url
stringlengths
52
124
post_id
stringlengths
17
17
title
stringlengths
2
248
author
stringlengths
2
49
content
stringlengths
22
295k
date
stringclasses
376 values
https://www.lesswrong.com/posts/Trnaa9skzuWBJsH7A/plausibility-of-getting-early-warning-shots-because-ais-can
Trnaa9skzuWBJsH7A
Plausibility of Getting Early Warning Shots because AIs can't coordinate?
the-cactus
I don't know if this is a well known argument that has already been responded to. If it is, just delete the post. An implicit assumption most people make when discussion takeover risk, is that any misaligned agent will stave off enacting takeover plans with less than very high probability of success. If the Agent has a...
2024-04-27
https://www.lesswrong.com/posts/md5n2EJxon7XNotDe/exploring-the-esoteric-pathways-to-ai-sentience-part-one
md5n2EJxon7XNotDe
Exploring the Esoteric Pathways to AI Sentience (Part One)
jeffreycaruso
We've been told by VCs and founders in the AI space that Human-level Artificial Intelligence (formerly AGI), followed by Superintelligence, will bring about a techno-utopia, if it doesn't kill us all first. In order to fulfill that dream, AI must be sentient, and that requires it have consciousness. Today, AI is neithe...
2024-04-27
https://www.lesswrong.com/posts/EpnRFR2zYSrEJGKGB/concept-dependency-edge-regular-lattice-graph
EpnRFR2zYSrEJGKGB
[Concept Dependency] Edge Regular Lattice Graph
johannes-c-mayer
This is a Concept Dependency Post. It may not be worth reading on its own, out of context. See the backlinks at the bottom to see which posts use this concept. Also known as periodically labeled lattice graphs in graph theory. Here is a concrete Edge Regular Lattice Graph: In this graph, the following pattern is repeat...
2024-04-26
https://www.lesswrong.com/posts/237cdeNLxAGWhSbDD/concept-dependency-concept-dependency-posts
237cdeNLxAGWhSbDD
[Concept Dependency] Concept Dependency Posts
johannes-c-mayer
This is a Concept Dependency Post. It may not be worth reading on its own, out of context. See the backlinks at the bottom to see which posts use this concept. See the backlinks at the bottom of the post. Every post starting with [Concept Dependency] is a concept dependency post, that describes a concept this post is u...
2024-04-26
https://www.lesswrong.com/posts/ZCi7GgwkfnpmWM3AF/wouldn-t-weak-ai-agents-provide-warning
ZCi7GgwkfnpmWM3AF
Wouldn't weak AI agents provide warning?
Mandatory Topic
If you are an advanced AI at ~70-120% of human level with some goal that is best optimized by destroying humanity you are unlikely to be able to optimize your goal because at your level of capability you are unlikely to be able to destroy all humans. However it still seems in your interest to try. If you are being trai...
2024-04-26
https://www.lesswrong.com/posts/CLS4FijfEFHzc5HEv/duct-tape-security
CLS4FijfEFHzc5HEv
Duct Tape security
KingSupernova
This is a linkpost for On Duct Tape and Fence Posts. Eliezer writes about fence post security. When people think to themselves "in the current system, what's the weakest point?", and then dedicate their resources to shoring up the defenses at that point, not realizing that after the first small improvement in that area...
2024-04-26
https://www.lesswrong.com/posts/XgC9XrrmuNqMAGYpe/ai-safety-sphere
XgC9XrrmuNqMAGYpe
AI Safety Sphere
zarsou9
[Update: This post is unfortunately no longer relevant as I have stopped developing the website. It is still kinda cool, but it can be difficult to have people collaborate on a mind map. Also there is less of an issue of finding relevant papers with recent advancements in search] AI Safety Sphere is a collaborative min...
2024-04-27
https://www.lesswrong.com/posts/xyL5kb8RBGLiupGLf/scaling-of-ai-training-runs-will-slow-down-after-gpt-5
xyL5kb8RBGLiupGLf
Scaling of AI training runs will slow down after GPT-5
maxime-riche
My credence: 33% confidence in the claim that the growth in the number of GPUs used for training SOTA AI will slow down significantly directly after GPT-5. It is not higher because of (1) decentralized training is possible, and (2) GPT-5 may be able to increase hardware efficiency significantly, (3) GPT-5 may be smalle...
2024-04-26
https://www.lesswrong.com/posts/7Pt9fogptmiSduXt9/spatial-attention-as-a-tell-for-empathetic-simulation
7Pt9fogptmiSduXt9
Spatial attention as a “tell” for empathetic simulation?
steve2152
(UPDATE Sept 2024: (1) I still think “local spatial attention” is a thing as described below. But I no longer think that it functions as a ‘tell’ for empathetic simulation, as in the headline claim. I have an alternate theory I like much better. Details coming in the next couple months hopefully. Also, (2) as noted bel...
2024-04-26
https://www.lesswrong.com/posts/hveHKFcs3LZh2rJXh/d-and-d-sci-long-war-defender-of-data-mocracy
hveHKFcs3LZh2rJXh
D&D.Sci Long War: Defender of Data-mocracy
aphyer
This is an entry in the 'Dungeons & Data Science' series, a set of puzzles where players are given a dataset to analyze and an objective to pursue using information from that dataset. STORY (skippable) You have the excellent fortune to live under the governance of The People's Glorious Free Democratic Republic of Earth...
2024-04-26
https://www.lesswrong.com/posts/GWNqHMQZnnAuwytgu/arch-anarchy
GWNqHMQZnnAuwytgu
Arch-anarchy
Peter lawless
By A [Editor: This article is reprinted from Extropy #5, Winter 1990. Extropy was published by The Extropy Institute] Call to Arms Down with the law of gravity! By what right does it counter my will? I have not pledged my allegiance to the law of gravity; I have learned to live under its force as one learns to live und...
2024-04-26
https://www.lesswrong.com/posts/XMKfsvczDhgbS5fxz/breadboarding-a-whistle-synth
XMKfsvczDhgbS5fxz
Breadboarding a Whistle Synth
jkaufman
With my electronic harp mandolin project I've been enjoying working with analog and embedded audio hardware. And a few weeks ago, after reading about Ugo Conti's whistle-controlled synth I wrote to him, he gave me a call, and we had a really interesting conversation. And my existing combination of hardware for my whi...
2024-04-26
https://www.lesswrong.com/posts/jsmNCj9QKcfdg8fJk/an-introduction-to-ai-sandbagging
jsmNCj9QKcfdg8fJk
An Introduction to AI Sandbagging
teun-van-der-weij
Summary: Evaluations provide crucial information to determine the safety of AI systems which might be deployed or (further) developed. These development and deployment decisions have important safety consequences, and therefore they require trustworthy information. One reason why evaluation results might be untrustwort...
2024-04-26
https://www.lesswrong.com/posts/sfWPjmfZY4Q5qFC5o/why-i-m-doing-pauseai
sfWPjmfZY4Q5qFC5o
Why I'm doing PauseAI
Josephm
GPT-5 training is probably starting around now. It seems very unlikely that GPT-5 will cause the end of the world. But it’s hard to be sure. I would guess that GPT-5 is more likely to kill me than an asteroid, a supervolcano, a plane crash or a brain tumor. We can predict fairly well what the cross-entropy loss will be...
2024-04-30
https://www.lesswrong.com/posts/ZxAWeiT8qNYppPbYA/llms-seem-relatively-safe
ZxAWeiT8qNYppPbYA
LLMs seem (relatively) safe
JustisMills
Post for a somewhat more general audience than the modal LessWrong reader, but gets at my actual thoughts on the topic. In 2018 OpenAI defeated the world champions of Dota 2, a major esports game. This was hot on the heels of DeepMind’s AlphaGo performance against Lee Sedol in 2016, achieving superhuman Go performance ...
2024-04-25
https://www.lesswrong.com/posts/qBDQQqQ9dWhJJ7Jt6/losing-faith-in-contrarianism
qBDQQqQ9dWhJJ7Jt6
Losing Faith In Contrarianism
omnizoid
Crosspost from my blog. If you spend a lot of time in the blogosphere, you’ll find a great deal of people expressing contrarian views. If you hang out in the circles that I do, you’ll probably have heard of Yudkowsky say that dieting doesn’t really work, Guzey say that sleep is overrated, Hanson argue that medicine doe...
2024-04-25
https://www.lesswrong.com/posts/utnS8tthytY8rT6fm/why-i-stopped-being-into-basin-broadness
utnS8tthytY8rT6fm
Why I stopped being into basin broadness
tailcalled
There was a period where everyone was really into basin broadness for measuring neural network generalization. This mostly stopped being fashionable, but I'm not sure if there's enough written up on why it didn't do much, so I thought I should give my take for why I stopped finding it attractive. This is probably a rep...
2024-04-25
https://www.lesswrong.com/posts/jGuXSZgv6qfdhMCuJ/refusal-in-llms-is-mediated-by-a-single-direction
jGuXSZgv6qfdhMCuJ
Refusal in LLMs is mediated by a single direction
andy-arditi
This work was produced as part of Neel Nanda's stream in the ML Alignment & Theory Scholars Program - Winter 2023-24 Cohort, with co-supervision from Wes Gurnee. This post is a preview for our upcoming paper, which will provide more detail into our current understanding of refusal. We thank Nina Rimsky and Daniel Palek...
2024-04-27
https://www.lesswrong.com/posts/PRZzA2wsFjaj4n8Pg/axrp-episode-29-science-of-deep-learning-with-vikrant-varma
PRZzA2wsFjaj4n8Pg
AXRP Episode 29 - Science of Deep Learning with Vikrant Varma
DanielFilan
YouTube link In 2022, it was announced that a fairly simple method can be used to extract the true beliefs of a language model on any given topic, without having to actually understand the topic at hand. Earlier, in 2021, it was announced that neural networks sometimes ‘grok’: that is, when training them on certain tas...
2024-04-25
https://www.lesswrong.com/posts/vdoiWXeouGsZEYgrg/improving-dictionary-learning-with-gated-sparse-autoencoders
vdoiWXeouGsZEYgrg
Improving Dictionary Learning with Gated Sparse Autoencoders
SenR
Authors: Senthooran Rajamanoharan*, Arthur Conmy*, Lewis Smith, Tom Lieberum, Vikrant Varma, János Kramár, Rohin Shah, Neel Nanda A new paper from the Google DeepMind mech interp team: Improving Dictionary Learning with Gated Sparse Autoencoders! Gated SAEs are a new Sparse Autoencoder architecture that seems to be a s...
2024-04-25
https://www.lesswrong.com/posts/Li4evP8nL7Xg4Wjjw/why-i-write-by-george-orwell-1946
Li4evP8nL7Xg4Wjjw
"Why I Write" by George Orwell (1946)
arjun-panickssery
People have been posting great essays so that they're "fed through the standard LessWrong algorithm." This essay is in the public domain in the UK but not the US. From a very early age, perhaps the age of five or six, I knew that when I grew up I should be a writer. Between the ages of about seventeen and twenty-four I...
2024-04-25
https://www.lesswrong.com/posts/8EyCQKuWo6swZpagS/superposition-is-not-just-neuron-polysemanticity
8EyCQKuWo6swZpagS
Superposition is not "just" neuron polysemanticity
LawChan
TL;DR: In this post, I distinguish between two related concepts in neural network interpretability: polysemanticity and superposition. Neuron polysemanticity is the observed phenomena that many neurons seem to fire (have large, positive activations) on multiple unrelated concepts. Superposition is a specific explanatio...
2024-04-26
https://www.lesswrong.com/posts/Ejy4rRpGwsr9fCriP/the-first-future-and-the-best-future
Ejy4rRpGwsr9fCriP
The first future and the best future
KatjaGrace
It seems to me worth trying to slow down AI development to steer successfully around the shoals of extinction and out to utopia. But I was thinking lately: even if I didn’t think there was any chance of extinction risk, it might still be worth prioritizing a lot of care over moving at maximal speed. Because there are m...
2024-04-25
https://www.lesswrong.com/posts/ixJMcbizgmXRXxFbT/nih-cancer-myths-myths
ixJMcbizgmXRXxFbT
NIH Cancer Myths Myths
belkarx
The NIH has a page called Cancer Myths and Misconceptions that you come across if you end up looking into cancer for long enough, aimed at bio-illiterate patients and their families. Around half the things on that page are wrong at face value, and a solid percentage of those are contradicted by the pages and studies th...
2024-04-25
https://www.lesswrong.com/posts/wfAF4v8kikxJoNqfL/social-lemon-markets
wfAF4v8kikxJoNqfL
social lemon markets
bhauth
I refuse to join any club that would have me as a member. — Groucho Marx Alice and Carol are walking on the sidewalk in a large city, and end up together for a while. "Hi, I'm Alice! What's your name?" Carol thinks: If Alice is trying to meet people this way, that means she doesn't have a much better option for meeting...
2024-04-25
https://www.lesswrong.com/posts/okufhDiGpqK3AYbDi/cybersecurity-of-frontier-ai-models-a-regulatory-review
okufhDiGpqK3AYbDi
Cybersecurity of Frontier AI Models: A Regulatory Review
deric-cheng
This article is part of a series of ~10 posts comprising a 2024 State of the AI Regulatory Landscape Review, conducted by the Governance Recommendations Research Program at Convergence Analysis. Each post will cover a specific domain of AI governance (e.g. incident reporting, safety evals, model registries, etc.). We’l...
2024-04-25
https://www.lesswrong.com/posts/bwjfXXf3GmGSe6rrP/bayesian-inference-without-priors
bwjfXXf3GmGSe6rrP
Bayesian inference without priors
DanielFilan
Epistemic status: party trick Why remove the prior One famed feature of Bayesian inference is that it involves prior probability distributions. Given an exhaustive collection of mutually exclusive ways the world could be (hereafter called ‘hypotheses’), one starts with a sense of how likely the world is to be described...
2024-04-24
https://www.lesswrong.com/posts/Y8rEA4e4DxafmeAbW/the-inner-ring-by-c-s-lewis
Y8rEA4e4DxafmeAbW
The Inner Ring by C. S. Lewis
saul-munn
Note: In @Nathan Young's words "It seems like great essays should go here and be fed through the standard LessWrong algorithm. There is possibly a copyright issue here, but we aren't making any money off it either." What follows is a full copy of the C. S. Lewis essay "The Inner Ring" the 1944 Memorial Lecture at King’...
2024-04-24
https://www.lesswrong.com/posts/2x8EJnEu4JZMiNbxd/this-is-water-by-david-foster-wallace
2x8EJnEu4JZMiNbxd
This is Water by David Foster Wallace
Nathan Young
Note: It seems like great essays should go here and be fed through the standard LessWrong algorithm. There is possibly a copyright issue here, but we aren't making any money off it either. What follows is a full copy of "This is Water" by David Foster Wallace his 2005 commencement speech to the graduating class at Keny...
2024-04-24
https://www.lesswrong.com/posts/BBCtWtg44Yeh6fire/is-being-a-trans-woman-or-just-low-t-20-iq
BBCtWtg44Yeh6fire
Is being a trans woman (or just low-T) +20 IQ?
lcmgcd
Note September 2024: I still think this question is quite important, but what I wrote in April was just a bit too toxic/insensitive, even by my own lenient standards. I tried to move the post to a pastebin, but it got blocked by their automatic content filters, which I guess confirms it's unnecessarily spicy. I think t...
2024-04-24
https://www.lesswrong.com/posts/ZNu7rQLnggos5wZt3/betadine-oral-rinses-for-covid-and-other-viral-infections
ZNu7rQLnggos5wZt3
Betadine oral rinses for covid and other viral infections
pktechgirl
Before we get started, this is your quarterly reminder that I have no medical credentials and my highest academic credential is a BA in a different part of biology (with a double major in computer science). In a world with a functional medical system no one would listen to me. Tl;dr povidone iodine probably reduces vir...
2024-04-24
https://www.lesswrong.com/posts/sH2WnakxL4qAmPP4P/at-last-chatgpt-does-shall-we-say-interesting-imitations-of
sH2WnakxL4qAmPP4P
At last! ChatGPT does, shall we say, interesting imitations of “Kubla Khan”
bill-benzon
Cross-posted from New Savanna. As you may know, I have a long-term interest in Coleridge’s “Kubla Khan” and have written a lot of posts about it. If you scan through those posts you’ll find that some of the more recent one’s involve ChatGPT. Here’s my first attempt to have ChatGPT parody the poem: How ChatGPT parodied ...
2024-04-24
https://www.lesswrong.com/posts/DNzzfBAZx8Laua4MB/magic-by-forgetting
DNzzfBAZx8Laua4MB
Magic by forgetting
avturchin
Epistemic – this post is more suitable for LW as it was 10 years ago Thought experiment with curing a disease by forgetting Imagine I have a bad but rare disease X. I may try to escape it in the following way: 1. I enter the blank state of mind and forget that I had X. 2. Now I in some sense merge with a very large num...
2024-04-24
https://www.lesswrong.com/posts/PTC7bZdZoqbCcAshW/changes-in-college-admissions
PTC7bZdZoqbCcAshW
Changes in College Admissions
Zvi
This post brings together various questions about the college application process, as well as practical considerations of where to apply and go. We are seeing some encouraging developments, but mostly the situation remains rather terrible for all concerned. Application Strategy and Difficulty Paul Graham: Colleges that...
2024-04-24
https://www.lesswrong.com/posts/qxakrNr3JEoRSZ8LE/1-page-outline-of-carlsmith-s-otherness-and-control-series
qxakrNr3JEoRSZ8LE
1-page outline of Carlsmith's otherness and control series
Nathan Young
Joe’s summary is here, these are my condensed takeaways in my own words. All links in this section are to the essays. Outline Carlsmith tackles two linked questions:How should we behave towards future beings (future humans, AIs etc)?What should our priors be about how AIs will behave towards us?So the first question:We...
2024-04-24
https://www.lesswrong.com/posts/LSDtWbE6acmff6Tig/ai-generated-music-as-a-method-of-installing-essential
LSDtWbE6acmff6Tig
AI Generated Music as a Method of Installing Essential Rationalist Skills
keltan
On April Fools the LW team released an album under the name of the Fooming Shoggoths. Ever since the amount that I think about rationality has skyrocketed. That's because I've been listening exclusively to it when I'd usually be listening to other music. Especially Thought That Faster (feat. Eliezer Yudkowsky). I now f...
2024-04-24
https://www.lesswrong.com/posts/NTezyvvHbmoZzCsGA/beta-tester-request-rallypoint-bounties
NTezyvvHbmoZzCsGA
Beta Tester Request: Rallypoint Bounties
marc/er
We are requesting beta testers for the closed beta release of Rallypoint: A platform for the creation of bountied requests related to AI safety. We offer a convenient means of requesting a good or service in exchange for a financial reward that is able to be crowdfunded. The utility of Rallypoint is best communicated t...
2024-05-25
https://www.lesswrong.com/posts/cfnjE4LJhX9qp9HhJ/electronic-harp-mandolin-prototype
cfnjE4LJhX9qp9HhJ
Electronic Harp Mandolin Prototype
jkaufman
While I'm not really at a clear stopping point, I wanted to write up my recent progress with the electronic harp mandolin project. If I go too much further without writing anything up I'm going to start forgetting things. First, a demo: Or, if you're prefer a different model: Since last time, I: Fixed my interference...
2024-04-24
https://www.lesswrong.com/posts/csHstEPagqs8wChhh/examples-of-highly-counterfactual-discoveries
csHstEPagqs8wChhh
Examples of Highly Counterfactual Discoveries?
johnswentworth
The history of science has tons of examples of the same thing being discovered multiple time independently; wikipedia has a whole list of examples here. If your goal in studying the history of science is to extract the predictable/overdetermined component of humanity's trajectory, then it makes sense to focus on such e...
2024-04-23
https://www.lesswrong.com/posts/hTgWygNXbEo4mAgXM/is-there-software-to-practice-reading-expressions
hTgWygNXbEo4mAgXM
Is there software to practice reading expressions?
lsusr
I took the Reading the Mind in the Eyes Test test today. I got 27/36. Jessica Livingston got 36/36. Reading expressions is almost mind reading. Practicing reading expressions should be easy with the right software. All you need is software that shows a random photo from a large database, asks the user to guess what it ...
2024-04-23
https://www.lesswrong.com/posts/uAQxLJtGrRQg3LgqA/let-s-design-a-school-part-1
uAQxLJtGrRQg3LgqA
Let's Design A School, Part 1
Sable
The American school system, grades K-12, leaves much to be desired. While its flaws are legion, this post isn’t about that. It’s easy to complain. This post is about how we could do better. To be clear, I’m talking about redesigning public education, so “just use the X model” where X is “charter” or “Montessori” or “ho...
2024-04-23
https://www.lesswrong.com/posts/eDaPLeLcEftrGSopD/wsj-inside-amazon-s-secret-operation-to-gather-intel-on
eDaPLeLcEftrGSopD
WSJ: Inside Amazon’s Secret Operation to Gather Intel on Rivals
TrevorWiesinger
The operation, called Big River Services International, sells around $1 million a year of goods through e-commerce marketplaces including eBay, Shopify, Walmart and Amazon AMZN 1.49%increase; green up pointing triangle.com under brand names such as Rapid Cascade and Svea Bliss. “We are entrepreneurs, thinkers, marketer...
2024-04-23
https://www.lesswrong.com/posts/R5EJn3TFpHDQBW3Af/on-minicircle
R5EJn3TFpHDQBW3Af
On Minicircle
Metacelsus
Max Berry has analyzed Minicircle's follistatin gene therapy, and I agree with his conclusion that it is unlikely to be effective. There are several aspects of their design that could be improved (in particular, using a more efficient lipid nanoparticle delivery system instead of PEI, and making a better choice of prom...
2024-04-23
https://www.lesswrong.com/posts/99P2GwacTwp2mjhxq/when-should-you-work-through-the-night-when-inspiration
99P2GwacTwp2mjhxq
(When) Should you work through the night when inspiration strikes you?
Chi Nguyen
A lot of the time, I'm not very motivated to work, at least on particular projects. Sometimes, I feel very inspired and motivated to work on a particular project that I usually don't feel (as) motivated to work on. Sometimes, this happens in the late evening or at night. And hence I face the question: To sleep or to wo...
2024-04-23
https://www.lesswrong.com/posts/jMccY2X5aRZmED2yv/manifold-exploring-real-cash-prizes
jMccY2X5aRZmED2yv
Manifold “exploring real cash prizes”
Rana Dexsin
Manifold Markets has announced that they intend to add cash prizes to their current play-money model, with a raft of attendant changes to mana management and conversion. I first became aware of this via a comment on ACX Open Thread 326; the linked Notion document appears to be the official one. The central change invol...
2024-04-23
https://www.lesswrong.com/posts/AfABcGZshpyx2nxiZ/book-review-deep-utopia
AfABcGZshpyx2nxiZ
Book review: Deep Utopia
PeterMcCluskey
Book review: Deep Utopia: Life and Meaning in a Solved World, by Nick Bostrom. Bostrom's previous book, Superintelligence, triggered expressions of concern. In his latest work, he describes his hopes for the distant future, presumably to limit the risk that fear of AI will lead to a The Butlerian Jihad-like scenario. W...
2024-04-23
https://www.lesswrong.com/posts/y4rz6BFENLBThxrHm/on-what-research-policymakers-actually-need
y4rz6BFENLBThxrHm
On what research policymakers actually need
MondSemmel
I saw this guest post on the Slow Boring substack, by a former senior US government official, and figured it might be of interest here. The post's original title is "The economic research policymakers actually need", but it seemed to me like the post could be applied just as well to other fields. Excerpts (totaling ~75...
2024-04-23
https://www.lesswrong.com/posts/gknc6NWCNuTCe8ekp/simple-probes-can-catch-sleeper-agents-1
gknc6NWCNuTCe8ekp
Simple probes can catch sleeper agents
montemac
This is a link post for the Anthropic Alignment Science team's first "Alignment Note" blog post. We expect to use this format to showcase early-stage research and work-in-progress updates more in the future. Twitter thread here. Top-level summary: In this post we present "defection probes": linear classifiers that use ...
2024-04-23
https://www.lesswrong.com/posts/c6oLNbKNDtGE5xpNL/dequantifying-first-order-theories-1
c6oLNbKNDtGE5xpNL
Dequantifying first-order theories
jessica.liu.taylor
The Löwenheim–Skolem theorem implies, among other things, that any first-order theory whose symbols are countable, and which has an infinite model, has a countably infinite model. This means that, in attempting to refer to uncountably infinite structures (such as in set theory), one "may as well" be referring to an onl...
2024-04-23
https://www.lesswrong.com/posts/w5DPADKQw2KdqujgE/vector-planning-in-a-lattice-graph
w5DPADKQw2KdqujgE
Vector Planning in a Lattice Graph
johannes-c-mayer
You want to get to your sandwich: Well, that’s easy. Apparently we are in some kind of grid world, which is presented to us in the form of a lattice graph, where each vertex represents a specific world state, and the edges tell us how we can traverse the world states. We just do BFS to go from S (where we are) to T (wh...
2024-04-23
https://www.lesswrong.com/posts/HEpufTdakGTTKgoYF/prolu-a-nonlinearity-for-sparse-autoencoders
HEpufTdakGTTKgoYF
ProLU: A Nonlinearity for Sparse Autoencoders
Glen Taggart
Abstract This paper presents ProLU, an alternative to ReLU for the activation function in sparse autoencoders that produces a pareto improvement over both standard sparse autoencoders trained with an L1 penalty and sparse autoencoders trained with a Sqrt(L1) penalty. ProLU(mi,bi)={miif mi+bi>0 and mi>00otherwiseSAEProL...
2024-04-23
https://www.lesswrong.com/posts/t7gqDrb657xhbKkem/uncovering-deceptive-tendencies-in-language-models-a
t7gqDrb657xhbKkem
Uncovering Deceptive Tendencies in Language Models: A Simulated Company AI Assistant
jarviniemi
Abstract: We study the tendency of AI systems to deceive by constructing a realistic simulation setting of a company AI assistant. The simulated company employees provide tasks for the assistant to complete, these tasks spanning writing assistance, information retrieval and programming. We then introduce situations whe...
2024-05-06
https://www.lesswrong.com/posts/JE37rMdXqhxSaoDDC/rejecting-television
JE37rMdXqhxSaoDDC
Rejecting Television
declan-molony
I didn’t use to be, but now I’m part of the 2% of U.S. households without a television. With its near ubiquity, why reject this technology? The Beginning of my Disillusionment Neil Postman’s book Amusing Ourselves to Death radically changed my perspective on television and its place in our culture. Here’s one illuminat...
2024-04-23
https://www.lesswrong.com/posts/RYCvRwwKumBTPrhax/thoughts-on-zero-points
RYCvRwwKumBTPrhax
Thoughts on Zero Points
anchpop
There’s a cool concept I’ve been thinking about. I first heard of it when reading Jesse Schell’s book “The Art of Game Design”. (Fun fact: Jesse Schell was my professor’s professor, aka my grand-professor.) Then I heard of it again in the LessWrong post “Choosing the Zero Point”. Having been exposed to it twice, I now ...
2024-04-23
https://www.lesswrong.com/posts/F6pf38EQMxtMA5J36/lw-frontpage-experiments-aka-take-the-wheel-shoggoth
F6pf38EQMxtMA5J36
LW Frontpage Experiments! (aka "Take the wheel, Shoggoth!")
Ruby
Update: June 20th After a few rounds of adjustments and careful examination of the data, we've decided to make the Enriched tab be the default for all logged-in users[1]. Anyone is who has not already switched tabs ever will be set to Enriched tab. If you dislike it, you can switch to the Latest tab, the prior default....
2024-04-23
https://www.lesswrong.com/posts/C7deNdJkdtbzPtsQe/funny-anecdote-of-eliezer-from-his-sister
C7deNdJkdtbzPtsQe
Funny Anecdote of Eliezer From His Sister
daniel-birnbaum
This comes from a podcast called 18Forty, of which the main demographic of Orthodox Jews. Eliezer's sister (Hannah) came on and talked about her Sheva Brachos, which is essentially the marriage ceremony in Orthodox Judaism. People here have likely not seen it, and I thought it was quite funny, so here it is: https://18...
2024-04-22
https://www.lesswrong.com/posts/46uzmEgzW6yjs9cQX/how-llms-work-in-the-style-of-the-economist
46uzmEgzW6yjs9cQX
How LLMs Work, in the Style of The Economist
utilistrutil
The Assignment: (4 hours) Write an Economist-style explainer article on how LLMs work. You’ve just started as an AI reporter at The Economist, and your editor’s realised there’s no good Economist Explains style piece on how LLMs work. They’ve asked you to write one. It should be 500 words, and in the style of other Eco...
2024-04-22
https://www.lesswrong.com/posts/3LuZm3Lhxt6aSpMjF/ai-regulation-is-unsafe
3LuZm3Lhxt6aSpMjF
AI Regulation is Unsafe
maxwell-tabarrok
Concerns over AI safety and calls for government control over the technology are highly correlated but they should not be. There are two major forms of AI risk: misuse and misalignment. Misuse risks come from humans using AIs as tools in dangerous ways. Misalignment risks arise if AIs take their own actions at the expe...
2024-04-22
https://www.lesswrong.com/posts/YREGdnrdQLwqANzKj/on-precise-out-of-context-steering
YREGdnrdQLwqANzKj
On precise out-of-context steering
jarviniemi
Meta: This is a minor and relatively unimportant problem I've worked on. I'll be brief in my writing. Thanks to Aaron Scher for lots of conversations on the topic. Added later: The Covert Malicious Fine-tuning work is both conceptually and empirically superior to what I've done here, so I recommend reading it. Summary ...
2024-05-03
https://www.lesswrong.com/posts/sKKxuqca9uhpFSvgq/priors-and-prejudice
sKKxuqca9uhpFSvgq
Priors and Prejudice
MathiasKirkBonde
I Imagine an alternate version of the Effective Altruism movement, whose early influences came from socialist intellectual communities such as the Fabian Society, as opposed to the rationalist diaspora. Let’s name this hypothetical movement the Effective Samaritans. Like the EA movement of today, they believe in doing ...
2024-04-22
https://www.lesswrong.com/posts/pyRdxwfLxB9gwJJ35/on-llama-3-and-dwarkesh-patel-s-podcast-with-zuckerberg
pyRdxwfLxB9gwJJ35
On Llama-3 and Dwarkesh Patel’s Podcast with Zuckerberg
Zvi
It was all quiet. Then it wasn’t. Note the timestamps on both of these. Dwarkesh Patel did a podcast with Mark Zuckerberg on the 18th. It was timed to coincide with the release of much of Llama-3, very much the approach of telling your story directly. Dwarkesh is now the true tech media. A meteoric rise, and well earne...
2024-04-22
https://www.lesswrong.com/posts/cRFtWjqoNrKmgLbFw/we-are-headed-into-an-extreme-compute-overhang
cRFtWjqoNrKmgLbFw
We are headed into an extreme compute overhang
devrandom
If we achieve AGI-level performance using an LLM-like approach, the training hardware will be capable of running ~1,000,000s concurrent instances of the model. Definitions Although there is some debate about the definition of compute overhang, I believe that the AI Impacts definition matches the original use, and I pre...
2024-04-26
https://www.lesswrong.com/posts/ojPJYci3AdgdHsKY8/motivation-gaps-why-so-much-ea-criticism-is-hostile-and-lazy
ojPJYci3AdgdHsKY8
Motivation gaps: Why so much EA criticism is hostile and lazy
lombertini
null
2024-04-22
https://www.lesswrong.com/posts/kz4taEHAcDkNGN6F4/should-we-break-up-google-deepmind
kz4taEHAcDkNGN6F4
Should we break up Google DeepMind?
hauke-hillebrandt
null
2024-04-22
https://www.lesswrong.com/posts/d396HCvYG7SSqg9Hh/take-scifs-it-s-dangerous-to-go-alone
d396HCvYG7SSqg9Hh
Take SCIFs, it’s dangerous to go alone
latterframe
Coauthored by Dmitrii Volkov1, Christian Schroeder de Witt2, Jeffrey Ladish1 (1Palisade Research, 2University of Oxford). We explore how frontier AI labs could assimilate operational security (opsec) best practices from fields like nuclear energy and construction to mitigate near-term safety risks stemming from AI R&D ...
2024-05-01
https://www.lesswrong.com/posts/K2BsvRGmN7WJTx2wP/what-should-our-containers-do
K2BsvRGmN7WJTx2wP
What should our containers do?
richard-henage
If we don't have free will, we're just minds stuck inside (biological) robot bodies and brains. Some argue that this would mean that there's no point in trying to determine which courses of action are right, or what facts are true, since we wouldn't have control over how we behave or what we believe. In other words, we...
2024-04-22
https://www.lesswrong.com/posts/oxsBpx9v3bgxraiPj/towards-a-formalization-of-the-agent-structure-problem
oxsBpx9v3bgxraiPj
Towards a formalization of the agent structure problem
Alex_Altair
In Clarifying the Agent-Like Structure Problem (2022), John Wentworth describes a hypothetical instance of what he calls a selection theorem. In Scott Garrabrant's words, the question is, does agent-like behavior imply agent-like architecture? That is, if we take some class of behaving things and apply a filter for age...
2024-04-29
https://www.lesswrong.com/posts/SozrRa4KtHW9W8GoH/geometrically-maximal-lottery-lotteries-exist
SozrRa4KtHW9W8GoH
(Geometrically) Maximal Lottery-Lotteries Exist
Lorxus
epistemic/ontological status: almost certainly all of the following - a careful research-grade writeup of a genuinely kinda shiny open(?) question in theoretical psephology that will likely never see actual serious non-cooked-up use;dedicated to a very dear cat;utterly dependent, for the entirety of the most interestin...
2024-05-03
https://www.lesswrong.com/posts/FhryNAFknqKAdDcYy/how-to-use-and-interpret-activation-patching
FhryNAFknqKAdDcYy
How to use and interpret activation patching
Stefan42
This is a write-up of Neel’s and my experience and opinions on best practices for doing Activation Patching. A arXiv PDF version of this post is available here (easier to cite). A previous version was shared with MATS Program scholars in July 2023 under the title "Everything Activation Patching". Pre-requisites: This p...
2024-04-24
https://www.lesswrong.com/posts/QTTCRytvyFteJgPwg/transfer-learning-in-humans
QTTCRytvyFteJgPwg
Transfer Learning in Humans
niplav
I examine the literature on transfer learning in humans. Far transfer is difficult to achieve, best candidate interventions are to practice at the edge of one's ability and make many mistakes, evaluate mistakes after one has made them, learn from training programs modeled after expert tacit knowledge, and talk about on...
2024-04-21
https://www.lesswrong.com/posts/nQCarXwqY6u6sozGN/i-created-an-asi-alignment-tier-list
nQCarXwqY6u6sozGN
I created an Asi Alignment Tier List
TimeGoat
Website: https://www.AsiATL.net/ Check out the mission page to learn about the mission of AsiATL. For now, I’m looking for General feedback / constructive criticismPeople to join the discord and hang outFeedback on the tier list contentMaking a case to demote a singular item is the easiest for me to process. Example: X...
2024-04-21
https://www.lesswrong.com/posts/Eu6Rsc6bThkAosFCY/the-losing-identity-of-twitter
Eu6Rsc6bThkAosFCY
The losing identity of Twitter
itay-dreyfus
What’s Twitter for you? That's a long-lasting trend I often see on my feed when people praise the blue bird for getting them a job, introducing them to new people, investors, and all this and that. What about me? I just wanted to get into dribbble — the then invite-only designer's social network which was at its peak a...
2024-04-21
https://www.lesswrong.com/posts/v9mdyNBfEE8tsTNLb/subjective-questions-require-subjective-information
v9mdyNBfEE8tsTNLb
Subjective Questions Require Subjective information
ben-lang
This summarizes a (possibly trivial) observation that I found interesting. Story An all-powerful god decides to play a game. They stop time, grab a random human, and ask them "What will you see next?". The human answers, then time is switched back on and the god looks at how well they performed. Most of the time the hu...
2024-04-23
https://www.lesswrong.com/posts/EZQcKfe7HpJi7c9nx/paper-the-ethics-of-advanced-ai-assistants-google-deepmind
EZQcKfe7HpJi7c9nx
Paper: "The Ethics of Advanced AI Assistants" -Google DeepMind
tristan-wegner
Google Deepmind just released a 200 page paper. A few Authors are from the community.
2024-04-21
https://www.lesswrong.com/posts/aQMrsMG9B3EqzpQJL/contra-chord-simplification
aQMrsMG9B3EqzpQJL
Contra Chord Simplification
jkaufman
Tomorrow, Lily and I will be leading a Kids Contra Jam at NEFFA (2pm in the Sudbury room!). We'll be playing off of Lily's tune list, but someone was asking about chords. I decided to have a go at writing out the simplest acceptable chords for each of the tunes we're planning. Each letter represents two downbeats: A...
2024-04-21
https://www.lesswrong.com/posts/wbnWNSCyjFKyknvBy/forget-everything-statistical-mechanics-part-1
wbnWNSCyjFKyknvBy
Forget Everything (Statistical Mechanics Part 1)
Jemist
EDIT: I somehow missed that John Wentworth and David Lorell are also in the middle of a sequence have written one post on this same topic here. I will see where this goes from here! This sequence will continue! Introduction to a sequence on the statistical thermodynamics of some things and maybe eventually everything. ...
2024-04-22
https://www.lesswrong.com/posts/ZN6L5ysKd35FEyGr6/a-couple-productivity-tips-for-overthinkers
ZN6L5ysKd35FEyGr6
A couple productivity tips for overthinkers
steve2152
1. If you find that you’re reluctant to permanently give up on to-do list items, “deprioritize” them instead I hate the idea of deciding that something on my to-do list isn’t that important, and then deleting it off my to-do list without actually doing it. Because once it’s off my to-do list, then quite possibly I’ll n...
2024-04-20
https://www.lesswrong.com/posts/aH9R8amREaDSwFc97/rapid-capability-gain-around-supergenius-level-seems
aH9R8amREaDSwFc97
Rapid capability gain around supergenius level seems probable even without intelligence needing to improve intelligence
Simon Skade
TLDR: Around Einstein-level, relatively small changes in intelligence can lead to large changes in what one is capable to accomplish.E.g. Einstein was a bit better than the other best physi at seeing deep connections and reasoning, but was able to accomplish much more in terms of impressive scientific output.There are ...
2024-05-06
https://www.lesswrong.com/posts/ErYqeQpz8TWAmgmzD/you-re-the-most-beautiful-girl-in-the-world-and
ErYqeQpz8TWAmgmzD
"You're the most beautiful girl in the world" and Wittgensteinian Language Games
Chris_Leong
Wittgenstein argues that we shouldn't understand language by piecing together the dictionary meaning of each individual word in a sentence, but rather that language should be understood in context as a move in a language game. Consider the phrase, "You're the most beautiful girl in the world". Many rationalists might s...
2024-04-20
https://www.lesswrong.com/posts/KE6X8peAKutYNXnx5/past-tense-features
KE6X8peAKutYNXnx5
Past Tense Features
Can Rager
TLDR I find past tense features in pythia-70m using a templated dataset. My high-level steps are: Creating a templated dataset that indicates past tense through a past progressive clauseFinding subsets of features that recover the original model performance with attribution patchingAnalyzing the feature effects by toke...
2024-04-20
https://www.lesswrong.com/posts/DHkkL2GxhxoceLzua/thoughts-on-seed-oil
DHkkL2GxhxoceLzua
Thoughts on seed oil
dynomight
A friend has spent the last three years hounding me about seed oils. Every time I thought I was safe, he’d wait a couple months and renew his attack: “When are you going to write about seed oils?” “Did you know that seed oils are why there’s so much {obesity, heart disease, diabetes, inflammation, cancer, dementia}?” “...
2024-04-20
https://www.lesswrong.com/posts/XHCPDNA38tc2YDftM/how-to-know-whether-you-are-an-idealist-or-a-physicalist
XHCPDNA38tc2YDftM
How to know whether you are an idealist or a physicalist/materialist
JackOfAllSpades
You have heard and perhaps even used the expression "observable universe", right? What is included in the purportedly observable universe? The moon? The whole of the moon? If you had heard the expression "observable universe" a century ago, would you have been including the far side of the moon in that category?
2024-04-20
https://www.lesswrong.com/posts/okNWgNrZiQNoTieMn/how-i-think-part-four-money-is-weird
okNWgNrZiQNoTieMn
How I Think, Part Four: Money is Weird
richard-henage
If you work for free, you're doing whoever you're working for a favor. If you work for money but never spend it, you're doing the world a favor[1]. Except... When you buy someone's goods or services for their set price, you're doing them a favor. When you work for someone at their set wage, you're doing them a favor. S...
2024-04-20
https://www.lesswrong.com/posts/RdjKL7uroCGTFtWPL/wisdomism-a-moral-theory-for-the-age-of-information
RdjKL7uroCGTFtWPL
WISDOMISM A Moral Theory for the Age of Information
Peter lawless
by Tom W. Bell[Editor: this article is reprinted from Extropy # 2, winter 1989. Extropy was published by the Extropy institute] In the last issue of EXTROPY, co-editor Max O'Connor presented a number of powerful arguments for amoralism in his article titled "Morality or Reality?". While I share many of Max's sentiments...
2024-04-19
https://www.lesswrong.com/posts/C5KAZQib3bzzpeyrg/full-post-progress-update-1-from-the-gdm-mech-interp-team
C5KAZQib3bzzpeyrg
[Full Post] Progress Update #1 from the GDM Mech Interp Team
neel-nanda-1
This is a series of snippets about the Google DeepMind mechanistic interpretability team's research into Sparse Autoencoders, that didn't meet our bar for a full paper. Please start at the summary post for more context, and a summary of each snippet. They can be read in any order. Activation Steering with SAEs Arthur C...
2024-04-19
https://www.lesswrong.com/posts/X5bXnA7WHopGoMH4X/daniel-dennett-has-died-1942-2024
X5bXnA7WHopGoMH4X
Daniel Dennett has died (1942-2024)
kave
Daniel Dennett, professor emeritus of philosophy at Tufts University, well-known for his work in philosophy of mind and a wide range of other philosophical areas, has died. Professor Dennett wrote extensively about issues related to philosophy of mind and cognitive science, especially consciousness. He is also recogniz...
2024-04-19
https://www.lesswrong.com/posts/bkzfhqo4Q6mEkwEp2/events-booking-new-callers
bkzfhqo4Q6mEkwEp2
Events Booking New Callers?
jkaufman
A friend asked whether anyone else had noticed a pattern where big contra dance events were generally booking more established callers since restarting. This could make a lot of sense: the established callers will be less "overplayed" than they had been, and many events will be less robust financially and so more risk...
2024-04-19
https://www.lesswrong.com/posts/v2FReduuvLWZA3jyr/what-is-the-best-way-to-talk-about-probabilities-you-expect
v2FReduuvLWZA3jyr
What is the best way to talk about probabilities you expect to change with evidence/experiments?
Will_Pearson
I was thinking about my p(doom) in the next 10 years and came up with something around 6%[1]. However that involves lots of current unknowns to me, like the nature of current human knowledge production (and the bottle necks involved) which impact my P(doom) to be either 3% or 15% depending upon what type of bottle neck...
2024-04-19
https://www.lesswrong.com/posts/o5yRZXTRFKuF49RAi/demonstrate-and-evaluate-risks-from-ai-to-society-at-the-ai
o5yRZXTRFKuF49RAi
Demonstrate and evaluate risks from AI to society at the AI x Democracy research hackathon
esben-kran
null
2024-04-19
https://www.lesswrong.com/posts/vsaMKdJNQHgSxWy2a/how-to-model-the-future-of-open-source-llms
vsaMKdJNQHgSxWy2a
How to Model the Future of Open-Source LLMs?
joel-burget
I previously expected open-source LLMs to lag far behind the frontier because they're very expensive to train and naively it doesn't make business sense to spend on the order of $10M to (soon?) $1B to train a model only to give it away for free. But this has been repeatedly challenged, most recently by Meta's Llama 3. ...
2024-04-19
https://www.lesswrong.com/posts/HpAr8k74mW4ivCvCu/summary-progress-update-1-from-the-gdm-mech-interp-team
HpAr8k74mW4ivCvCu
[Summary] Progress Update #1 from the GDM Mech Interp Team
neel-nanda-1
Introduction This is a progress update from the Google DeepMind mechanistic interpretability team, inspired by the Anthropic team’s excellent monthly updates! Our goal was to write-up a series of snippets, covering a range of things that we thought would be interesting to the broader community, but didn't yet meet our ...
2024-04-19
https://www.lesswrong.com/posts/HCNK2AFovWzRm2NZw/the-power-of-finite-and-the-weakness-of-infinite-binary
HCNK2AFovWzRm2NZw
The power of finite and the weakness of infinite binary point numbers
Disbeliever
Take a sequence of the numbers 0 and 1. You can interpret this sequence as a real number between zero and one. Just write 0.( your sequence). This can be interpreted as a binary number. For instance 0.1101 is 1/2+1/4+0/8+1/16. Now, I have read here about the might of irrational numbers, whose sequences go on and on, ne...
2024-04-20
https://www.lesswrong.com/posts/HZsQrv5ELLCBYTrfy/testing-for-parallel-reasoning-in-llms
HZsQrv5ELLCBYTrfy
Testing for parallel reasoning in LLMs
meeri
Summary We study language models' capability to perform parallel reasoning in one forward pass. To do so, we test GPT-3.5's ability to solve (in one token position) one or two instances of algorithmic problems. We consider three different problems: repeatedly iterating a given function, evaluating a mathematical expres...
2024-05-19
https://www.lesswrong.com/posts/beFF5GyBy4FntYFRQ/if-digital-goods-in-virtual-worlds-increase-gdp-do-we
beFF5GyBy4FntYFRQ
If digital goods in virtual worlds increase GDP, do we actually become richer?
no77e-noi
Noah Smith, in this article, argues that the Metaverse could enable economic growth to increase a lot and sharply decouple itself from real-world resource usage. By creating markets in which we buy and sell immaterial things, world GDP would grow. He also says, rightly, that GDP correlates with the well-being of a nati...
2024-04-19
https://www.lesswrong.com/posts/RPdG2fSiSPLkCSPjg/experiment-on-repeating-choices
RPdG2fSiSPLkCSPjg
Experiment on repeating choices
KatjaGrace
People behave differently from one another on all manner of axes, and each person is usually pretty consistent about it. For instance: how much to spend money how much to worry how much to listen vs. speak how much to jump to conclusions how much to work how playful to be how spontaneous to be how much to prepare How m...
2024-04-19
https://www.lesswrong.com/posts/XhhuSWvg9X2sq4JC5/the-thermodynamics-of-death
XhhuSWvg9X2sq4JC5
The Thermodynamics of Death
Peter lawless
by M.C. Price [Editor: This article is reprinted from Extropy #6, Summer 1990. Extropy was published by The Extropy institute] First this little planet with its winds and ways, and then all the laws of mind and matter that restrain him. Then the planets about him, and at last out across immensity to the stars. And when...
2024-04-19
https://www.lesswrong.com/posts/wonC6LQArRcFXgebZ/backyard-office
wonC6LQArRcFXgebZ
Backyard Office
jkaufman
In 2020 I renovated the small building in our backyard which had fallen into disrepair. It was zoned for use as a home office, and had electric but not plumbing. I wrote about how I was thinking about insulating it and comparing framing options but then apparently I never got around to writing up how I finished it! I...
2024-04-19
https://www.lesswrong.com/posts/YCFsfQzAnXW4GEKS5/cohesion-and-business-problems
YCFsfQzAnXW4GEKS5
Cohesion and business problems
adamzerner
In software development there is a concept called cohesion. It works like this. Suppose you have the following functions:[1] function getArea(radius) { ... } function getCircumference(radius) { ... } function sendWelcomeEmail(user) { ... } function updatePassword(user, newPassword) { ... } function getTemperatureInFahr...
2024-04-19
https://www.lesswrong.com/posts/foseau7Eqg22mCkpY/hydrogen-tube-transport
foseau7Eqg22mCkpY
hydrogen tube transport
bhauth
Elon Musk's Hyperloop proposal had substantial public interest. With various initial Hyperloop projects now having failed, I thought some people might be interested in a high-speed transportation system that's...perhaps not "practical" per se, but at least more-practical than the Hyperloop approach. aerodynamic drag in...
2024-04-18
https://www.lesswrong.com/posts/DfqxDezqCeFJPjYsL/i-m-open-for-projects-sort-of
DfqxDezqCeFJPjYsL
I'm open for projects (sort of)
cousin_it
I left Google a month ago, and right now don't work. Writing this post in case anyone has interesting ideas what I could do. This isn't an "urgently need help" kind of thing - I have a little bit of savings, right now planning to relax some more weeks and then go into some solo software work. But I thought I'd write th...
2024-04-18