id
stringlengths
36
36
source
stringclasses
15 values
formatted_source
stringclasses
13 values
text
stringlengths
2
7.55M
756352b2-e023-47a6-8954-ef5696c12bd2
StampyAI/alignment-research-dataset/alignmentforum
Alignment Forum
VOI is Only Nonnegative When Information is Uncorrelated With Future Action Attention conservation notice: Most of this post is a recap of the standard arguments for UDT, but the way in which the standard proof for value of information being nonnegative breaks down in some decision theory scenarios probably isn't comm...
de7eb03f-0534-439f-9404-95489f1c3091
trentmkelly/LessWrong-43k
LessWrong
Capitalism as the Catalyst for AGI-Induced Human Extinction Capitalism as the Catalyst for AGI-Induced Human Extinction  By A. Nobody   ----------------------------------------   Introduction: The AI No One Can Stop As the world races toward Artificial General Intelligence (AGI)—a machine capable of human-level ...
133e2834-d271-43e7-a392-4f6cb87c7845
trentmkelly/LessWrong-43k
LessWrong
Weekly LW Meetups: Austin; Bratislava, Slovakia; Cambridge MA; Melbourne; Sydney; Washington DC This summary was posted to LW main on August 17th, and has been moved to discussion. There are upcoming irregularly scheduled Less Wrong meetups in: * Bratislava, Slovakia - the first LW meetup: 18 August 2012 05:00PM ...
6e6181e8-799e-43ab-82a7-95cb04e03eb2
trentmkelly/LessWrong-43k
LessWrong
[SEQ RERUN] Three Fallacies of Teleology Today's post, Three Fallacies of Teleology was originally published on 25 August 2008. A summary (taken from the LW wiki):   > Theories of teleology have a few problems. First, theories of teleology often wind up drawing causal arrows from the future to the past. It also lead...
90b78130-4720-4ff6-b66c-d84c69c9fcb5
trentmkelly/LessWrong-43k
LessWrong
Approaches to Electronic Contra I've been thinking about the various ways that musicians have approached combining elements of electronic dance music with contra dance. Examples: * Lisa Greenleaf, Concord 2008 * YDW 2008 * Asheville 2010 * Perpetual eMotion, Dance Flurry 2011 * Double Apex, BIDA 2011 * DJ Imp...
13d6fe5c-3b83-4dc7-bcea-2b51a44dd552
trentmkelly/LessWrong-43k
LessWrong
Given one AI, why not more? Suppose there is a threshold of capability beyond which an AI may pose a non-negligible existential risk to humans. What is the argument against this reasoning: If one AI passes or seems likely to pass this threshold, then humans, to lower x-risk, ought to push other AI past this threshold...
4fe90cc5-ac68-472f-9799-05d91453ae19
trentmkelly/LessWrong-43k
LessWrong
Meetup : Seattle-Vancouver Kilomeetup Discussion article for the meetup : Seattle-Vancouver Kilomeetup WHEN: 18 May 2013 11:54:51AM (-0700) WHERE: 1600, ne 68 st, Seattle For May long weekend, I will be going to Seattle to visit jsalvatier. It's a kilomeetup because it's more than a meetup but not really a megamee...
5030b2d8-38c6-4939-902f-aeaafb759adb
trentmkelly/LessWrong-43k
LessWrong
Linkpost: Github Copilot productivity experiment > We recruited 95 professional developers, split them randomly into two groups, and timed how long it took them to write an HTTP server in JavaScript. One group used GitHub Copilot to complete the task, and the other one didn’t. We tried to control as many factors as we...
1a81c0a9-1dcd-41e1-8c3f-1de24cbf09cb
StampyAI/alignment-research-dataset/alignmentforum
Alignment Forum
The Goodhart Game > > In this paper, we argue that adversarial example defense papers have, to date, mostly considered abstract, toy games that do not relate to any specific security concern. Furthermore, defense papers have not yet precisely described all the abilities and limitations of attackers that would be rele...
6c941892-5f28-4651-9178-03822eec268f
trentmkelly/LessWrong-43k
LessWrong
Open Thread Fall 2024 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. This is als...
cb5d311a-c0fa-4ceb-9737-331f6cbd2e89
trentmkelly/LessWrong-43k
LessWrong
Personal AI Planning LLMs are getting much more capable, and progress is rapid. I use them in my daily work, and there are many tasks where they're usefully some combination of faster and more capable than I am. I don't see signs of these capability increases stopping or slowing down, and if they do continue I expect ...
54fe2497-fbe6-4d0a-9c4d-14739525ecbc
StampyAI/alignment-research-dataset/lesswrong
LessWrong
Efficient Cross-Domain Optimization **Previously in series**:  [Measuring Optimization Power](/lw/va/measuring_optimization_power/) Is Deep Blue "intelligent"?  It was powerful enough at optimizing chess boards to defeat Kasparov, perhaps the most skilled chess player humanity has ever fielded. A bee builds hives,...
0628cbb3-c800-4469-bf6f-01b7d7cabf32
trentmkelly/LessWrong-43k
LessWrong
[SEQ RERUN] Archimedes's Chronophone Today's post, Archimedes's Chronophone was originally published on March 23, 2007. A summary (taken from the LW wiki): > Imagine that Archimedes of Syracuse invented a device that allows you to talk to him. Imagine the possibilities for improving history! Unfortunately, the device...
0d81436c-2f23-45d7-915b-09952ac8b8c2
trentmkelly/LessWrong-43k
LessWrong
More Questions about Trees This is a follow-up to his previous Questions about Trees which was an inquiry "about the puzzling diversity of forest trees." This one looks into questions like: * Why do the leaves of trees have such varied shapes? * Why are trees so tall? * Why are trees trees (rather than other kind...
c4e3c690-1738-4754-be87-744dfa0ec8c0
StampyAI/alignment-research-dataset/arxiv
Arxiv
Imitation Learning from Observations by Minimizing Inverse Dynamics Disagreement 1 Introduction --------------- A crucial aspect of intelligent robots is their ability to perform a task of interest by imitating expert behaviors from raw sensory observations [[5](#bib.bib5)]. Towards this goal, GAIL [[16](#bib.bib16...
25227ee1-76bf-4aa3-8c27-8f408f65a75f
StampyAI/alignment-research-dataset/special_docs
Other
Trusted Partners FEBRUARY 2021 Trusted Partners Human-Machine Teaming and the Future of Military AI CSET Issue Brief AUTHORS Margarita Konaev Tina Huang Husanjot Chahal Center for Security and Emerging Technology | 2 Executive Summary The Department of Defense wants to harness AI-enabled tools and systems to suppo...
3512f549-f73f-467c-8d2e-70c81eee6b8a
StampyAI/alignment-research-dataset/lesswrong
LessWrong
AI cooperation in practice You know that [automated proof verifiers](http://en.wikipedia.org/wiki/Proof_verification) exist, right? And also that programs can [know their own source code](http://en.wikipedia.org/wiki/Quine_(computing))? Well, here's a puzzle for you: *Consider a program A that knows its own source c...
8f3318c9-2636-4051-8cef-72bebdade851
trentmkelly/LessWrong-43k
LessWrong
Rationality Power Tools Summary: Rationalists should win; however, it could take a really long time before a technological singularity or uploading provide powerful technology to aid rationalists in achieving their goals. It's possible today to create assistant computer software to help direct human effort and provide...
03c3d448-1e1e-475c-b88e-330fcf12f99b
trentmkelly/LessWrong-43k
LessWrong
The robust beauty of improper linear models It should come as no surprise to people on this list that models often outperform experts. But these are generally finely calibrated models, integrating huge amounts of data, so this seems less surprising. How can the poor experts compete against that? But sometimes the mod...
bf6c1592-a018-4c39-8595-3a5ba8b03ed3
StampyAI/alignment-research-dataset/eaforum
Effective Altruism Forum
Upcoming speaker series on emerging tech, national security & US policy careers *There is an upcoming virtual speaker series on emerging tech, national security, and US policy careers and wanted to share this opportunity. I’ve heard some of these speakers before think the series could be* *really helpful for anyone in...
2053fa36-ef86-47eb-90c8-5ea94719fba0
trentmkelly/LessWrong-43k
LessWrong
Paphos Published June 2, 2025. "The first thing you need to know about Cyprus is that everything is Hotel. School is Hotel. Restaurant is Hotel. Home is Hotel." [1] Leptos Estates owns everything on the island, as well as NUP, so naturally NUP is in a hotel. Because Leptos Estates is in the business of making and ru...
a3da4f70-16f7-47be-a8b4-f92471fb4ac8
trentmkelly/LessWrong-43k
LessWrong
Edward Nelson claims proof of inconsistency in Peano Arithmetic We've discussed Edward Nelson's beliefs and work before. Now, he claims to have a proof of a contradiction in Peano Arithmetic; which if correct is not that specific to PA but imports itself into much weaker systems. I'm skeptical of the proof but haven't...
03287a0c-48ff-465e-a1dd-14557f64eb98
StampyAI/alignment-research-dataset/aisafety.info
AI Safety Info
What are the different possible AI takeoff speeds? In a slow takeoff AI capabilities improve gradually, giving us plenty of time to adapt. In a moderate takeoff we might see accelerating progress, but we still won’t be caught off guard by a dramatic change. In a fast or hard takeoff, AI quickly goes from being not ver...
0e034695-c48a-4a12-bd5c-01f5ed1bac35
StampyAI/alignment-research-dataset/lesswrong
LessWrong
Why no major LLMs with memory? One thing that I'm slightly puzzled by is that an obvious improvement to LLMs would be adding some kind of long-term memory that would allow them to retain more information than fits their context window. Naively, I would imagine that even just throwing some recurrent neural net layers ...
530aad21-40f2-48d4-a66e-1e07de3b5c7c
trentmkelly/LessWrong-43k
LessWrong
When Can You Use a Travel Adapter? We're currently in the UK, where outlets look like: Since we're from the US, none of the gear we have will plug into these sockets. What if we use a travel adapter? plug intentionally partially inserted to show pins This accepts the pins from a US plug, and connects them dire...
df32c0b9-534b-47ee-ac70-071642239b3f
trentmkelly/LessWrong-43k
LessWrong
What is life? (This post is part of the attempt to do a write up after every Cambridge_UK meetup, of something raised at the meetup.)   What is life?   Eliezer, in "The First World Takeover", talks about the search for objects with increasing numbers of bits of functional complexity.  He framed the difference betw...
9f14e157-ac0a-440c-8cea-654675fa2a78
trentmkelly/LessWrong-43k
LessWrong
The Advantages of Being Technical > Someone told me that each equation I included in the book would halve the sales. I therefore resolved not to have any equations at all. — Stephen Hawking, in his book A Brief History of Time Whoever gave this advice to Mr. Hawking is probably partly responsible for the success, fi...
fc0e7f16-9736-4cd6-9349-12f304cd0c0b
trentmkelly/LessWrong-43k
LessWrong
Intuition In the previous post, I've said, > [...] in both schemes, Factored Cognition includes a step that is by definition non-decomposable. In Ideal Debate, this is step is judging the final statement. In HCH, it is solving a problem given access to the subtrees. This step is also entirely internal to the human. ...
a0d4c36b-b3c9-485b-abef-685e2d5e8e93
trentmkelly/LessWrong-43k
LessWrong
It is easy to expose users' secret web habits, say researchers
2ea8b97b-48a4-4dcb-8c6c-4134de6cff3f
trentmkelly/LessWrong-43k
LessWrong
On learning difficult things I have been autodidacting quite a bit lately. You may have seen my reviews of books on the MIRI course list. I've been going for about ten weeks now. This post contains my notes about the experience thus far. Much of this may seem obvious, and would have seemed obvious if somebody had tol...
8a171fa4-63fc-424a-8d6a-25c921592639
trentmkelly/LessWrong-43k
LessWrong
AI Alignment Prize: Super-Boxing ,,,,,,,,, In this post I propose an (allegedly...) secure AI boxing method. For reasons elaborated at the end of the post, this proposal is probably unsafe, and is certainly not practical. Still, I've elected to share my thoughts on this in the hope of receiving valuable feedback, a...
6b3e6138-c5d0-4c5d-8ff1-336b7945de1e
StampyAI/alignment-research-dataset/eaforum
Effective Altruism Forum
How to create curriculum for self-study towards AI alignment work? Hello! I discovered EA and alignment work in 2022. I currently work in the humanities field, have taken math & cs college-level courses but have no relevant degrees. I want to create a curriculum for myself to follow in 2023 so that I can eventually a...
1ac924e8-6a66-45fa-b164-e233b1502611
trentmkelly/LessWrong-43k
LessWrong
Infra-Miscellanea This post is a writeup of some miscellaneous results that don't cohere into a single big post, but might possibly be of interest. Also, in mostly unrelated news, this stuff can be a bit tricky to learn for the first time. So, I was thinking of assembling an infra version of the fixpoint exercise she...
8b19b038-4fc5-41fe-b0a4-708aa5abbfd2
trentmkelly/LessWrong-43k
LessWrong
Preferences from (real and hypothetical) psychology papers Research projects I'm planning to start two research projects on model splintering/reward generalisation and learning the preferences of irrational agents. Within those projects, I'm aiming to work on subprojects that are: 1. Posed in terms that are familia...
e7f7691c-2d82-4f4e-a1bc-9a7ac46ac878
LDJnr/LessWrong-Amplify-Instruct
LessWrong
"I am writing a post on scenario-based planning, and thought it would be useful to get some example scenario symptoms on the topic of near-term AGI. So the scenario-based planning thought exercise goes like this: Imagine it is ten years from now and AGI has happened (for the thought experiment just imagine you are look...
c4aa3773-da25-434d-b4c9-64c8e608b83b
trentmkelly/LessWrong-43k
LessWrong
Truthseeking, EA, Simulacra levels, and other stuff Elizabeth I've been on this big kick talking about truthseeking in effective altruism. I started with vegan advocacy because it was the most legible, but now need to move on to the deeper problems. Unfortunately those problems are still not that legible, and I end up...
9e091c80-caf7-4952-bd02-f8a8e30a457a
trentmkelly/LessWrong-43k
LessWrong
R0 Is Not Counterfactual Someone who helps organize an N95-required dance recently wrote to a group of organizers: > R0 is the number of people that an infected individual is likely to infect. R0 was 5.4 in Dec 2022. ... People who assume increased risk for themselves are also assuming increased risk for 5.4 other ...
01622c58-1a4f-4539-b833-6041e5b4d4ff
trentmkelly/LessWrong-43k
LessWrong
Current UK government levers on AI development This is a link post for this collection of current UK government levers on AI development. At the end of 2022, I made a collection of information on current UK government levers on AI development, focused on levers which seem to me to have potentially significant implica...
3b17c6bf-967a-4269-b322-1d0442e68dd4
trentmkelly/LessWrong-43k
LessWrong
[Linkpost] Why AI Safety Camp struggles with fundraising (FBB #2) Crossposted on The Field Building Blog and the EA forum.
b3671b14-bb34-4dcd-9f4d-3690c25798f9
trentmkelly/LessWrong-43k
LessWrong
What exercises go best with 3 blue 1 brown's Linear Algebra videos? 3 blue 1 brown is a youtube channel that teaches math concepts. I've found it a much better introduction than other resources I've looked at. The Essence of Calculus series was particularly good. But one issue is that, while the videos come with a fe...
ae9f50b5-7f7a-4927-9d76-15f061ba3327
trentmkelly/LessWrong-43k
LessWrong
No Summer Harvest: Why AI Development Won't Pause Cleo Nardo recently highlighted Yudkowsky and others discussing an "AI Summer Harvest" (or simply "Summer Harvest"). This potential near-future era would involve a pause in research on state-of-the-art (SOTA) models, enabling us to fully enjoy the economic benefits of ...
111a8c63-cf2a-4374-8901-d38d68868aa7
StampyAI/alignment-research-dataset/alignmentforum
Alignment Forum
Inner and outer alignment decompose one hard problem into two extremely hard problems *TL;DR:* One alignment strategy is to 1) capture “what we want” in a loss function to a very high degree (“robust grading”), 2) use that loss function to train the AI, and 3) get the AI to exclusively care about optimizing that objec...
2a7b32cf-6385-430d-8308-b23a7193476c
trentmkelly/LessWrong-43k
LessWrong
Overview: AI Safety Outreach Grassroots Orgs We’ve been looking for joinable endeavors in AI safety outreach over the past weeks and would like to share our findings with you. Let us know if we missed any and we’ll add them to the list. For comprehensive directories of AI safety communities spanning general interest,...
6eb61f35-9692-4f9d-981f-4b95fb7b371c
StampyAI/alignment-research-dataset/special_docs
Other
Statement on AI Risk ### Contents [Statement](#open-letter)[Signatories](#signatories)[Sign the statement](#sign)###### **AI experts, journalists, policymakers, and the public are increasingly discussing a broad spectrum of important and urgent risks from AI. Even so, it can be difficult to voice concerns about some ...
fc21320c-963b-4140-a53b-cc1445372a50
trentmkelly/LessWrong-43k
LessWrong
AI as a Civilizational Risk Part 5/6: Relationship between C-risk and X-risk The general model presented in the previous parts suggests that the US and potentially other parts of the world will suffer some form of civilizational collapse, potentially as early as 10 years from now. In the absence of decisive action to...
087eb05a-caa3-4e7f-9398-818112de7763
LDJnr/LessWrong-Amplify-Instruct
LessWrong
"Once upon a time, in my reckless youth, when I knew not the Way of Bayes, I gave a Mysterious Answer to a mysterious-seeming question. Many failures occurred in sequence, but one mistake stands out as most critical: My younger self did not realize that solving a mystery should make it feel less confusing. I was trying...
287af102-f881-43b0-97b8-f735337863f6
StampyAI/alignment-research-dataset/eaforum
Effective Altruism Forum
Foundations for a Longtermist Foreign Policy "[There are some decisions that are made only by governments, and some of those decisions are highly consequential. They include decisions like going to war, or what weapon systems will be fielded, or how technologies will be embedded within larger critical systems. It mak...
345d55c5-d440-46a5-af9e-35a20dcfd317
LDJnr/LessWrong-Amplify-Instruct
LessWrong
"Summary: I think it’s important for surveys about the future of technology or society to check how people's predictions of the future depend on their beliefs about what actions or responsibilities they and others will take on. Moreover, surveys should also help people to calibrate their beliefs about those responsibil...
44784e9b-d281-4bc8-bea6-d29d61142eb8
trentmkelly/LessWrong-43k
LessWrong
Externalized reasoning oversight: a research direction for language model alignment Summary In this post, I’ll present a research direction that I’m interested in for alignment of pretrained language models. TL;DR: Force a language model to think out loud, and use the reasoning itself as a channel for oversight. If t...
6e892c22-9172-4f50-baa1-5d12e0b0e63e
trentmkelly/LessWrong-43k
LessWrong
Content and Takeaways from SERI MATS Training Program with John Wentworth Introduction I participated in the training program for John Wentworth’s stream of SERI MATS, and overall, it was a very good experience! This post is meant to convey the content of the workshops, so that others can carry out the exercises on th...
dc30fab7-27f1-496b-83b3-5f0cc2fe0f77
trentmkelly/LessWrong-43k
LessWrong
Barcoding LLM Training Data Subsets. Anyone trying this for interpretability? Main point: seeking information on this potential strategy for improving neural network interpretability, alignment, and reliability.   Segmenting training data and injecting unique "barcoding" tokens into each data subset. These unique toke...
4bbc77a7-e7ac-4ae3-aee8-50f263bc5805
trentmkelly/LessWrong-43k
LessWrong
Bayesian Adjustment Does Not Defeat Existential Risk Charity (This is a long post. If you’re going to read only part, please read sections 1 and 2, subsubsection 5.6.2, and the conclusion.) 1. Introduction Suppose you want to give some money to charity: where can you get the most bang for your philanthropic buck? On...
aaace824-7f56-4927-9b73-0d0dda41f846
trentmkelly/LessWrong-43k
LessWrong
Meetup : Moscow, Memory Tricks Discussion article for the meetup : Moscow, Memory Tricks WHEN: 24 November 2013 04:00:00PM (+0400) WHERE: Russia, Moscow, ulitsa L'va Tolstogo 16 We will have a report about how memory works, tips and tricks to remember, spaced repetition and Anki. There will be a discussion section...
1104a7de-7554-4820-adaa-7f5db8a9fd9a
trentmkelly/LessWrong-43k
LessWrong
"A Note on the Compatibility of Different Robust Program Equilibria of the Prisoner's Dilemma" > We study a program game version of the Prisoner's Dilemma, i.e., a two-player game in which each player submits a computer program, the programs are given read access to each other's source code and then choose whether to ...
25c1edd9-d31b-428f-8b4c-6db43ff348bf
trentmkelly/LessWrong-43k
LessWrong
Trial and error in policy making I’m working in government in the UK, and interested in rationality in policy making: at the individual level, but also how we can build systems (in terms of requirements during policy creation, creation of independent bodies, incentive structures for officials/ministers…) that encourag...
f35c0843-e1d4-4255-bc8f-57144cb879ad
trentmkelly/LessWrong-43k
LessWrong
More Cryonics Probability Estimates There are a lot of steps that all need to go correctly for cryonics to work. People who had gone through the potential problems, assigning probabilities, had come up with odds of success between 1:4 and 1:435. About a year ago I went through and collected estimates, finding other pe...
5a5d3604-7af2-4348-b4d8-2ac017661448
trentmkelly/LessWrong-43k
LessWrong
Understanding the two-head strategy for teaching ML to answer questions honestly This post is the result of my attempts to understand what’s going on in these two posts from summer 2021: Paul Christiano: https://www.alignmentforum.org/posts/QqwZ7cwEA2cxFEAun/teaching-ml-to-answer-questions-honestly-instead-of Evan H...
5c38464c-346d-4a35-8065-51cd32bb4a94
StampyAI/alignment-research-dataset/alignmentforum
Alignment Forum
The blue-minimising robot and model splintering A long time ago, Scott introduced the [blue-minimising robot](https://www.lesswrong.com/posts/hQHuXuRGZxxWXaPgg/the-blue-minimizing-robot): > > Imagine a robot with a turret-mounted camera and laser. Each moment, it is programmed to move forward a certain distance an...
0fe52864-4c56-45b1-84b9-f74e91611f4d
trentmkelly/LessWrong-43k
LessWrong
AGI safety from first principles: Introduction This is the first part of a six-part report called AGI safety from first principles, in which I've attempted to put together the most complete and compelling case I can for why the development of AGI might pose an existential threat. The report stems from my dissatisfacti...
d5436a86-b159-4a8d-b09c-ee702031aaab
trentmkelly/LessWrong-43k
LessWrong
Values Are Real Like Harry Potter Imagine a TV showing a video of a bizarre, unfamiliar object - let’s call it a squirgle. The video was computer generated by a one-time piece of code, so there's no "real squirgle" somewhere else in the world which the video is showing. Nonetheless, there's still some substantive sens...
0eb44662-901d-4475-ae84-76dc52985429
trentmkelly/LessWrong-43k
LessWrong
Meetup : DC Meetup: Boardgames Discussion article for the meetup : DC Meetup: Boardgames WHEN: 18 November 2012 03:00:00PM (-0500) WHERE: National Portrait Gallery We'll be meeting to play games and hang out, and possibly try a new variation of Zendo. Discussion article for the meetup : DC Meetup: Boardgames
2a89a402-260f-427a-b5cd-5d010aa3ea3d
trentmkelly/LessWrong-43k
LessWrong
Thiel on secrets and indefiniteness Some excerpts from Peter Thiel's 2014 book Zero to One that I've repeatedly come back to over the years: > [...] Why has so much of our society come to believe that there are no hard secrets left? It might start with geography. There are no blank spaces left on the map anymore. If ...
692ed83e-7794-487f-8e4c-ca72525a05fa
trentmkelly/LessWrong-43k
LessWrong
How do AI timelines affect how you live your life? This question is more about personal decision-making rather than for example deciding to work on AI safety for altruistic reasons. If I were thinking about this from a purely selfish perspective, it seems pretty likely that if I expect transformative AI to arrive in ...
5477d944-d3e3-4392-948d-1358cffe6a69
trentmkelly/LessWrong-43k
LessWrong
Does a site exist for keeping track of casual wagers with others? I run into two situations where I want to wager: * With a friend, about some future outcome * With someone I'm having a debate with on the internet, as a way to settle the debate and test our beliefs With friends, I've been tracking these wagers wit...
721635fe-4b84-4507-8b37-57243f42167c
trentmkelly/LessWrong-43k
LessWrong
Meetup : Vienna meetup Discussion article for the meetup : Vienna meetup WHEN: 28 September 2012 07:00:00PM (+0200) WHERE: Favoritenstraße 55, 110 Vienna, Austria Let's have a meetup in Vienna! The Cafe Illy at Favoritenstraße 55 is on the southeast corner of Favoritenstraße and Johannitergasse. You can take the u...
6f7c5466-27e2-4a68-9ab0-3a21bce08439
trentmkelly/LessWrong-43k
LessWrong
Dead men tell tales: falling out of love with SIA SIA is the Self Indication Assumption, an anthropic theory about how we should reason about the universe given that we exist. I used to love it; the argument that I've found most convincing about SIA was the one I presented in this post. Recently, I've been falling out...
e0dab0f3-4a03-4c5a-9a1d-004cd09af332
trentmkelly/LessWrong-43k
LessWrong
The Byronic Hero Always Loses Why do we root for the antiheroes? Walter White, Light Yagami, Lucifer Morningstar, Doctor Frankenstein... It seems to be natural to root for the bad guys, and perhaps even more natural for rationalists. “I must confess I have always had some sympathy with villains. Heroism makes fine e...
07899802-e91e-4f2b-be1d-f625005f24eb
StampyAI/alignment-research-dataset/eaforum
Effective Altruism Forum
Chilean AIS Hackathon Retrospective TL;DR ----- * We hosted an AI Safety “Thinkathon” in Chile. We had participation from 40 students with differing skill levels and backgrounds, with groups totalling 13 submissions. * We see potential in: + Similar introductory events aiming for a broad audience + Collaborating mo...
3c71bc1a-e8e1-4567-9f47-4a2c8089f70c
trentmkelly/LessWrong-43k
LessWrong
A Strange Argument about the potential Importance of Multiverse Theory Multiverse Theory is the science of guessing at the shape of the state space of all which exists, once existed, will exist, or exists without any temporal relation to our present. Multiverse theory attempts to model the unobservable, and it is very...
2f38dc46-ee03-4c28-8d25-e9b69e3051bf
trentmkelly/LessWrong-43k
LessWrong
Algorithmic Similarity The Problem If someone asked you and me to each write a python program to solves some novel problem then it wouldn't be surprising if our solutions looked quite different. If we then tried explaining to the other person the way our own program worked, one thing that might happen is that we would...
edae8b7b-c6cb-4bab-97b2-e8eadc230d4a
StampyAI/alignment-research-dataset/special_docs
Other
Shaping the Terrain of AI Competition Shaping the Terrain of AI CompetitionJUNE 2020 AUTHOR Tim Hwang Center for Security and Emerging Technology2Established in January 2019, the Center for Security and Emerging Technology (CSET) at Georgetown’s Walsh School of Foreign Service is a research organization fo - cused ...
1cae0fce-a828-4a42-af9c-9c53fb72ee4d
StampyAI/alignment-research-dataset/lesswrong
LessWrong
Bing chat is the AI fire alarm *Epistemic status*: Mostly speculative. Panicking and shouting "Wolf" while everyone else is calm is a risky move, status-wise. The good thing is, I don't have any status, so I volunteer to be one of those weirdos who panic when everyone else is calm with some hope it could trigger a [r...
c0de492c-1057-458b-8d29-81868cce76cb
LDJnr/LessWrong-Amplify-Instruct
LessWrong
"Followup to: Are Your Enemies Innately Evil?, Policy Debates Should Not Appear One-Sided "The financial crisis is not the crisis of capitalism. It is the crisis of a system that has distanced itself from the most fundamental values of capitalism, which betrayed the spirit of capitalism." -- Nicolas Sarkozy During the ...
c4423c3d-d92d-4c2c-a4dd-8cfac8ee875b
StampyAI/alignment-research-dataset/blogs
Blogs
fully aligned singleton as a solution to everything fully aligned singleton as a solution to everything --------------------------------------------------- when i misplace my keys, [building a fully aligned superintelligent singleton (FAS)](outlook-ai-risk-mitigation.html) that takes over the world and realizes my v...
8b324ebb-9ec2-424a-b3f1-cdcca7debb8f
StampyAI/alignment-research-dataset/lesswrong
LessWrong
Activation additions in a simple MNIST network Abstract ======== Team Shard's recent activation addition methodology for [steering GPT-2-XL](https://www.lesswrong.com/posts/5spBue2z2tw4JuDCx/steering-gpt-2-xl-by-adding-an-activation-vector) provokes many questions about what the structure of internal model computatio...
5f11d0b4-60cd-42d7-b407-37c514c18b81
trentmkelly/LessWrong-43k
LessWrong
The appeal of fictional conflict Robert Wiblin asks why stories celebrate conflict rather than compromise: > As I was watching the film Avatar and the cinemagoers around me were cheering on the Na’vi heroes in their fight against human invaders, I couldn’t help but wonder how many of us would actually want to live al...
043574cf-07a3-4198-9c88-5e540a76bb5f
trentmkelly/LessWrong-43k
LessWrong
Meetup : Utrecht: Debiasing techniques Discussion article for the meetup : Utrecht: Debiasing techniques WHEN: 21 September 2014 02:00:00PM (+0200) WHERE: Utrecht, Netherlands We have biweekly meetups in a pub in Utrecht, near Central Station. This time we will discuss debiasing techniques. For details, please look...
7a7ae701-25d8-4eb0-a5b7-adae70f9cc9f
trentmkelly/LessWrong-43k
LessWrong
Vows & Declaration This is the eighth and final section of the wedding ceremony of Ruby & Miranda. See the Sequence introduction for more info.  ---------------------------------------- IMAGE 7 PROJECTED ON PLANETARIUM Brienne:  We’ve spoken of the darkness, and of our hope to overcome it. We’ve spoken of the light...
4cccb26a-5b62-42b9-8da8-ecde6f6de796
trentmkelly/LessWrong-43k
LessWrong
Retroactive If-Then Commitments An if-then commitment is a framework for responding to AI risk: "If an AI model has capability X, then AI development/deployment must be halted until mitigations Y are put in place." As an extension of this approach, we should consider retroactive if-then commitments. We should behave ...
05fcd89d-0cb4-4793-b4cb-42e7adb56b3c
trentmkelly/LessWrong-43k
LessWrong
AIXI and Existential Despair It has been observed on Less Wrong that a physical, approximate implementation of AIXI is unable to reason about its own embedding in the universe, and therefore is apt to make certain mistakes: for example, it is likely to destroy itself for spare parts, and is unable to recognize itself ...
d6819805-d7ff-466e-9413-04642a65b8b8
trentmkelly/LessWrong-43k
LessWrong
Deference and Decision-Making You're talking with a leading AI researcher about timelines to AGI. After walking through various considerations, they tell you: "Looking at the current pace of development, I'd put the probability of AGI within 3 years at about 70%. But here's the thing—I know several really sharp resear...
e2a8806f-0fa5-4fc2-86d7-b1c7437cfaa3
trentmkelly/LessWrong-43k
LessWrong
Apprenticeship Online If humans stopped imitating, civilization would vanish. It is 6:37 in the morning as I’m sitting down to write this, late in August, and I and the three-year-old are down in the old stable we use as study. As I look over my notes, she observes me. Mist is leaking through a crack in the window. T...
00efd2b6-43cb-444e-80c6-722ac2281784
trentmkelly/LessWrong-43k
LessWrong
Open thread, September 8-14, 2014 If it's worth saying, but not worth its own post (even in Discussion), then it goes here. 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. 3. Open Threads should be posted in Discussion, and n...
da01588a-2cb5-4ed1-a7e2-a902e4dfd883
LDJnr/LessWrong-Amplify-Instruct
LessWrong
"I am teaching a class, and I write upon the blackboard three numbers: 2-4-6. “I am thinking of a rule,” I say, “which governs sequences of three numbers. The sequence 2-4-6, as it so happens, obeys this rule. Each of you will find, on your desk, a pile of index cards. Write down a sequence of three numbers on a card, ...
a433be4c-ed5c-45ab-95a8-78ab05313fec
trentmkelly/LessWrong-43k
LessWrong
Sentinel's Global Risks Weekly Roundup #15/2025: Tariff yoyo, OpenAI slashing safety testing, Iran nuclear programme negotiations, 1K H5N1 confirmed herd infections. Executive summary The Trump administration backtracked from his tariff plan, reducing tariffs from most countries to 10%, but imposing tariffs of 145% on...
0fb4ed4f-81e0-490d-9830-ee432dd3540b
trentmkelly/LessWrong-43k
LessWrong
Group Rationality Diary, September 1-15 This is the public group instrumental rationality diary for September 1-15. > 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 s...
d007f938-eba1-45f5-995b-f58e27a5e7ab
trentmkelly/LessWrong-43k
LessWrong
GPT-4 Specs: 1 Trillion Parameters? I know that OpenAI has tried to keep GPT-4's specs under wraps, but I've also heard some reports that anonymous sources are converging at a trillion parameters for the model. Can anyone confirm or deny, or qualify that statement? https://the-decoder.com/gpt-4-has-a-trillion-paramete...
7ded246d-b046-4a1f-b792-9830dc62b271
StampyAI/alignment-research-dataset/blogs
Blogs
Using Unity to Help Solve Intelligence #### A wide range of environments In the pursuit of artificial general intelligence (AGI), we seek to create agents that can achieve goals in a wide range of environments. As our agents master the environments we create, we must continually create new environments that probe as-...
9e7b3a56-866a-4607-895c-1bfadb327b09
trentmkelly/LessWrong-43k
LessWrong
Knightian Uncertainty from a Bayesian perspective Some people have maintained that there are events to which there's no rational basis for assigning probabilities. For example, John Maynard Keynes wrote of "uncertainty" in the following sense: > "By `uncertain' knowledge, let me explain, I do not mean merely to disti...
b8c998ac-4186-4882-bd6b-1fdf9c8d794a
trentmkelly/LessWrong-43k
LessWrong
Emergence of superintelligence from AI hiveminds: how to make it human-friendly? "Hive mind" may not be the ideal term. What I'm really envisioning is something modelled on the existing relationship between human users and frontier LLMs, but with the "users" actually just being LLMs too, or simple programs that mediat...
1c3c9c1f-4579-448a-9402-0796e9a716f7
trentmkelly/LessWrong-43k
LessWrong
Dealing with UFO claims A common response in the recent LessWrong threads about UFO's is rationalists immediately going into a state of wanting to translate the news into probabilities of the existence of aliens instead of taking the facts for what they are and thinking about what should happen based on the revealed f...
844af6e6-d5d2-4212-ab65-0195174238af
trentmkelly/LessWrong-43k
LessWrong
COEDT Equilibria in Games From Jessica's earlier post about conditional oracles, the question was asked: what is the equilibrium concept for games with more than one COEDT agent? This post will (partially) answer that question, and provide a link to a tool to visualize 2-player 2-move COEDT equilibria sets. To beg...
7c74241a-4f19-4c54-adb4-bc09a2867150
trentmkelly/LessWrong-43k
LessWrong
Let's Design A School, Part 2.3 School as Education - The Curriculum (Phase 2, Specific) In the previous post, we outlined three phases that students would go through, where each student matriculated through them at their own speed. Phase 1 was literacy and numeracy. Phase 2 was core civilizational requirements and ...
f7ba3649-20d1-470e-929e-545f8393e56d
trentmkelly/LessWrong-43k
LessWrong
UDT1.01: Plannable and Unplanned Observations (3/10) The Omnipresence of Unplanned Observations Time to introduce some more concepts. If an observation is "any data you can receive which affects your actions", then there seem to be two sorts of observations. A plannable observation is the sort of observation where yo...
eeb5ada0-fee0-420a-99a0-bbf22946898d
StampyAI/alignment-research-dataset/special_docs
Other
Fundamental lssues Of Artificial Intelligence [] Volume 376 Synthese LibraryStudies in Epistemology, Logic, Methodology, and Philosophy of Science Editor-in-Chief Otávio Bueno Department of Philosophy, University of Miami, CORAL GABLES, Florida, USA More information about this series at http://​www.​springer.​co...
09fc4e02-e239-4896-8b0d-996e2e828699
trentmkelly/LessWrong-43k
LessWrong
Code Generation as an AI risk setting Historically, it has been difficult to persuade people of the likelihood of AI risk because the examples tend to sound “far-fetched” to audiences not bought in on the premise. One particular problem with many traditional framings for AI takeover is that most people struggle to ima...
5923b715-9b51-49f4-839b-f53de35d4898
StampyAI/alignment-research-dataset/eaforum
Effective Altruism Forum
A Theologian's Response to Anthropogenic Existential Risk Hi all, This is very much someone outside the bailiwick of this forum looking in, but I was told it could be interesting to share this article I wrote recently.  I'm a Catholic priest, with a prior background in Electronic Engineering, currently working on a ...
c227efda-1e0f-4a52-ad30-f09ddd5fb7b4
trentmkelly/LessWrong-43k
LessWrong
The one where Quirrell is an egg Egg-based parody of HPMOR on r/HPMOR. I don't know why this exists. Spoilers, you have been warned. How to describe it? It's some combination of the following two quotes. > “The observational study is taller than a dozen men! It reaches journalists so quickly that it can melt common ...
92b0693e-9787-4337-99f4-faeb630caac2
trentmkelly/LessWrong-43k
LessWrong
Announcing the Complice Less Wrong Study Hall (If you're familiar with the backstory of the LWSH, you can skip to paragraph 5. If you just want the link to the chat, click here: LWSH on Complice) The Less Wrong Study Hall was created as a tinychat room in March 2013, following Mqrius and ShannonFriedman's desire to c...
b79f0238-92be-443e-9912-c4d8ee001125
trentmkelly/LessWrong-43k
LessWrong
[Link] IDA 9/14: The Scheme This is a linkpost for https://app.grasple.com/#/level/1669 Every Thursday for 4 weeks, we will be posting lessons about Iterated Distillation and Amplification. They're largely based on Paul Christiano's sequence here on LW. He graciously allowed us to use his work. This is the third sec...