH/letter-from-leading-soviet-academicians-to-party-and |
# Turbocharging
**Epistemic status**: Mixed
*The concepts underlying the Turbocharging model (such as classical and operant conditioning, neural nets, and distinctions between procedural and declarative knowl- edge) are all well-established and well-understood, and the "further resources" section for this class is on... | https://www.lesswrong.com/posts/3Jggtk5otfDoL7Lpq/turbocharging |
# I want to donate some money (not much, just what I can afford) to AGI Alignment research, to whatever organization has the best chance of making sure that AGI goes well and doesn't kill us all. What are my best options, where can I make the most difference per dollar?
I don't understand this field well, I'm hoping y... | https://www.lesswrong.com/posts/SuwaxLK8kBwtQ6x4S/i-want-to-donate-some-money-not-much-just-what-i-can-afford |
# (Summary) Sequence Highlights - Thinking Better on Purpose
Rationality is a very useful thing to learn, but while there is [good curated reading material](https://www.lesswrong.com/library), it's not always easy reading. More than once I've wanted to introduce someone to the topics, but couldn't hope for them to dig... | https://www.lesswrong.com/posts/NZEyHpGGCG2cZnyqf/summary-sequence-highlights-thinking-better-on-purpose |
# Againstness
**Epistemic status**: Mixed
*The concepts underlying the Againstness model (such as the division of the autonomic nervous system into the sympathetic and parasympathetic subsystems, or the bidirectional relationship between physiology and stress response) are all well-established and well-understood. Th... | https://www.lesswrong.com/posts/QnZRvud7fXfpdxBY3/againstness |
# What are the Red Flags for Neural Network Suffering? - Seeds of Science call for reviewers
[Seeds of Science](https://www.theseedsofscience.org/) is a new journal (funded through Scott Alexander's [ACX grants program](https://astralcodexten.substack.com/p/acx-grants-results)) that publishes speculative or non-tradit... | https://www.lesswrong.com/posts/h28jWPYxoB4iAQtum/what-are-the-red-flags-for-neural-network-suffering-seeds-of |
# Two-year update on my personal AI timelines
I worked on my [draft report on biological anchors for forecasting AI timelines](https://www.lesswrong.com/posts/KrJfoZzpSDpnrv9va/draft-report-on-ai-timelines) mainly between ~May 2019 (three months after the release of GPT-2) and ~Jul 2020 (a month after the release of G... | https://www.lesswrong.com/posts/AfH2oPHCApdKicM4m/two-year-update-on-my-personal-ai-timelines |
# Externalized reasoning oversight: a research direction for language model alignment
Summary
-------
In this post, I’ll present a research direction that I’m interested in for alignment of pretrained language models.
TL;DR: **Force a language model to think out loud, and use the reasoning itself as a channel for ov... | https://www.lesswrong.com/posts/FRRb6Gqem8k69ocbi/externalized-reasoning-oversight-a-research-direction-for |
# Announcing Squiggle: Early Access
*Update: We've announced* [*a $1,000 prize for early experimentation with Squiggle*](https://forum.effectivealtruism.org/posts/ZrWuy2oAxa6Yh3eAw/usd1-000-squiggle-expe... | https://www.lesswrong.com/posts/mv6nuN7nhnNpj9MLr/announcing-squiggle-early-access |
# Transformer language models are doing something more general
*Epistemic status: speculative.*
First, a few paper titles:
* [Pretrained Transformers as Universal Computation Engines](https://arxiv.org/abs/2103.05247)
* [Can Wikipedia Help Offline Reinforcement Learning?](https://arxiv.org/abs/2201.12122) (Answe... | https://www.lesswrong.com/posts/YwqSijHybF9GFkDab/transformer-language-models-are-doing-something-more-general |
# Precursor checking for deceptive alignment
This post is primarily an excerpt from “[Acceptability Verification: a Research Agenda](https://www.alignmentforum.org/posts/GeabLEXYP7oBMivmF/acceptability-verification-a-research-agenda)” that I think is useful enough on its own such that I’ve spun it off into its own pos... | https://www.lesswrong.com/posts/pRt4E3nmPBtWZiT4A/precursor-checking-for-deceptive-alignment |
# Three pillars for avoiding AGI catastrophe: Technical alignment, deployment decisions, and coordination
Generated by Midjourney
Cross-posted to the EA Forum: [Three pillars for avoiding AGI catastroph... | https://www.lesswrong.com/posts/cm5dCKYCamotzEMxq/three-pillars-for-avoiding-agi-catastrophe-technical |
# Clapping Lower
Clapping is normally relatively high pitch:
([mp3](https://www.jefftk.com/three-high-claps.mp3))
[](https://www.jefftk.com/high-clap-spectrum-big.png)
I find clapping annoyingly high, so I usually clap a bit differently:
([mp3](https://www.jefftk.... | https://www.lesswrong.com/posts/aQurKGKyCaFGZozLM/clapping-lower |
# Surprised by ELK report's counterexample to Debate, IDA
Summary
-------
I had assumed the original ELK report had fundamental objections to Debate and IDA in terms of their robustness. Re-reading the report, I was surprised to find that that the only counterexample it provides for these proposals is that they don't... | https://www.lesswrong.com/posts/Qsc3G2HemFWLobDSw/surprised-by-elk-report-s-counterexample-to-debate-ida |
# High Reliability Orgs, and AI Companies
[*Epistemic Effort*](https://www.lesswrong.com/posts/oDy27zfRf8uAbJR6M/epistemic-effort)*: Rough notes from a shallow dive. Looked into this for a few hours. I didn't find a strong takeaway but think this is probably a useful jumping-off-point for further work.*
Most likely, ... | https://www.lesswrong.com/posts/FBoyR2rt29oYvazsE/high-reliability-orgs-and-ai-companies |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.