text stringlengths 300 320k | source stringlengths 52 154 |
|---|---|
# Should LW have an official list of norms?
*To get this written and shared quickly, I haven't polished it much and the English/explanation is a little rough. Seemed like the right tradeoff though.*
Recently, a few users have written their sense of norms for rationalist discourse, i.e. [Basics of Rationalist Discours... | https://www.lesswrong.com/posts/gugkWsfayJZnicAew/should-lw-have-an-official-list-of-norms |
# How Many Bits Of Optimization Can One Bit Of Observation Unlock?
So there’s this thing where a system can perform more bits of optimization on its environment by observing some bits of information from its environment. Conjecture: observing an additional $N$ bits of information can allow a system to perform *at most... | https://www.lesswrong.com/posts/6rZroG3mztowktJzp/how-many-bits-of-optimization-can-one-bit-of-observation |
# Archetypal Transfer Learning: a Proposed Alignment Solution that solves the Inner & Outer Alignment Problem while adding Corrigible Traits to GPT-2-medium
**NOTE: An updated discussion on what Archetypal Transfer Learning is all about can be found in this** [**LessWrong post.**](https://www.lesswrong.com/posts/DMtzw... | https://www.lesswrong.com/posts/pu6D2EdJiz2mmhxfB/archetypal-transfer-learning-a-proposed-alignment-solution |
# A simple presentation of AI risk arguments
[Here is a draft of an accordion-style AGI risk FAQ.](https://workflowy.com/s/faq-why-would-ai-be/2kHollRW8uqinUsV)
I'm interested in three types of feedback:
1. The best place to host a similar format
2. Feedback on content
3. Thoughts about this general approach
Th... | https://www.lesswrong.com/posts/mJqabqwAb3QzZcu9T/a-simple-presentation-of-ai-risk-arguments |
# Transcript and Brief Response to Twitter Conversation between Yann LeCunn and Eliezer Yudkowsky
Yann LeCun is Chief AI Scientist at Meta.
This week, Yann engaged with Eliezer Yudkowsky on Twitter, [doubling down on Yann’s position that it is dangerously irresponsible to talk about smarter-than-human AI as an existe... | https://www.lesswrong.com/posts/tcEFh3vPS6zEANTFZ/transcript-and-brief-response-to-twitter-conversation |
# My version of Simulacra Levels
> People act as if there are four truth-values: *True, False, Cringe,* and *Based.*
--[David Udell (paraphrase)](https://www.lesswrong.com/posts/KGLJeeF6psTzHiuiR/david-udell-s-shortform?commentId=mFGFuDbAhgkLArdqa)
This post lays out my own version of the [Simulacra Levels idea](htt... | https://www.lesswrong.com/posts/uS2RH8c7xT8DiSzaD/my-version-of-simulacra-levels |
# $250 prize for checking Jake Cannell's Brain Efficiency
This is to announce a $250 prize for [spotchecking](https://www.lesswrong.com/tag/epistemic-spot-check#:~:text=Epistemic%20Spot%20Checks%20is%20a,has%20to%20give%20the%20author.) or otherwise indepth reviewing Jacob Cannell's technical claims concerning thermod... | https://www.lesswrong.com/posts/Xo4cqLrAKjdNzpRSa/usd250-prize-for-checking-jake-cannell-s-brain-efficiency-1 |
# I was Wrong, Simulator Theory is Real
*\[Epistemic Status: Excitedly writing up my new thoughts. I literally just fixed one mistake, so its possible there are others. Not a finalized research product.\]*
**Overview**
------------
Fixing a small bug in my [recent study](https://aizi.substack.com/p/research-report-i... | https://www.lesswrong.com/posts/mfn32QHwKb55afHq4/i-was-wrong-simulator-theory-is-real |
# Chad Jones paper modeling AI and x-risk vs. growth
Looks like this Chad Jones paper was [just posted today](https://twitter.com/ChadJonesEcon/status/1651219721205977088). Abstract:
> Advances in artificial intelligence (A.I.) are a double-edged sword. On the one hand, they may increase economic growth as A.I. augme... | https://www.lesswrong.com/posts/EhHdZ5yBgEvLLx6Pw/chad-jones-paper-modeling-ai-and-x-risk-vs-growth |
# Infra-Bayesianism naturally leads to the monotonicity principle, and I think this is a problem
**Introduction**
----------------
The [monotonicity principle](https://www.youtube.com/watch?v=kmPFjpEibu0) is a famously uncomfortable consequence of [Infra-Bayesian Physicalism](https://www.lesswrong.com/posts/gHgs2e2J5... | https://www.lesswrong.com/posts/yykNvq257zBLDNmJo/infra-bayesianism-naturally-leads-to-the-monotonicity |
# Recent Database Migration - Report Bugs
Hey all,
LessWrong is currently in the process of migrating databases (from mongo to postgres). Today we migrated a large number of important collections, including Posts, Comments, Tags, Votes, and others related to those. If you notice any **new bugs** related to posts, c... | https://www.lesswrong.com/posts/WSTB698bTdnK8Yrgn/recent-database-migration-report-bugs |
# LM Situational Awareness, Evaluation Proposal: Violating Imitation
Motivation
----------
“[Playing the training game](https://www.lesswrong.com/posts/pRkFkzwKZ2zfa3R6H/without-specific-countermeasures-the-easiest-path-to#While_humans_are_in_control__Alex_would_be_incentivized_to__play_the_training_game_)” most like... | https://www.lesswrong.com/posts/kkaBC9Epydj3m6ZsA/lm-situational-awareness-evaluation-proposal-violating |
# A very non-technical explanation of the basics of infra-Bayesianism
**Introduction**
================
As a response to John Wentworth's [public request](https://www.lesswrong.com/posts/ig62gqK2xRPv79Wsr/usd500-bounty-contest-explain-infra-bayes-in-the-language-of), I try to explain the basic structure of infra-Baye... | https://www.lesswrong.com/posts/NdJsWDS7Aq4xqoumk/a-very-non-technical-explanation-of-the-basics-of-infra |
# Contra Yudkowsky on Doom from Foom #2
*This is a follow up and partial rewrite to/of an [earlier part #1 post](https://www.lesswrong.com/posts/Lwy7XKsDEEkjskZ77/contra-yudkowsky-on-ai-doom) critiquing EY's specific argument for doom from AI go foom, and a partial clarifying response to [DaemonicSigil's reply](https:... | https://www.lesswrong.com/posts/LF3DDZ67knxuyadbm/contra-yudkowsky-on-doom-from-foom-2 |
# Automating the Breath Pulse
One of the pieces of my [rhythm stage setup](https://www.jefftk.com/news/jammer) is a breath controller, which measures how much I blow. It's a good fit for anything where I want to have some flexible imprecise control, but one thing I've ended up using it for a lot is just a pulse:
Here... | https://www.lesswrong.com/posts/JgLgDqQdBaRNGhhk2/automating-the-breath-pulse |
# AI chatbots don't know why they did it
*\[Epistemic status: a basic explanation, high confidence.\]*
Chatting with an AI seems like chatting with a person, so it’s natural to ask an AI chatbot to explain its answers. But it doesn’t work how you’d expect. An AI chatbot doesn't know how it decided what to write, ... | https://www.lesswrong.com/posts/sneGMrPkQhBzkJFJd/ai-chatbots-don-t-know-why-they-did-it |
# Romance, misunderstanding, social stances, and the human LLM
1\. Cross-sex friendships
=========================
I saw [a tweet](https://twitter.com/uncatherio/status/1649182838754009089) recently that was talking about one of the possible conditions where (heterosexual) men and women have an easy time being “just ... | https://www.lesswrong.com/posts/mhjAwsxTMmqFNbKLQ/romance-misunderstanding-social-stances-and-the-human-llm |
# AI doom from an LLM-plateau-ist perspective
*(in the form of an FAQ)*
Q: What do you mean, “LLM plateau-ist”?
=======================================
**A:** As background, I think it’s obvious that there will *eventually* be **“transformative AI” (TAI)** that would radically change the world.[^76tsq17s1nk]
I’m i... | https://www.lesswrong.com/posts/KJRBb43nDxk6mwLcR/ai-doom-from-an-llm-plateau-ist-perspective |
# AI #9: The Merge and the Million Tokens
There were two big developments this week.
One is that Google merged Google Brain and DeepMind into the new Google DeepMind. DeepMind head and founder Demis Hassabis is in charge of the new division. We will see how this plays out in practice, seems very good for Google.
The... | https://www.lesswrong.com/posts/qtEgaxSFxYanT5QbJ/ai-9-the-merge-and-the-million-tokens |
# Interview with Paul Christiano: How We Prevent the AI’s from Killing us
[0:00](https://www.youtube.com/watch?v=GyFkWb903aU&t=0s) Intro
[9:20](https://www.youtube.com/watch?v=GyFkWb903aU&t=560s) Percentage Likelihood of Death by AI
[11:24](https://www.youtube.com/watch?v=GyFkWb903aU&t=684s) Timing
[19:15](http... | https://www.lesswrong.com/posts/vFiZCrt5cp2ic9ois/interview-with-paul-christiano-how-we-prevent-the-ai-s-from |
# Second-Level Empiricism: Reframing the Two-Child Puzzle
Empiricism is the belief that knowledge derives from sensory experience, or, by some definitions, experiment. However, empiricism itself can show that it is not the best tool for a given scenario. This article provides an example of when this happened to me.
*... | https://www.lesswrong.com/posts/cd868irGDsEqyDySk/second-level-empiricism-reframing-the-two-child-puzzle |
# My views on “doom”
I’m often asked: “what’s the probability of a really bad outcome from AI?”
There are many different versions of that question with different answers. In this post I’ll try to answer a bunch of versions of this question all in one place.
#### Two distinctions
Two distinctions often lead to confu... | https://www.lesswrong.com/posts/xWMqsvHapP3nwdSW8/my-views-on-doom |
# Infrafunctions and Robust Optimization
$\newcommand{\bol}[1]{\textbf{#1}}$
[Proofs are in this link](https://www.lesswrong.com/posts/NsThHtSMDvah6p5oB/infrafunctions-proofs)
This will be a fairly important post. Not one of those obscure result-packed posts, but something a bit more fundamental that I hope to refer... | https://www.lesswrong.com/posts/d96dDEYMfnN2St3Bj/infrafunctions-and-robust-optimization |
# Infrafunctions Proofs
$$\newcommand{\bol}[1]{\textbf{#1}}$$
**Proposition 1:**
*Given some compact metric space of options $X$, if $U:X\to\mathbb{R}$ is a bounded function, $\{\mu|\forall\nu\in\Delta X:\mu(U)\ge\nu(U)\}=\Delta\{x|\forall y\in X:U(x)\ge U(y)\}$*
PROOF: To do this, we must show that anything in the... | https://www.lesswrong.com/posts/NsThHtSMDvah6p5oB/infrafunctions-proofs |
# The LW crossroads of purpose
Hello everyone reading,
I'm writing this because I do not believe fixing peripheral things on LW is enough. And even though I wrote about it in a [comment](https://www.lesswrong.com/posts/gugkWsfayJZnicAew/?commentId=J99Q6mK3G29Bnodce), I wanted to elaborate and highlight the issue... | https://www.lesswrong.com/posts/duT32SKT6FshrNgH2/the-lw-crossroads-of-purpose |
# Tuning your Cognitive Strategies
*The blogpost author (*[*SquirrelInHell*](https://www.lesswrong.com/users/squirrelinhell) *on LessWrong) died awhile ago. I'm not sure who's currently paying for their website or how long it'll be up. I don't have the rights to this, but decided it was important enough to have on Les... | https://www.lesswrong.com/posts/bbB4pvAQdpGrgGvXH/tuning-your-cognitive-strategies |
# Simultaneous Footbass and Footdrums
For the past year or so I've had two options for my feet with my [rhythm stage setup](https://www.jefftk.com/news/jammer). I can play bass:
Or I can play drums:
But when I play mandolin I want both!
I tried a system where I play bass but whichever drum pedal I tap triggers a ki... | https://www.lesswrong.com/posts/KADwbpuS7zwp2dy5G/simultaneous-footbass-and-footdrums |
# Podcast with Divia Eden and Ronny Fernandez on the strong orthogonality thesis
On my side podcast, ["The Filan Cabinet"](https://thefilancabinet.com/), I invited Ronny Fernandez and Divia Eden to talk about the strong orthogonality thesis, and whether it's true. Seems like people here might also be interested. Podca... | https://www.lesswrong.com/posts/3ocLFajphtyHTgNxx/podcast-with-divia-eden-and-ronny-fernandez-on-the-strong |
# Dropout can create a privileged basis in the ReLU output model.
# Abstract
This post summarises some interesting results on dropout in the ReLU output model. The ReLU output model, introduced [here](https://transformer-circuits.pub/2022/toy_model/index.html) is a minimal toy model of a neural network which can e... | https://www.lesswrong.com/posts/uSdFFTATPFJz4pQyB/dropout-can-create-a-privileged-basis-in-the-relu-output |
# AI safety logo design contest, due mid-May
Please have a look at the updated version of this post on the EA Forum.
This contest is about creating a logo to represent AI safety research.
I’m being supported by an independent funder and people from [nonlinear.org](https://www.nonlinear.org/) and am taking care of a... | https://www.lesswrong.com/posts/zBxzQEHCsfK5diCgz/ai-safety-logo-design-contest-due-mid-may |
# cyberpunk raccoons
There are many stories about supervillains that have a secret base that manufactures something and is mostly self-contained, which would require an entire parallel economy. Sometimes, like in James Bond stories, there are human workers living there. The trend now is towards having such secret base... | https://www.lesswrong.com/posts/aojgYqb3gZhCPTE3B/cyberpunk-raccoons |
# A Clearer Thinking tool that teaches you to use Internal Family Systems concepts
One of our recently-released Clearer Thinking tools, [Have Better Conversations with Yourself](https://programs.clearerthinking.org/internal_family_systems.html), walks you through key concepts from Internal Family Systems (IFS). While ... | https://www.lesswrong.com/posts/pJpZ62v8fWXPmeymq/a-clearer-thinking-tool-that-teaches-you-to-use-internal |
# The Social Alignment Problem
*TLDR: I think public outreach is a very hopeful path to victory. More importantly, extended large-scale conversation around questions such as whether public outreach is a hopeful path to victory would be very likely to decrease p(doom).*
You’re a genius mechanical engineer in prison. Y... | https://www.lesswrong.com/posts/ecuAGrZGtavgsPs4w/the-social-alignment-problem |
# [Linkpost] Sam Altman's 2015 Blog Posts Machine Intelligence Parts 1 & 2
I'm often surprised more people haven't read Open AI CEO Sam Altman's 2015 blog posts [Machine Intelligence Part 1](https://blog.samaltman.com/machine-intelligence-part-1) & [Part 2](https://blog.samaltman.com/machine-intelligence-part-2). In m... | https://www.lesswrong.com/posts/QnBZkNJNbJK9k5Xi7/linkpost-sam-altman-s-2015-blog-posts-machine-intelligence |
# LLMs and computation complexity
*Epistemic status: Speculative. I've built many large AI systems in my previous HFT career but have never worked with generative AIs. I am leveling up in LLMs by working things out from base principles and observations. All feedback is very welcome.*
Tl;dr: An LLM cannot solve comput... | https://www.lesswrong.com/posts/XNBZPbxyYhmoqD87F/llms-and-computation-complexity |
# "notkilleveryoneism" sounds dumb
"AI safety" and "AI alignment" now often mean "making AI not say racist/illegal/etc things". The term "AI notkilleveryoneism" is now sometimes being used as a replacement that refers specifically to ASI-related risks.
I think "notkilleveryoneism" is the wrong term to use, for reason... | https://www.lesswrong.com/posts/jMzBhCRrr7otmqcvK/notkilleveryoneism-sounds-dumb |
# Locating Fulcrum Experiences
*Note: There are two appendices at the end of this post. The first contains a summary of the steps I’ve described here. The second is a glossary of key terms I’ve introduced in this essay.*
Once you’ve booted up some curiosity and original seeing—perhaps even identified a question that’... | https://www.lesswrong.com/posts/Gs6oMWEFsL4gGocBf/locating-fulcrum-experiences |
# [SEE NEW EDITS] No, *You* Need to Write Clearer
**This post is aimed solely at people in AI alignment/safety.**
*EDIT 3 October 2023: This post did not even mention, let alone account for, how somebody should post [half-baked/imperfect/hard-to-describe](https://www.lesswrong.com/posts/f73Z3Qzo67mgvXCbw/half-baked-p... | https://www.lesswrong.com/posts/mLubC65xXekk5tkug/see-new-edits-no-you-need-to-write-clearer |
# One bit of observation can unlock many of optimization - but at what cost?
[This question](https://www.lesswrong.com/posts/6rZroG3mztowktJzp/) by johnswentworth nerd sniped me, so I ended up thinking a lot about the relationship between information and control over the world in the simplified scenario of a single ta... | https://www.lesswrong.com/posts/dfTm26pvq7yQp8mR3/one-bit-of-observation-can-unlock-many-of-optimization-but |
# Research agenda: Supervising AIs improving AIs
\[This post summarizes some of the work done by Owen Dudney, Roman Engeler and myself (Quintin Pope) as part of the SERI MATS shard theory stream.\]
TL;DR
=====
Future prosaic AIs will likely shape their own development or that of successor AIs. We're trying to make s... | https://www.lesswrong.com/posts/7e5tyFnpzGCdfT4mR/research-agenda-supervising-ais-improving-ais |
# Will GPT-5 be able to self-improve?
I want to try to go over some of the objections imagine people are having. I don't think I fully understand the opposing viewpoint here, so hopefully it will be clarified for me in the comments.
1\. LLMs are not truly generating novel ideas, they are just interpolating between ex... | https://www.lesswrong.com/posts/zwAHF5tmFDTDD6ZoY/will-gpt-5-be-able-to-self-improve |
# Which technologies are stuck on initial adoption?
Video calls have been with us for a while. Except, they were rarely used. IME, people sometimes had Skype calls with relatives abroad and that's about it. And then, COVID happened. Suddenly, Zoom skyrocketed, with Google Meet not far behind. The reason is obvious.
... | https://www.lesswrong.com/posts/fZPwyW2rkZzZBryuR/which-technologies-are-stuck-on-initial-adoption |
# Horizontal vs vertical generality
*This is a note/definition I want to put out there. While the term may sound similar to breadth vs depth of knowledge, it is importantly distinct.*
Let's define intelligence as the ability to process information and make decisions. (This is a very broad definition and includes not ... | https://www.lesswrong.com/posts/Kratt3f8KZsuzTeJG/horizontal-vs-vertical-generality |
# Biological DOOM: a brief overview of biological computation
*(no, not* [*that kind of biological doom*](https://www.nti.org/about/programs-projects/project/global-catastrophic-biological-risks/)*)*
[DOOM](https://en.wikipedia.org/wiki/Doom_(1993_video_game)) is a classic first-person shooter game released in 1993 b... | https://www.lesswrong.com/posts/guxDHdwaRt9AbZAKx/biological-doom-a-brief-overview-of-biological-computation |
# Accuracy of arguments that are seen as ridiculous and intuitively false but don't have good counter-arguments
I think one of the biggest reasons I am worried about AI doom is that there doesn't seem to be very good counter-arguments. Most of them are just bad, and the ones that aren't a bit vague (usually something ... | https://www.lesswrong.com/posts/M37dXLYJo3kckry8Y/accuracy-of-arguments-that-are-seen-as-ridiculous-and |
# Forum Proposal: Karma Transfers
On LessWrong, we can weak- and strong-upvote posts, which adds to the other user's karma without impacting our own. As we gain karma, we are able to award more karma. I have nearly 10,000 LessWrong Karma, and my strong upvote awards 8 karma at no cost to me.
Many users offer monetary... | https://www.lesswrong.com/posts/bfFjLF9LrFLqMq9ax/forum-proposal-karma-transfers |
# The voyage of novelty
*[Metadata: crossposted from [https://tsvibt.blogspot.com/2023/01/the-voyage-of-novelty.html](https://tsvibt.blogspot.com/2023/01/the-voyage-of-novelty.html). First completed January 17, 2023. This essay is more like research notes than exposition, so context may be missing, the use of terms ma... | https://www.lesswrong.com/posts/aFZju8Kh4MWJaChav/the-voyage-of-novelty |
# Connectomics seems great from an AI x-risk perspective
Context
=======
Numerous people are in a position to accelerate certain areas within science or technology, whether by directing funds and resources, or by working in the area directly. But which areas are best to accelerate?
One possible consideration (among ... | https://www.lesswrong.com/posts/ybmDkJAj3rdrrauuu/connectomics-seems-great-from-an-ai-x-risk-perspective |
# Simulators Increase the Likelihood of Alignment by Default
Alignment by Default is the idea that achieving alignment in artificial general intelligence (AGI) may be more straightforward than initially anticipated. When an AI possesses a comprehensive and detailed world model, it inherently represents human values wi... | https://www.lesswrong.com/posts/yzGPLdEqa2ytT7MY2/simulators-increase-the-likelihood-of-alignment-by-default |
# Fundamental Uncertainty: Chapter 7 - Why is truth useful?
*N.B. This is a chapter in a book about truth and knowledge. It is the first draft. I have since revised it. You can find the most up-to-date info/version on the* [*book's website*](https://www.fundamentaluncertainty.com/)*.*
Last chapter we proved that our ... | https://www.lesswrong.com/posts/Q4zBhYobwkGBGuh7v/fundamental-uncertainty-chapter-7-why-is-truth-useful |
# Vaccine Policies Need Updating
It's [reasonably common](https://www.jefftk.com/p/boston-social-dance-covid-requirements) for events like social dances to require vaccination. This usually looks like requiring attendees be "fully vaccinated", "up to date", or "fully vaccinated and boosted", and cashes out to the peop... | https://www.lesswrong.com/posts/LuDH8ALr4dWczgZmW/vaccine-policies-need-updating |
# Discussion about AI Safety funding (FB transcript)
Kat Woods recently wrote a [Facebook post](https://www.facebook.com/katxiowoods/posts/pfbid0ngoH3xG2tZAQyoGsCK9oNcDJNPbmmD7VX4BqrPB5GyN2UnbW8dck4xVwkXsDKXgHl) about [Nonlinear's new funding program](https://forum.effectivealtruism.org/posts/Qoecey2umNjcqEGHP/apply-t... | https://www.lesswrong.com/posts/FdHRkGziQviJ3t8rQ/discussion-about-ais-funding-fb-transcript |
# A small update to the Sparse Coding interim research report
This is a linkpost to a [**set of slides**](https://docs.google.com/presentation/d/1lP3u8NupGyWtrECoxr4zroemUZpCDd5k3ghuBbORWHU/edit?usp=sharing) containing an update to a project that was the subject of a previous post ([\[Interim research report\] Taking... | https://www.lesswrong.com/posts/DezghAd4bdxivEknM/a-small-update-to-the-sparse-coding-interim-research-report |
# How can one rationally have very high or very low probabilities of extinction in a pre-paradigmatic field?
It is generally accepted in the local AI alignment circles that the whole field is pre-paradigmatic, in the Kuhnian sense (phase 1, as summarized [here](https://www.hoddereducation.co.uk/media/Documents/magazin... | https://www.lesswrong.com/posts/R6kGYF7oifPzo6TGu/how-can-one-rationally-have-very-high-or-very-low |
# neuron spike computational capacity
The following is taken from [this blog post](https://bhauth.com/blog/biology/neuron%20networks.html). I'm posting it here now because of [this argument](https://www.lesswrong.com/posts/Xo4cqLrAKjdNzpRSa/usd250-prize-for-checking-jake-cannell-s-brain-efficiency-1?commentId=Q9qh6JJw... | https://www.lesswrong.com/posts/dTpKX5DdygenEcMjp/neuron-spike-computational-capacity |
# Safety standards: a framework for AI regulation
*Purpose: provide a public resource on safety standards that introduces the ideas I've heard people talk about and some new ideas. I don't know which ideas are new because the discourse has been largely non-public. I think safety standards provide a clear regulatory di... | https://www.lesswrong.com/posts/j2LD87wT3dpr4m7rs/safety-standards-a-framework-for-ai-regulation |
# Hell is Game Theory Folk Theorems
*[content warning: simulated very hot places; extremely bad Nash equilibria]*
*(based on [a Twitter thread](https://twitter.com/jessi_cata/status/1652433243751485441))*
Rowan: "If we succeed in making aligned AGI, we should punish those who committed cosmic crimes that decreased t... | https://www.lesswrong.com/posts/d2HvpKWQ2XGNsHr8s/hell-is-game-theory-folk-theorems |
# LessWrong Community Weekend 2023 [Applications now closed]
***Update:** \[Applications now closed\]*
September 22-25th is the 10th annual LessWrong Community Weekend (LWCW). This is Europe's largest r... | https://www.lesswrong.com/posts/f2vFYmwFGu3HLL9QB/lesswrong-community-weekend-2023-applications-now-closed-1 |
# LessWrong Community Weekend 2023 [Applications now closed]
**When: **Friday 22nd September - Monday 25th September 2023
**Where:** [jh-wannsee.de](http://jh-wannsee.de/) (Berlin)
**Prices: **
* Regular ticket: €200
* Supporter ticket: €250/300/400
**Apply here: **\[Applications now closed\]
* * *
, Robin Hanson writes
> There have been over 100K UFO sightings reported worldwide since 1940. Roughly 5% or so are “strong” events, which seem rather hard to explain due to either many witnesses, especial... | https://www.lesswrong.com/posts/kD5zEeK9ihQXq2KTL/where-is-all-this-evidence-of-ufos |
# The Rocket Alignment Problem, Part 2
Previously (Eliezer Yudkowsky): [The Rocket Alignment Problem](https://intelligence.org/2018/10/03/rocket-alignment/).
Recently we had a failure to launch, and a failure to communicate around that failure to launch. This post explores that failure to communicate, and the attempt... | https://www.lesswrong.com/posts/9XdoF8AjKwEDZbC8x/the-rocket-alignment-problem-part-2 |
# What Boston Can Teach Us About What a Woman Is
*"It's all about that semantic space map."*
*\[Originally posted* [*7/14/2022 on Substack*](https://ymeskhout.substack.com/p/what-boston-can-teach-us-about-what)*.\]*
One of the coolest illustrations of societal consensus is this cartography research project from a fe... | https://www.lesswrong.com/posts/uBjKXbTwumJ3ruR5b/what-boston-can-teach-us-about-what-a-woman-is |
# Distinguishing misuse is difficult and uncomfortable
USPS delivers ibuprofen and vicodin just the same because it doesn't care enough to open and test every bottle of pills. If they did carefully inspect the contents of every package, they still couldn't distinguish other "shipping misuse" — money for your granddaug... | https://www.lesswrong.com/posts/cCc5vkLCeDukRHLsJ/distinguishing-misuse-is-difficult-and-uncomfortable |
# Shah (DeepMind) and Leahy (Conjecture) Discuss Alignment Cruxes
Preface
=======
In December 2022, Rohin Shah (DeepMind) and Connor Leahy (Conjecture) discussed why Leahy is pessimistic about AI risk, and Shah is less so. Below is a summary and transcript.
Summary
=======
Leahy expects discontinuities - capabiliti... | https://www.lesswrong.com/posts/cnn3kkC6kDqRkLe7W/shah-deepmind-and-leahy-conjecture-discuss-alignment-cruxes |
# In favor of steelmanning
It's different than passing someone's Intellectual Turing Test
*Epistemic status: Others have found my take here useful, but I haven’t engaged with the entire debate on the subject. Fairly confident that it’s useful to me, though see caveat.*
Eliezer Yudkowsky recently disclaimed steelmann... | https://www.lesswrong.com/posts/wzEMZsxRdscFCj8be/in-favor-of-steelmanning |
# The Apprentice Thread 2
A while back, [@Zvi](https://www.lesswrong.com/users/zvi?mention=user) [set up a Schelling point for mentorship](https://www.lesswrong.com/posts/scNCmwaduCgJmCBYh/the-apprentice-thread). I, personally, had some pretty good experiences from that thread, and wanted to continue the tradition, n... | https://www.lesswrong.com/posts/ztwDeXLZaquKNdf7y/the-apprentice-thread-2 |
# Geoff Hinton Quits Google
The NYTimes reports that Geoff Hinton has quit his role at Google:
> On Monday, however, he officially joined a growing chorus of critics who say those companies are racing toward danger with their aggressive campaign to create products based on generative artificial intelligence, the tech... | https://www.lesswrong.com/posts/ri5qQoAiDGwckcZGh/geoff-hinton-quits-google |
# What 2025 looks like
**I wrote almost all of this in mid-March before the FLI Open Letter and Eliezer's TIME piece. Weirdly, after just six weeks I'd likely write something different. This isn't as finished/polished as I'd like, but better to ship it as is than languish incomplete forever.**
* * *
Not quite two ye... | https://www.lesswrong.com/posts/JeMGZNZ6tuBWJHqvi/what-2025-looks-like |
# Getting Your Eyes On
The first phase of naturalism is about identifying crucial experiences that you want to study in greater depth. It’s a lot like preparing slides for observation under a microscope.
This second phase, “Getting Your Eyes On”, is about heightening your awareness of those crucial experiences. It’s ... | https://www.lesswrong.com/posts/pMduTNnhamWYn7KzC/getting-your-eyes-on |
# AXRP Episode 21 - Interpretability for Engineers with Stephen Casper
[YouTube link](https://www.youtube.com/watch?v=lmqfyYn_WJw&list=PLmjaTS1-AiDeqUuaJjasfrM6fjSszm9pK&index=21)
Lots of people in the field of machine learning study ‘interpretability’, developing tools that they say give us useful information about ... | https://www.lesswrong.com/posts/9mzRsm5GyMYwvGSZi/axrp-episode-21-interpretability-for-engineers-with-stephen |
# Summaries of top forum posts (24th - 30th April 2023)
We've just passed the half year mark for this project! If you're reading this, **please consider taking** [**this 5 minute survey**](https://forms.gle/F1URnwsfKqPoTDJt7) \- all questions optional. If you listen to the podcast version, we have a separate survey fo... | https://www.lesswrong.com/posts/qrpwqSmHq8shX4Rg6/summaries-of-top-forum-posts-24th-30th-april-2023 |
# Open & Welcome Thread - May 2023
If it’s worth saying, but not worth its own post, here's a place to put it.
If you are new to LessWrong, here's the place to introduce yourself. Personal stories, anecdotes, or just general comments on how you found us and what you hope to get from the site and community are invited... | https://www.lesswrong.com/posts/aPzhXZw7Zbm7PqXkZ/open-and-welcome-thread-may-2023 |
# Technological unemployment as another test for rationalist winning
Ultimately, [rationalism should help people win](https://www.lesswrong.com/tag/rationality-is-systematized-winning). Scott Alexander [claims](https://www.lesswrong.com/posts/MajyZJrsf8fAywWgY/a-lesswrong-crypto-autopsy) that the surge of the price of... | https://www.lesswrong.com/posts/KAK7wpsBb6apSDmEg/technological-unemployment-as-another-test-for-rationalist |
# Some Thoughts on Virtue Ethics for AIs
This post argues for the desirability and plausibility of AI agents whose values have a structure I call ‘praxis-based.’ The idea draws on various aspects of virtue ethics, and basically amounts to an RL-flavored take on that philosophical tradition.
Praxis-based values as I... | https://www.lesswrong.com/posts/eutmuwTpHCb4xYZfo/some-thoughts-on-virtue-ethics-for-ais |
# An Impossibility Proof Relevant to the Shutdown Problem and Corrigibility
The Incompatibility of a Utility Indifference Condition with Robustly Making Sane Pure Bets
-------------------------------------------------------------------------------------------
### Summary
It is provably impossible for an agent to rob... | https://www.lesswrong.com/posts/MBemd8k9uHFDEKzad/an-impossibility-proof-relevant-to-the-shutdown-problem-and |
# AGI safety career advice
People often ask me for career advice related to AGI safety. This post ([now also translated into Spanish](https://docs.google.com/document/d/12X5U-vYsiZkpSxTNAFTpcirX9p_UBR2Y_3DdglARtwU/edit?usp=sharing)) summarizes the advice I most commonly give. I’ve split it into three sections: general... | https://www.lesswrong.com/posts/ho63vCb2MNFijinzY/agi-safety-career-advice |
# Systems that cannot be unsafe cannot be safe
***Epistemic Status:** Trying to clarify a confusion people outside of the AI safety community seem to have about what safety means for AI systems.*
In engineering and design, there is a process that includes, among other stages, specification, creation, verification and... | https://www.lesswrong.com/posts/jiXMZHGmEf7qPrKPc/systems-that-cannot-be-unsafe-cannot-be-safe |
# Five Worlds of AI (by Scott Aaronson and Boaz Barak)
A relatively short April 27, 2023 post attempting to classify AI outcomes into 5 main cases:
* **AI-Fizzle**
* Not-AI-Fizzle
* Civilization recognizably continues
* **Futurama**
* **AI-Dystopia**
* Civilization does no... | https://www.lesswrong.com/posts/Kfr7PgXRQow9kbGx7/five-worlds-of-ai-by-scott-aaronson-and-boaz-barak |
# AI Safety Newsletter #4: AI and Cybersecurity, Persuasive AIs, Weaponization, and Geoffrey Hinton talks AI risks
Welcome to the AI Safety Newsletter by the [Center for AI Safety](https://www.safe.ai/). We discuss developments in AI and AI safety. No technical background required.
Subscribe [here](https://newsletter... | https://www.lesswrong.com/posts/MSw5y88tDbyp8WKo9/ai-safety-newsletter-4-ai-and-cybersecurity-persuasive-ais |
# Avoiding xrisk from AI doesn't mean focusing on AI xrisk
tl;dr: To avoid extinction, focus more on the positive opportunities for the future.
AI is an existential risk and an extinction risk - there is a non-insignificant probability that AI/AGI[^rlws7fphgo] will become super-powered, and a non-insignificant probab... | https://www.lesswrong.com/posts/Wigdk6s4xsC8fqhYT/avoiding-xrisk-from-ai-doesn-t-mean-focusing-on-ai-xrisk |
# Does descaling a kettle help? Theory and practice
I've heard that descaling a kettle makes it more efficient, and can save me time and money.
Okay, but how much? For some reason my my intuition says it'll be basically unnoticable. Let's try to figure it out, and then actually try it.
### The Models
There's a firs... | https://www.lesswrong.com/posts/4fPsFvDx8Q3KkiFyC/does-descaling-a-kettle-help-theory-and-practice |
# Are Emergent Abilities of Large Language Models a Mirage? [linkpost]
> Recent work claims that large language models display emergent abilities, abilities not present in smaller-scale models that are present in larger-scale models. What makes emergent abilities intriguing is two-fold: their sharpness, transitioning ... | https://www.lesswrong.com/posts/o3r2Qru84Rz5LsQk9/are-emergent-abilities-of-large-language-models-a-mirage |
# A Case for the Least Forgiving Take On Alignment
1\. Introduction
----------------
The field of AI Alignment is a pre-paradigmic one, and the primary symptom of that is the wide diversity of views across it. Essentially every senior researcher has their own research direction, their own idea of what the core proble... | https://www.lesswrong.com/posts/3JRBqRtHBDyPE3sGa/a-case-for-the-least-forgiving-take-on-alignment |
# Averting Catastrophe: Decision Theory for COVID-19, Climate Change, and Potential Disasters of All Kinds
***Averting Catastrophe*** **is a book by** [**Cass Sunstein**](https://en.wikipedia.org/wiki/Cass_Sunstein) **that NYU Press published** [**in April 2021**](https://nyupress.org/9781479808489/averting-catastroph... | https://www.lesswrong.com/posts/ANnpcyBMkeY6n9BYp/averting-catastrophe-decision-theory-for-covid-19-climate |
# Communication strategies for autism, with examples
This is a description of communication strategies which have been helpful to me in social situations where I was confused. It is relatively unedited; some of the content may be obvious and banal, or redundant.
\[Edited to move out of footnote, thanks [Aorou](http... | https://www.lesswrong.com/posts/bQjqcx5AkLwqprzEc/communication-strategies-for-autism-with-examples |
# How much do personal biases in risk assessment affect assessment of AI risks?
Had a stray thought that I think is worth exploring: perhaps disagreements over how to respond to AI risks are heavily influenced by personal biases in how to assess risk in general such that very different conclusions can be drawn from ev... | https://www.lesswrong.com/posts/RzGJB7GoosQmfALuE/how-much-do-personal-biases-in-risk-assessment-affect |
# Monthly Roundup #6: May 2023
#### Bad News
[Wild pigs continue to multiply and do increasing amounts of damage as they spread throughout America](https://www.texasmonthly.com/travel/texas-feral-hog-problem-swine-country/). Perhaps in the past we might have been willing to do what it would take to solve such problem... | https://www.lesswrong.com/posts/jEnKmiphDtEBNcopM/monthly-roundup-6-may-2023 |
# Finding Neurons in a Haystack: Case Studies with Sparse Probing
Abstract
========
> Despite rapid adoption and deployment of large language models (LLMs), the internal computations of these models remain opaque and poorly understood. In this work, we seek to understand how high-level human-interpretable features ar... | https://www.lesswrong.com/posts/yXiu6DBxKKWXC8Ygx/finding-neurons-in-a-haystack-case-studies-with-sparse |
# Slowing AI: Crunch time
It is often asserted that time will be more valuable and higher leverage in "crunch time," shortly before critical models are deployed.
This brief post provides a starting point for consideration of three questions:
1. What is more time (before dangerous AI) good for?
2. What is different... | https://www.lesswrong.com/posts/tZnDC4zZEbH3bdEcX/slowing-ai-crunch-time |
# Progress links and tweets, 2023-05-03
**The Progress Forum**
----------------------
* [Reflections from seeing the Starship launch firsthand](https://progressforum.org/posts/NskMryyCmndEx2wtY)
* [When America received a message from the future: the 1939 World’s Fair](https://progressforum.org/posts/6JS3HS4hvXbg... | https://www.lesswrong.com/posts/TzeHQNwsriEsahxDC/progress-links-and-tweets-2023-05-03 |
# List of notable people who believe in AI X-risk?
Remembering Scott's [AI Researchers On AI Risk](https://slatestarcodex.com/2015/05/22/ai-researchers-on-ai-risk/). The debate about AI is heating up, and a lot of people rely on authority figures for knowing what to believe. Do we have a current list of notable AI X-r... | https://www.lesswrong.com/posts/EoYqXbLFx9LF7haTo/list-of-notable-people-who-believe-in-ai-x-risk |
# Formalizing the "AI x-risk is unlikely because it is ridiculous" argument
There is a lot of good writing on technical arguments against AI x-risk (such as [Where I agree and disagree with Eliezer](https://www.lesswrong.com/posts/CoZhXrhpQxpy9xw9y/where-i-agree-and-disagree-with-eliezer) (which mostly argues for more... | https://www.lesswrong.com/posts/zxanhCijprzDo2Tgn/formalizing-the-ai-x-risk-is-unlikely-because-it-is |
# AGI rising: why we are in a new era of acute risk and increasing public awareness, and what to do now
\[Added **13Jun:** Submitted to OpenPhil AI Worldviews Contest - [this pdf version](https://www.dropbox.com/s/4qs7w9sqtql0cyq/%5BOpenPhil%20AI%20Worldviews%20Contest%5D%20AGI%20rising_%20why%20we%20are%20in%20a%20ne... | https://www.lesswrong.com/posts/YRfQss3TPfKAyGxLf/agi-rising-why-we-are-in-a-new-era-of-acute-risk-and |
# «Boundaries/Membranes» and AI safety compilation
In this post I outline every post I could find that meaningfully connects the concept of «Boundaries/Membranes» ([tag](https://www.lesswrong.com/tag/boundaries-membranes-technical)) with AI safety.[^h6pjihkqz8o] This seems to be a booming subtopic: interest has picked... | https://www.lesswrong.com/posts/fjgoMaBenyXcRDrbX/boundaries-membranes-and-ai-safety-compilation |
# Alignment Research @ EleutherAI
## The past and future of AI alignment at Eleuther
Initially, EleutherAI focused mainly on supporting open source research. AI alignment was something that was acknowledged by many of the core members as important, but it was not the primary focus. We mainly had discussions about the... | https://www.lesswrong.com/posts/LkkTWBahQBRSaYxot/alignment-research-eleutherai |
# Prizes for matrix completion problems
Here are two self-contained algorithmic questions that have come up in our research. We're offering a bounty of $5k for a solution to either of them—either an algorithm, or a lower bound under any hardness assumption that has appeared in the literature.
**Question 1 (existence ... | https://www.lesswrong.com/posts/pJrebDRBj9gfBE8qE/prizes-for-matrix-completion-problems |
# How MATS addresses “mass movement building” concerns
Recently, many AI safety movement-building programs have been criticized for attempting to grow the field too rapidly and thus:
1. Producing more aspiring alignment researchers than there are jobs or training pipelines;
2. Driving the wheel of AI hype and progr... | https://www.lesswrong.com/posts/iD6tYjLLFFn4LXgnt/how-mats-addresses-mass-movement-building-concerns |
# [New] Rejected Content Section
tl:dr;
======
* The LessWrong reviews all first time submissions from new users
* In the last month, we've built a new process where we reject some of those submissions (with explanation) – you can now see what was rejected from the main site on the new Rejected Content section
* ... | https://www.lesswrong.com/posts/PKGTKdRDneLA7bY54/new-rejected-content-section |
# Has the Symbol Grounding Problem just gone away?
A few years ago, the symbol grounding problem was widely considered a significant challenge in AI discussions. I believed that it would likely be addressed as a side effect of capability improvements, without requiring specific breakthroughs or attention, unlike other... | https://www.lesswrong.com/posts/CH3eo4QW9ZA5cWppn/has-the-symbol-grounding-problem-just-gone-away |
# We don’t need AGI for an amazing future
Most of the public discussions about AI safety seem to assume that there is only one plausible path for AI development: from narrow, specialized AI to broad AGI to superintelligence. So we seemingly have to decide between an amazing future with AGI at the risk that it will des... | https://www.lesswrong.com/posts/qh2jYyuXRrCuqEXPK/we-don-t-need-agi-for-an-amazing-future |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.