id
stringlengths
36
36
source
stringclasses
15 values
formatted_source
stringclasses
13 values
text
stringlengths
2
7.55M
504cc6e7-6f21-456d-9903-803c6a7d0f4a
trentmkelly/LessWrong-43k
LessWrong
Expected Utility, Geometric Utility, and Other Equivalent Representations In Scott Garrabrant's excellent Geometric Rationality sequence, he points out an equivalence between modelling an agent as * Maximizing the expected logarithm of some quantity V, E[ln(V)] * Maximizing the geometric expectation of V, G[V] And...
794a6c1d-5fb1-4d0d-a58e-4c870fd7d286
StampyAI/alignment-research-dataset/eaforum
Effective Altruism Forum
The Orthogonality Thesis is Not Obviously True (Crosspost of [this](https://benthams.substack.com/p/the-orthogonality-thesis-is-not-obviously) on my blog).   *The basic case* ================ > If you really accept the practical version of the Orthogonality Thesis, then it seems to me that you can’t regard educati...
e78c6396-885a-48bb-a544-e84143ec11d0
trentmkelly/LessWrong-43k
LessWrong
LessWrong Now Has Dark Mode One of the more commonly requested features on LessWrong is a dark mode. Well, now we have it! To activate, hover over the user-menu in the top right, open the Themes submenu, and click on Dark Mode there. Notes: * This feature is beta, by which I mean there may be aesthetic problems, es...
e6a8cd31-cb99-48d9-9499-ea3ff654ef41
StampyAI/alignment-research-dataset/blogs
Blogs
Some survey results! *By Katja Grace, 8 June 2017* We put the main results of [our survey](http://aiimpacts.org/2016-expert-survey-on-progress-in-ai/) of machine learning researchers on AI timelines online recently—see **[here](https://arxiv.org/abs/1705.08807)** for the paper. Apologies for the delay—we are tryin...
8df7861a-b8e3-4996-b73c-db889b25733f
trentmkelly/LessWrong-43k
LessWrong
Prediction: The Defense Department Will Blame Trump for the Slow Response on Jan. 7, 2021 Context Recently a rioting group of protestors attacked congress after attending a presidential speech. The members from both parties and the senate narrowly escaped. One capitol policeman and one protestor were killed. After bei...
d6add8e9-65e1-43c3-af79-fc9c0bf9c1b7
trentmkelly/LessWrong-43k
LessWrong
Do you like excessive sugar? In "Adaptation-Executers, not Fitness-Maximisers", Eliezer Yudkowsky writes: "Fifty thousand years ago, the taste buds of Homo sapiens directed their bearers to the scarcest, most critical food resources—sugar and fat. Calories, in a word. Today, the context of a taste bud's function has ...
484d4ae8-825b-4726-8c11-33be8e04c0e8
trentmkelly/LessWrong-43k
LessWrong
Formalized math: dream vs reality (Disclaimer: this post is intended as an enthusiastic introduction to a topic I knew absolutely nothing about till yesterday. Correct me harshly as needed.) We programmers like to instantly pigeonhole problems as "trivial" and "impossible". The AI-Box has been proposed as candidate f...
b9917864-2f20-46dc-ab20-e7a945d0853f
StampyAI/alignment-research-dataset/alignmentforum
Alignment Forum
[AN #162]: Foundation models: a paradigm shift within AI Alignment Newsletter is a weekly publication with recent content relevant to AI alignment around the world. Find all Alignment Newsletter **[resources here](http://rohinshah.com/alignment-newsletter/)**. In particular, you can look through **[this spreadsheet](h...
1513a7eb-d94f-418d-b6b5-97b6ef1597eb
trentmkelly/LessWrong-43k
LessWrong
Chronic perfectionism through the eyes of school reports A short post for a short thought. I was sifting through old school reports, and there's only one comment which is repeated in pretty much every report from 3 years old all the way through to 16. 3 years old: "Has a hard time dealing with disappointments. Tends ...
7041d0fd-9493-4139-9d5a-9e70a6efeb49
StampyAI/alignment-research-dataset/lesswrong
LessWrong
AI Alignment: A Comprehensive Survey We have just released an academic survey of AI alignment. We identify four main categories of alignment research: 1. Learning from feedback (e.g. scalable oversight) 2. Learning under distribution shift 3. Assurance (e.g. interpretability) 4. Governance We mainly focused on acad...
a8e81c60-2ccb-4576-b113-25c2125a9683
StampyAI/alignment-research-dataset/lesswrong
LessWrong
Stanovich on CEV [Keith Stanovich](http://web.mac.com/kstanovich/Site/Home.html) is a leading expert on the cogsci of rationality, but he also also written on a problem related to CEV, that of the "rational integration" of our preferences. Here he is on pages 81-86 of *[Rationality and the Reflective Mind](http://www....
be20d147-2f95-4bf2-b703-d55b3a6bd94f
trentmkelly/LessWrong-43k
LessWrong
The Costs of Reliability A question that used to puzzle me is “Why can people be so much better at doing a thing for fun, or to help their friends and family, than they are at doing the exact same thing as a job?” I’ve seen it in myself and I’ve seen it in others. People can be hugely more productive, creative, intel...
4d9b3596-4c16-4e9f-abc1-d6b612f418cc
StampyAI/alignment-research-dataset/special_docs
Other
Understanding Recommenders.. [![Luke Thorburn](https://miro.medium.com/v2/resize:fill:48:48/1*pgY6plckE3fmSouBjppMHg.jpeg)](/@lukethorburn?source=collection_home---------0----------------------------)[Luke Thorburn](/@lukethorburn?source=collection_home---------0----------------------------)[·May 7](/understanding-rec...
a2665580-0def-4372-95ab-aff1697be97f
trentmkelly/LessWrong-43k
LessWrong
The Lens That Sees Its Flaws Light leaves the Sun and strikes your shoelaces and bounces off; some photons enter the pupils of your eyes and strike your retina; the energy of the photons triggers neural impulses; the neural impulses are transmitted to the visual-processing areas of the brain; and there the optical inf...
f57a82b2-8fe2-47a8-8570-9fb31f965d0c
StampyAI/alignment-research-dataset/blogs
Blogs
what happens when you die? what happens when you die? -------------------------- contrary to popular (secular) belief, i don't believe nothing happens. considering that [you are your information system](you-are-your-information-system.html) and [nothing else](persistent-data-structures-consciousness.html), any fut...
80fa7416-9428-4de2-9d6f-e8b08365c2cf
trentmkelly/LessWrong-43k
LessWrong
Weekly LW Meetups This summary was posted to LW Main on December 9th. The following week's summary is here. Irregularly scheduled Less Wrong meetups are taking place in: * [Amsterdam] Meetup #8 - Reversed stupidity is not intelligence: 11 December 2016 03:11PM * Munich: 10 December 2016 04:00PM * [Zurich] Decembe...
629608a2-b65f-43d4-8b0e-f8a7d62421e7
trentmkelly/LessWrong-43k
LessWrong
Beware Isolated Demands For Rigor I. From Identity, Personal Identity, and the Self by John Perry: > "There is something about practical things that knocks us off our philosophical high horses. Perhaps Heraclitus really thought he couldn’t step in the same river twice. Perhaps he even received tenure for that contri...
5c995e04-8ce5-4456-8ebd-4b8f6c1f10d6
trentmkelly/LessWrong-43k
LessWrong
Productivity: Working towards a summary of what we know Epistemic effort * Spent 20-30 hours researching what we know about productivity. About an hour a day for two weeks, plus a few more days where I dedicated more than an hour to studying productivity. * Skimmed through the entire archives of Cal Newport's blog (...
fb43acc9-3627-4dd2-84a7-e8039ed83771
trentmkelly/LessWrong-43k
LessWrong
[Exploratory] Seperate exploratory writing from public writing Disclaimer: This is an exploratory writing post. No checking for typos or other editing was done. It seems that when I am writing, I basically almost always wirte for myself. This is fast and easy. I do find it useful to do this, for generating and fleshi...
0039dffc-ff05-49ad-b957-9815d65f0650
trentmkelly/LessWrong-43k
LessWrong
AI #24: Week of the Podcast In addition to all the written developments, this was a banner week for podcasts. I would highlight four to consider listening to. 1. Dario Amodei of Anthropic went on The Lunar Society to talk to Dwarkesh Patel. We got our best insight so far into where Dario’s head is at, Dwarkesh is e...
d92b7e04-004f-4955-b2f3-7dc269e21e9e
trentmkelly/LessWrong-43k
LessWrong
Science-informed normativity The debate over moral realism is often framed in terms of a binary question: are there ever objective facts about what’s moral to do in a given situation? The broader question of normative realism is also framed in a similar way: are there ever objective facts about what’s rational to do i...
b2425db4-d96d-4119-8776-50212835d9ef
trentmkelly/LessWrong-43k
LessWrong
Flowchart: How might rogue AIs defeat all humans? Epistemic status: I put this together for myself mostly off the top of my head without carefully cross-referencing other sources. I'm hoping to flesh this out a lot more and do a much more thorough research job of it at some point, but in the meantime I've decided to e...
551fadca-4212-49c4-a8f9-9a23d9fc4bbc
trentmkelly/LessWrong-43k
LessWrong
Guesstimate Algorithm for Medical Research This document is aimed at subcontractors doing medical research for me. I am sharing it in the hope it is more broadly useful, but have made no attempts to make it more widely accessible.  Intro Guesstimate is a tool I have found quite useful in my work, especially in makin...
3ec526d8-7646-4ff1-bd7d-85426ba59635
trentmkelly/LessWrong-43k
LessWrong
The value of Now. I am an easily bored Omega-level being, and I want to play a game with you. I am going to offer you two choices.  Choice 1: You spend the next thousand years in horrific torture, after which I restore your local universe to precisely the state it is in now (wiping your memory in the process), and h...
928d0269-6a8f-4faf-ba00-d4a783cf223a
trentmkelly/LessWrong-43k
LessWrong
A summary of aligning narrowly superhuman models This post was written under Evan Hubinger’s direct guidance and mentorship, as a part of the Stanford Existential Risks Institute ML Alignment Theory Scholars (MATS) program. Where not indicated, ideas are summaries of the documents mentioned, not original contributions...
c95331f6-22bd-4d88-a718-530f4c2b9459
trentmkelly/LessWrong-43k
LessWrong
Values, Valence, and Alignment I have previously advocated for finding a mathematically precise theory for formally approaching AI alignment. Most recently I couched this in terms of predictive coding and longer ago I was thinking in terms of a formalized phenomenology, but further discussions have helped me realize t...
5f12f4d3-465f-48f9-9326-54d1559edd13
trentmkelly/LessWrong-43k
LessWrong
[Altruist Support] LW Go Foom In which I worry that the Less Wrong project might go horribly right. This post belongs to my Altruist Support sequence. Every project needs a risk assessment. There's a feeling, just bubbling under the surface here at Less Wrong, that we're just playing at rationality. It's rationalit...
448d9630-eddd-4326-8644-fe06a334cdb5
trentmkelly/LessWrong-43k
LessWrong
Intelligence in Commitment Races A distillation of my understanding of the commitment races problem. Greaser Courage It's 1950 and you're a greaser in a greaser gang. Recreationally, you're driving your heavy American cars (without seatbelts) at each other at speed, and seeing who swerves first. Whoever swerves firs...
32b28f21-9b71-4d54-99d4-aa24b4c177d1
trentmkelly/LessWrong-43k
LessWrong
The Benefits of Distillation in Research Epistemic status: I wrote this six months ago, but I never felt satisfied with it, so I kind of forgot about it. At this point, I'll just release it. It is similar to Chris Olah's article on research debt, but it has a different perspective on presenting the mechanisms.  (The ...
f73c21b1-9ae4-4b99-84df-c751acbbb597
StampyAI/alignment-research-dataset/eaforum
Effective Altruism Forum
AGI safety from first principles *In [this report](https://www.alignmentforum.org/s/mzgtmmTKKn5MuCzFJ) (which I'm linking from the Alignment Forum) I have attempted to put together the most compelling case for why the development of AGI might pose an existential threat. It stems from my dissatisfaction with existing a...
9ddeafb1-6a06-4082-8e0c-a2128b3c906b
trentmkelly/LessWrong-43k
LessWrong
Melting democracy Liquid democracy, also known as delegative democracy, is a proposed hybrid between direct democracy and representative democracy. In liquid democracy, as in direct democracy, everyone is able to vote directly on every issue. Of course, in a modern society it’d be prohibitively time-consuming for ever...
3ec74705-a4ae-42b8-8d55-1654d2edfd97
StampyAI/alignment-research-dataset/alignmentforum
Alignment Forum
[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 of...
b1ccba05-cd4a-469a-b48b-79ad729f3b6e
trentmkelly/LessWrong-43k
LessWrong
Open Thread, Aug. 22 - 28, 2016 Notes for future OT posters: 1. Please add the 'open_thread' tag. 2. Check if there is an active Open Thread before posting a new one. (Immediately before; refresh the list-of-threads page before posting.) 3. Open Threads should start on Monday, and end on Sunday. 4. Unflag the tw...
ab16dcc0-1a75-4f58-906f-d94d09f9da60
trentmkelly/LessWrong-43k
LessWrong
How do health systems work in adequate worlds? I'm confused about Health Systems in general, the same way I was confused about the Education System until I read "The case against education". I don't have a good model of what are the different agents interacting in this system, nor the differences among different count...
c2d110fb-9c06-4823-8a9a-c83e07ce5394
trentmkelly/LessWrong-43k
LessWrong
The "Outside the Box" Box Whenever someone exhorts you to "think outside the box", they usually, for your convenience, point out exactly where "outside the box" is located.  Isn't it funny how nonconformists all dress the same... In Artificial Intelligence, everyone outside the field has a cached result for brilliant...
366e1a23-95ae-4b11-b985-f53c4304e75d
trentmkelly/LessWrong-43k
LessWrong
High school advice What advice does the Lesswrong community have in general for high school students who stumbled on this community?  Looking for all types of responses, it's fine whether your response is focused on achieving something in the AI/effective altruism/rationality community, or whether it is focused on a ...
d569b053-c030-4de0-b12f-955e2f454d83
trentmkelly/LessWrong-43k
LessWrong
Affordance Widths This article was originally a post on my tumblr. I'm in the process of moving most of these kinds of thoughts and discussions here. ---------------------------------------- Okay. There’s a social interaction concept that I’ve tried to convey multiple times in multiple conversations, so I’m going to...
5c4409fc-baa8-4225-a285-3174aad9548d
trentmkelly/LessWrong-43k
LessWrong
AI Alignment Research Overview (by Jacob Steinhardt) I'm really excited to see someone outline all the work they think needs solving in AI alignment - to describe what the problem looks like, what a solution looks like, and what work has been done so far. Especially from Jacob, who is a coauthor of the Concrete Proble...
08857be5-ca69-4f4f-8bde-021b59cd748a
trentmkelly/LessWrong-43k
LessWrong
Disagreeables and Assessors: Two Intellectual Archetypes Epistemic status: uncertain. I removed most weasel words for clarity, but that doesn't mean I'm very confident. Don’t take this too literally. I'd love to see a cluster analysis to see if there's actually a trend, this is a rough guess.[1] I'm interested in feed...
f3d326d9-6c63-4cec-b84f-2f6adac0c3d6
StampyAI/alignment-research-dataset/aisafety.info
AI Safety Info
Wouldn't a superintelligence be smart enough to know right from wrong? As far as we know, morality is a human construct - the universe itself does not encode morality in the laws of physics. There is no good reason to expect an arbitrary mind, which could be very different from our own, to share our values. A sufficie...
f95b1620-eec0-4a70-ad52-1b724d6e9968
trentmkelly/LessWrong-43k
LessWrong
An open response to Wittkotter and Yampolskiy A response to this paper. https://asi-safety-lab.com/DL/Kill-Switch-For_ASI_EW_21_12_14.pdf A substantial fraction of my argument comes down to. It is plausible that a hypothetical hypercompetent civilization could coordinate to build such infrastructure. However we don'...
5f1659a6-29c2-4946-9779-9e283dc06e22
trentmkelly/LessWrong-43k
LessWrong
Reframing the burden of proof: Companies should prove that models are safe (rather than expecting auditors to prove that models are dangerous) Evaluations of large language models (“model evals”) are one of the most commonly discussed AI governance ideas. The idea is relatively straightforward: we want to be able to u...
fd2433cf-d7b4-438f-bf6e-8bdb683f4638
trentmkelly/LessWrong-43k
LessWrong
What are the most interesting / challenging evals (for humans) available? I want to build a nice testing ground for human rationality (that is to say, solving arbitrary complex problems in different domains using limited information and time) This would be a lot of annoying work to assemble, but (un)fortunately there...
7062cc9c-ec90-4cb6-a3c1-508fdc9a8020
trentmkelly/LessWrong-43k
LessWrong
Elevator pitches/responses for rationality / AI I'm trying to develop a large set of elevator pitches / elevator responses for the two major topics of LW: rationality and AI. An elevator pitch lasts 20-60 seconds, and is not necessarily prompted by anything, or at most is prompted by something very vague like "So, I ...
0db6b2af-c2a1-4d14-9731-375c792d329a
trentmkelly/LessWrong-43k
LessWrong
Bird-eye view visualization of LLM activations I’m starting to learn about mechanistic interpretability, and I’m seeing lots of great visualizations of transformer internals, but somehow I’ve never seen the whole large model’s internal state shown at once, on one image. So I made this visualization, for Llama-2-7B. A...
f92dc122-637c-41ab-921e-bbf6faf4ce94
trentmkelly/LessWrong-43k
LessWrong
Respecting your Local Preferences In this post, I give a application of geometric rationality to a toy version of a real problem. A Conflicted Agent Let's say you are an agent with two partially conflicting goals. Part of you wants to play a video game, and part of you wants to save the world and tile the multiverse...
a8daae64-4228-4bdf-8c72-e546f07d3761
trentmkelly/LessWrong-43k
LessWrong
Parable of the vanilla ice cream curse (and how it would prevent a car from starting!) Story from https://www.cs.cmu.edu/~wkw/humour/carproblems.txt (my thoughts follow): > For those of us who understand that the obvious is not > always the solution, and that the facts, no matter how implausible, > are still the fact...
fe7d4019-1ad3-42d3-a5b2-5dc9af42ffed
trentmkelly/LessWrong-43k
LessWrong
TopicTitle'); DROP TABLE Users EXCEPT (AdeleneDawner OR Kevin); -- --;
356c3886-6b22-4e37-9a6b-020a5623a298
trentmkelly/LessWrong-43k
LessWrong
Ignorance in parenting Followup to: Strategic ignorance and plausible deniability My in-law always says: "For children it is easier be forgiven then to get permission."  EDIT: This post is superseeded by my Book Review: Kazdin's The Everyday Parenting Toolkit I recommend reading only that. The remaining insight of t...
04fe48d4-bcd0-4855-b9d7-ec7c5df99449
StampyAI/alignment-research-dataset/arbital
Arbital
Mutually exclusive and exhaustive A set of propositions is "mutually exclusive and exhaustive" when exactly one of the propositions must be true. For example, of the two propositions "The sky is blue" and "It is not the case that the sky is blue", exactly one of those must be the case. Therefore, the [probabilities]...
68d5b8cf-1a72-4307-8d74-e428f6f9ad6c
trentmkelly/LessWrong-43k
LessWrong
Incremental AI Risks from Proxy-Simulations Abstract: Numerical simulations are versatile predictive tools that permit explorations of complex systems. The ability of LLM agents to simulate real-world scenarios will expand the AI risk landscape. In the proxy-simulation threat model, a user (or a deceptively aligned AI...
fa511064-ef8e-4e54-80cf-8c6b4d051e72
trentmkelly/LessWrong-43k
LessWrong
Torturing people for fun Most of the usual thought experiments that justify expected utilitarialism trade off fun for fun, or suffering for suffering. Here's a situation which mixes the two. You are offered to press a button that will select a random person (not you) and torture them for a month. In return the machine...
fe3f06f3-4649-4463-9701-66fc16fdf0b6
trentmkelly/LessWrong-43k
LessWrong
Understandable vs justifiable vs useful This is a quickly written post, just sharing a rough conceptual distinction that I’ve found useful in everyday life and that I think others might find useful too. Your boss just expected you to stay late yesterday. Your partner just shot you a pretty damn sharp comment. Someone...
3f8409ff-b462-4e31-9d89-30549eb89976
trentmkelly/LessWrong-43k
LessWrong
Meetup : Meetup #6 - Still Amsterdam! Discussion article for the meetup : Meetup #6 - Still Amsterdam! WHEN: 13 November 2016 03:10:00PM (+0100) WHERE: Meester Treublaan 18, 1097 DP Amsterdam, Netherlands Ready for another meetup? Last time was a terrific success, so we want to build on that (also EAGx Oxford is in...
fca21ddd-a1f4-4722-a58d-c66e86d89caa
trentmkelly/LessWrong-43k
LessWrong
How to think about and deal with OpenAI Eliezer Yudkowsky writes on twitter: > Nothing else Elon Musk has done can possibly make up for how hard the "OpenAI" launch trashed humanity's chances of survival; previously there was a nascent spirit of cooperation, which Elon completely blew up to try to make it all be abou...
77bf1772-01f1-42ee-80a9-6f557cddaa42
trentmkelly/LessWrong-43k
LessWrong
Yann LeCun on AGI and AI Safety Yann recently gave a presentation at MIT on Objective-Driven AI with his specific proposal being based upon a Joint Embedding Predictive Architecture. He claims that his proposal will make AI safe and steerable, so I thought it was worthwhile copying the slides at the end which provide...
e47ed19f-f39d-4331-bbdd-e995877d4365
StampyAI/alignment-research-dataset/lesswrong
LessWrong
Robustness & Evolution [MLAISU W02] Welcome to this week’s ML Safety Report where we talk about robustness in machine learning and the human-AI dichotomy. Stay until the end to check out several amazing competitions you can participate in today. Watch this week's MLAISU on [YouTube](https://youtu.be/bT0VD0lGO_Q) or l...
fc6cba07-917a-4be0-b8dd-2638c8d44df3
trentmkelly/LessWrong-43k
LessWrong
Could Patent-Trolling delay AI timelines? The US patent system is broadly considered to be broken, and patent-trolls make a healthy profit abusing it. Agressively pursuing and enforcing software patents does not strike me as a sufficient or desirable means to delaying AI timelines, but I can imagine it as tool in a sl...
676b2105-1cbb-42b5-8af9-c2813c63b5f8
trentmkelly/LessWrong-43k
LessWrong
Newcomblike problems are the norm This is crossposted from my blog. In this post, I discuss how Newcomblike situations are common among humans in the real world. The intended audience of my blog is wider than the readerbase of LW, so the tone might seem a bit off. Nevertheless, the points made here are likely new to m...
61987c73-6f90-450a-b414-5995fd88abb4
StampyAI/alignment-research-dataset/blogs
Blogs
AI Timelines: Where the Arguments, and the "Experts," Stand *Audio also available by searching Stitcher, Spotify, Google Podcasts, etc. for "Cold Takes Audio"* [Today’s world Transformative AI Digital people World of Misaligned AI World run by Something else or or Stable, galaxy-wide civilizati...
b230616c-6e9e-4af1-9b59-a353b4a446e5
trentmkelly/LessWrong-43k
LessWrong
Meetup : UMD Meetup Discussion article for the meetup : UMD Meetup WHEN: 05 October 2011 04:00:00PM (-0400) WHERE: University of Maryland We're meeting in Terrapin Room B of the Student Involvement Suite in STAMP. Discussion article for the meetup : UMD Meetup
f6484413-2988-450b-aa74-393ddd58c450
trentmkelly/LessWrong-43k
LessWrong
Open thread, January 25- February 1 If it's worth saying, but not worth its own post (even in Discussion), then it goes here.
a89dae02-1a3e-4644-9b3b-0666869982a1
trentmkelly/LessWrong-43k
LessWrong
Which Issues in Conceptual Alignment have been Formalised or Observed (or not)? One aspect of alignment research I find interesting is improving how we understand potential threat / failure modes by making them more concrete or rigorous in some way. When I think of making these concrete, I’m thinking of theoretical re...
07a91b8e-c84e-40c0-b1d7-50b4cb761889
trentmkelly/LessWrong-43k
LessWrong
Making Exceptions to General Rules Suppose you make a general rule, ie. "I won't eat any cookies". Then you encounter a situation that legitimately feels exceptional , "These are generally considered the best cookies in the entire state". This tends to make people torn between two threads of reasoning: 1) Clearly the...
753ae6ab-cd9b-41a7-a8a1-c1ba3e6a7194
trentmkelly/LessWrong-43k
LessWrong
New article from Oren Etzioni (Cross-posted from EA Forum.) This just appeared in this week’s MIT Technology Review: Oren Etzioni, “How to know if AI is about to destroy civilization.” Etzioni is a noted skeptic of AI risk. Here are some things I jotted down: Etzioni’s key points / arguments: * Warning signs that ...
598951f6-6c7d-4977-9af5-5699bbdfcfdd
trentmkelly/LessWrong-43k
LessWrong
[Linkpost] 7 Swedish Words to Import
a8ff7c72-e33a-4815-a815-f94cc2819b97
trentmkelly/LessWrong-43k
LessWrong
Crafting Polysemantic Transformer Benchmarks with Known Circuits Notes:  * This research was performed as part of Adrià Garriga-Alonso’s MATS 6.0 stream. * If an opinion is stated in this post saying that "we" hold the opinion, assume it's Evan's opinion (Adrià is taking a well-deserved vacation at the time of writ...
396c3392-391b-4915-98aa-2ea38fb30b61
StampyAI/alignment-research-dataset/arxiv
Arxiv
Self-Modification of Policy and Utility Function in Rational Agents 1 Introduction --------------- Agents that are part of the environment they interact with may have the opportunity to self-modify. For example, humans can in principle modify the circuitry of their own brains, even though we currently lack the tech...
c4319d1a-9dde-4e0e-877a-85f335f549a9
StampyAI/alignment-research-dataset/special_docs
Other
Transhumanist FAQ 3.0 Transhumanist FAQ ================= The Transhumanist FAQ was developed in the mid-1990s and in 1998 became a formal FAQ through the inspirational work of transhumanists, including Alexander Chislenko, Max More, Anders Sandberg, Natasha Vita-More, Eliezer Yudkowsky, Arjen Kamphius, and many othe...
3c45bd75-1e0b-4775-a366-12b6f7224f73
trentmkelly/LessWrong-43k
LessWrong
Room For More Funding In AI Safety Is Highly Uncertain (Crossposted to the Effective Altruism Forum) ---------------------------------------- Introduction In effective altruism, people talk about the room for more funding (RFMF) of various organizations. RFMF is simply the maximum amount of money which can be donat...
fbaf8ee6-2759-4b8c-8dbb-9badc5a7ee1e
trentmkelly/LessWrong-43k
LessWrong
Announcing ILIAD2: ODYSSEY We are pleased to announce ILIAD2: ODYSSEY—a 5-day conference bringing together 100+ researchers to build scientific foundations for AI alignment. This is the 2nd iteration of ILIAD, which was first held in the summer of 2024.  ***Apply to attend by June 1!*** * When: Aug 25-29, 2025 * W...
2b31afde-40a9-46d8-a67e-ac9e3563858e
StampyAI/alignment-research-dataset/blogs
Blogs
Is there a tradeoff between immediate and longer-term AI safety efforts? Something I [often](https://www.wired.com/2017/02/ai-threat-isnt-skynet-end-middle-class/) [hear](http://uk.businessinsider.com/killer-robots-biases-artificial-intelligence-ai-2017-10?r=US&IR=T) in the machine learning community and media article...
80964a4c-eb7c-48a4-a08e-7389b3190f6d
trentmkelly/LessWrong-43k
LessWrong
The Leopold Model: Analysis and Reactions Previously: On the Podcast, Quotes from the Paper This is a post in three parts. The first part is my attempt to condense Leopold Aschenbrenner’s paper and model into its load bearing elements and core logic and dependencies. Two versions here, a long version that attempts ...
9d9207df-a7ea-4048-b7fd-a87e502c71ce
trentmkelly/LessWrong-43k
LessWrong
When are the most important times to wash your hands? I set a five minute timer to brainstorm some, but I bet that there are many that I've missed, and furthermore, it would be helpful to know which ones are most important / impactful. * I'm about to eat food * After I go to the bathroom * After I touch a package ...
b8e33684-fd83-4903-9a41-5575e533412c
trentmkelly/LessWrong-43k
LessWrong
aspirational leadership Crossposted from my blog. I’m going to first describe a leader’s responsibility to their team because I don’t think it’s possible to develop good principles for working and communicating effectively with individuals without understanding the broader environment that the leader-individual relat...
53669d54-bf69-4b38-8e63-3f026622005f
trentmkelly/LessWrong-43k
LessWrong
A Short Intro to Humans A Preface to Intro to Humans  I used to understand humans. We were God's avatars on earth, sent to witness His Glory and enjoy His Grace - maybe up to a measurment error. Than God died, leaving some huge holes in my worldview. To address one of these holes I bought an Introduction to Psychology...
4b44ff66-7001-4b0e-8f8d-d4a855b58723
StampyAI/alignment-research-dataset/lesswrong
LessWrong
A philosopher's critique of RLHF In the spring, I went to a talk with Brian Christian at Yale. He talked about his book, *The Alignment Problem*, and then there was an audience Q&A. There was a really remarkable question in that Q&A, which I have transcribed here. It came from the Yale philosophy professor [L.A. Paul]...
2f5cd046-dca6-44dd-8034-3d99f35e59ea
StampyAI/alignment-research-dataset/eaforum
Effective Altruism Forum
Begging, Pleading AI Orgs to Comment on NIST AI Risk Management Framework The National Institute of Standards and Technology (NIST) is seeking public comment until April 29 on its [Draft AI Risk Management Framework](https://www.nist.gov/system/files/documents/2022/03/17/AI-RMF-1stdraft.pdf).  NIST will produce a seco...
6d71d2a6-4bc0-428e-91cf-f4c97c03db64
trentmkelly/LessWrong-43k
LessWrong
Group Rationality Diary, February 2017 This is the public group rationality diary for February, 2017. It's a place to record and chat about it if you have done, or are actively doing, things like: * Established a useful new habit * Obtained new evidence that made you change your mind about some belief * Decided ...
5ef331e1-6f99-40de-9326-51133013bee9
trentmkelly/LessWrong-43k
LessWrong
Only humans can have human values Ethics is not geometry Western philosophy began at about the same time as Western geometry; and if you read Plato you'll see that he, and many philosophers after him, took geometry as a model for philosophy. In geometry, you operate on timeless propositions with mathematical operato...
306f5a95-3517-49de-b35b-f25589dbfaaa
StampyAI/alignment-research-dataset/eaforum
Effective Altruism Forum
Asking for online resources why AI now is near AGI How many percentage of ability are AI at current compared to the ability required for human-like intelligence? I heard some people in EA think it's already more than 50%, which shocked me a lot. I'm an definitely AI outsider, I can't imagine how AI conductsvhard tasks...
cca1e4ff-f3df-4002-aaa3-e4d0a4d8d330
trentmkelly/LessWrong-43k
LessWrong
What should a telic science look like? A brief (re-)introduction In case you don't want to read introductory post of this sequence, here is a quick recap of the broad context for this work. This post will try to clarify my objectives, and the next one will start to provide some concrete formalisms. When we want to un...
e31210d4-0f33-4607-b04d-d7ca6487ee5d
trentmkelly/LessWrong-43k
LessWrong
Hiroshima Day On August 6th, in 1945, the world saw the first use of atomic weapons against human targets.  On this day 63 years ago, humanity lost its nuclear virginity.  Until the end of time we will be a species that has used fission bombs in anger. Time has passed, and we still haven't blown up our world, despite...
400c37dd-7211-4b20-868d-4417725a3d28
trentmkelly/LessWrong-43k
LessWrong
How hard is it to disguise my gait? Gait recognition technology ( = identifying a person by the way they walk) is getting good, and I find this trend unnerving.  Current context: It's already deployed in China by a company called Watrix, which claims 94% accuracy up to 50 meters away from any angle. Japan, the US, an...
60ba0c45-b8f4-4be8-8949-9256d91d8c3e
trentmkelly/LessWrong-43k
LessWrong
From "Coulda" and "Woulda" to "Shoulda": Predicting Decisions to Minimize Regret for Partially Rational Agents TRIGGER WARNING: PHILOSOPHY.  All those who believe in truly rigorous, scientifically-grounded reasoning should RUN AWAY VERY QUICKLY. Abstract: Human beings want to make rational decisions, but their decisi...
bebbcc32-9941-4eb7-a3cc-a99907864348
StampyAI/alignment-research-dataset/eaforum
Effective Altruism Forum
AI Safety: Applying to Graduate Studies **Acknowledgements** -------------------- A huge thank you to Stephen Casper, Rachel Freedman, Lewis Hammond, and Pablo Moreno for taking the time to join our panel and answer questions! **Introduction** ---------------- On November 7, 2021, [AI Safety Support](https://www.ai...
99f07298-4260-4119-9637-b60b195212a5
trentmkelly/LessWrong-43k
LessWrong
FW25 Color Stats We’ve been here before. I’ve posted about color trends in fashion in 2023, twice in 2024, and now I’m doing it again. Methodology Recap Fashion operates on a seasonal cycle; designer ready-to-wear fall/winter collections are released in the spring (so stores have time to buy them). Vogue Magazine ki...
ded489cc-d251-4d11-8bc9-78ed60b38253
trentmkelly/LessWrong-43k
LessWrong
Is there any metric measuring ~"proportion of people creating extra value"? Imagine a simplified model of the world, where the population suddenly increases by 1000 individuals. Of these newcomers, perhaps 50 will become policemen to ensure safety, 30 will serve as doctors to maintain health, 25 will take on roles as ...
b73ef26b-bc98-4cdf-a004-b0666970d0f3
LDJnr/LessWrong-Amplify-Instruct
LessWrong
""Why do humans reason" (PDF), a paper by Hugo Mercier and Dan Sperber, reviewing an impressive amount of research with a lot of overlap with themes previously explored on Less Wrong, suggests that our collective efforts in "refining the art of human rationality" may ultimately be more successful than most individual e...
4dfc67f1-f472-4c66-a0cc-dd5fde073ab3
trentmkelly/LessWrong-43k
LessWrong
How big of an issue are patent trolls to the average startup? There are many individual stories about patent trolls being problematic, but I have a hard time estimating how big of an effect their activity happens to be overall. Can someone point me to guide analysis of the problem overall that estimates their impact?
cd3452ba-6bab-4f94-87eb-ce19abac643b
trentmkelly/LessWrong-43k
LessWrong
Natural language alignment I think a potentially promising and undertheorized approach to AI safety, especially in short timelines, is natural language alignment (NLA), a form of AI-assisted alignment in which we leverage the model’s rich understanding of human language to help it develop and pursue a safe[1] notion o...
4179c313-82fb-4d6a-8620-e190ca164ea7
trentmkelly/LessWrong-43k
LessWrong
The Flexibility of Abstract Concepts I can discuss Daoist ideas with Taiwanese friends easily even if they have no background in Daoism. But when I try the same thing with white people it often feels as if I'm trying to explain quantum field theory to someone who has never heard of mathematics. It is easier for me to ...
01d39b83-b955-4cb9-9af9-98654bb597fa
trentmkelly/LessWrong-43k
LessWrong
When do utility functions constrain? The Problem This post is an exploration of a very simple worry about the concept of utility maximisers - that they seem capable of explaining any exhibited behaviour. It is one that has, in different ways, has been brought up many times before. Rohin Shah, for example, complained ...
211e8423-b910-4868-8e73-e3872bf53d60
trentmkelly/LessWrong-43k
LessWrong
Weekly LW Meetups: Baltimore, Houston, Chicago, Melbourne, and Penn State There are upcoming irregularly scheduled Less Wrong meetups in: * Baltimore meetup: 01 October 2011 02:00PM * Houston Meetup - Saturday, October 1: 01 October 2011 03:00PM * Chicago Meetup: 02 October 2011 03:00PM * Pittsburgh Meetup: 05 Oc...
0a236e2a-0960-42bb-82c6-33f4f9020ce0
trentmkelly/LessWrong-43k
LessWrong
Welcome to Columbus SSC Meetup [Edit With Your Details] (The following are our suggestions for what kind of information is best to include in the welcome post of your group, feel free to replace them with whatever you think is best) What kind of events does your group usually run? What does it usually do? How freque...
687c11da-2b5b-46f3-b101-72681013f209
StampyAI/alignment-research-dataset/arxiv
Arxiv
Learning to Play No-Press Diplomacy with Best Response Policy Iteration ### 1 Introduction Artificial Intelligence methods have achieved exceptionally strong competitive play in board games such as Go, Chess, Shogi [[108](#bib.bib108), [110](#bib.bib110), [109](#bib.bib109)], Hex [[2](#bib.bib2)], Poker [[85](#bib...
0196fafd-68bc-414f-bc3f-c7443e58a283
trentmkelly/LessWrong-43k
LessWrong
[Link] 'Something feels wrong with the state of philosophy today.' http://blog.talkingphilosophy.com/?p=6424 Some interesting thoughts on the current state of philosophy, and Less Wrong gets cited a couple of times. > Something feels wrong with the state of philosophy today. From whence hast this sense of ill-bodi...
3bfb6ee5-8840-4e85-a270-015c428988d6
trentmkelly/LessWrong-43k
LessWrong
Meetup : Moscow LW lecture centre meetup: The New Foundation Discussion article for the meetup : Moscow LW lecture centre meetup: The New Foundation WHEN: 25 January 2015 02:00:00PM (+0300) WHERE: Moscow, Mitnaya, 52 Welcome to the first meetup of Moscow LW lecture centre (don't mix up with LW club, which meets in ...
1829e234-1137-4d68-ae63-3e0e2686f5d7
trentmkelly/LessWrong-43k
LessWrong
Oops It's Time To Overthrow the Organizer Day! Summary: Sometimes, your regular organizer catches the pandemic, or moves to the bay, or has a newborn, or does a bunch of sketchy stuff, or just doesn’t have time anymore. If the usual organizers suddenly weren’t around, would your community be able to keep going?  Tags...
424e998f-254d-4acf-ac14-127c4ef08822
trentmkelly/LessWrong-43k
LessWrong
Acausal Now: We could totally acausally bargain with aliens at our current tech level if desired tl;dr. Weird forms of cooperation become humanly possible with the correct mechanisms. It might be interesting to try them out for practice/research purposes, even if there is not much to gain directly from aliens. Note: ...