id
stringlengths
36
36
source
stringclasses
15 values
formatted_source
stringclasses
13 values
text
stringlengths
2
7.55M
4225f277-bda7-4df6-a612-1886a27b1dde
StampyAI/alignment-research-dataset/eaforum
Effective Altruism Forum
Map of maps of interesting fields I love seeing cool visualizations (maps) of interesting intellectual fields. I’m also a big fan of making [lists](https://forum.effectivealtruism.org/posts/6trt8mTsKfqJJbfJa/suggestion-eas-should-post-more-summaries-and-collections)! Accordingly, I compiled a bunch of maps of fields ...
b24806d1-559b-475e-bf96-34c5fd631b3f
trentmkelly/LessWrong-43k
LessWrong
On what research policymakers actually need 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...
598970ee-20b4-4993-a0d0-d5c12eb2f835
trentmkelly/LessWrong-43k
LessWrong
Rhetoric for the Good The topics of rationality and existential risk reduction need their own Richard Dawkins. Their own Darwin. Their own Voltaire. Rhetoric moves minds. Students and masochists aside, people read only what is exciting. So: Want to make an impact? Be exciting. You must be heard before you can turn h...
d6269362-67c5-436f-8cf6-33e174a9ca52
trentmkelly/LessWrong-43k
LessWrong
Meetup : Melbourne Rationality Dojo, January - Lightning Talks Discussion article for the meetup : Melbourne Rationality Dojo, January - Lightning Talks WHEN: 15 January 2017 04:00:00PM (+1100) WHERE: Janet Lady Clarke Memorial Rotunda, Queen Victoria Garden, Melbourne The next meetup for Less Wrong Melbourne will ...
34246193-79f8-4c2e-b9da-7fbfb7862212
trentmkelly/LessWrong-43k
LessWrong
Something to Protect In the gestalt of (ahem) Japanese fiction, one finds this oft-repeated motif:  Power comes from having something to protect. I'm not just talking about superheroes that power up when a friend is threatened, the way it works in Western fiction.  In the Japanese version it runs deeper than that. I...
e1d2bd7d-ec06-4b05-9da7-9ea4c6f2a201
StampyAI/alignment-research-dataset/alignmentforum
Alignment Forum
If interpretability research goes well, it may get dangerous I've [historically](https://twitter.com/robbensinger/status/1602835145488113664) been pretty publicly [supportive](https://www.lesswrong.com/posts/3pinFH3jerMzAvmza/on-how-various-plans-miss-the-hard-bits-of-the-alignment) of interpretability research. I'm s...
b024d61a-ec9e-4d7c-8e22-2d9460044d4e
trentmkelly/LessWrong-43k
LessWrong
Which animals can suffer? In this recent paper, the author study whether the octopi can suffer. Well, how are you going to answer that? Ask him? It turns out there are two ways you can decide whether the animal can suffer or cannot. The first is simple. You just define that vertebrates can suffer (therefore experimen...
2f4d2bd8-89df-4c49-aef1-ca3cc0ad60f3
trentmkelly/LessWrong-43k
LessWrong
Assuming Nails Tangential followup to Defeating Ugh Fields in Practice. Somewhat related to Privileging the Hypothesis. Edited to add: I'm surprised by negative/neutral reviews. This means that either I'm simply wrong about what counts as interesting, or I haven't expressed my point very well. Based on commenter resp...
24cf8a35-97fe-43cf-a0eb-c105c6f36e4b
trentmkelly/LessWrong-43k
LessWrong
How to Give Coming AGI's the Best Chance of Figuring Out Ethics for Us [Note: this is a slightly edited version of an essay I entered into the AI Impacts essay contest on the Automation of Wisdom and Philosophy - entries due July 14, 2024. Crossposted to the EA Forum.] TL;DR A few possible scenarios are put forth to...
07a91e1e-0e68-421e-b2e8-20a19d33a94e
trentmkelly/LessWrong-43k
LessWrong
Fear of centralized power vs. fear of misaligned AGI: Vitalik Buterin on 80,000 Hours Vitalik Buterin wrote an impactful blog post, My techno-optimism. I found this discussion of one aspect on 80,00 hours much more interesting. The remainder of that interview is nicely covered in the host's EA Forum post.  My techno ...
f5ce067e-9d2b-48dd-916a-53b0ddaf11b8
trentmkelly/LessWrong-43k
LessWrong
Launching the Forecasting AI Progress Tournament I'm excited to let you know that Metaculus is launching the Forecasting AI Progress Tournament, a large-scale forecasting tournament with $50,000 in prizes. Participation is open to all. The tournament is intended to provide a proof of concept for probabilistic forecas...
3aef8140-5516-4e9b-88d7-b19082851b1c
trentmkelly/LessWrong-43k
LessWrong
Alienation and meta-ethics (or: is it possible you should maximize helium?) (Cross-posted from Hands and Cities) In a previous post, I tried to gesture at the possibility of a certain kind of wholeheartedness in ethical life. In this post, I want to examine a question about meta-ethics that seems to me important to t...
6551c5ec-4d22-49f8-8f54-20225f11c32f
trentmkelly/LessWrong-43k
LessWrong
Lying to Save Humanity You are an astronaut on an expedition to Mars when you encounter an Alien. Luckily, it speaks English, and you begin an exchange of cultures: history, art, technology, and food, whereupon the Alien interrupts: “Our species has strong feelings about ice cream flavors. Tell me, what is your favor...
2be0754c-9712-4d54-a0a7-04666eb3d7d4
trentmkelly/LessWrong-43k
LessWrong
Arguments against constructivism (in education)? Seems to me that constructivist education is already "in the water supply" of Less Wrong. My tweet-length definition of constructivism would be "teaching students so that they construct mental gears-level models of the subject, as opposed to just memorizing the teachers...
c0c55095-8605-423a-93aa-709a0e24d9f4
trentmkelly/LessWrong-43k
LessWrong
Covid 7/23: The Second Summit Last week: Covid 7/16: Becoming the Mask The news continues to be cautiously good. Positive test rates are likely peaking. Deaths are increasing slower than one would have expected, and could well peak a few weeks from now without having risen much further. This, in context, is what su...
9a2e7574-6b29-4d7d-bd3b-e775e53deab4
trentmkelly/LessWrong-43k
LessWrong
Speed & Performance is our current top priority TLDR: Fewer bugfixes and support for the next week or two while we are focusing on making the site fast and stable. General shift in direction towards more stable and polished features over a large breadth of slightly buggy features. It's been one week since the launch ...
9b3fd9a4-4327-472d-a206-033d47fa98dd
trentmkelly/LessWrong-43k
LessWrong
My attempt to explain Looking, insight meditation, and enlightenment in non-mysterious terms Epistemic status: pretty confident. Based on several years of meditation experience combined with various pieces of Buddhist theory as popularized in various sources, including but not limited to books like The Mind Illuminate...
44cca6b3-40eb-4b4a-bba7-ab923ad639b4
trentmkelly/LessWrong-43k
LessWrong
Chaotic era: avoid or survive? In the world of the “Three-body problem” novel by Liu Cixin, stable eras, when the planet is rotating just around one sun, and everything is quite predictable, are changing to chaotic eras, when all three suns are involved, and nothing is possible to predict.  In our history, we have rel...
7b637aad-d118-4fd6-8f65-315f199e77ca
LDJnr/LessWrong-Amplify-Instruct
LessWrong
"The dominant model about status in LW seems to be one of relative influence. Necessarily, it's zero-sum. So we throw up our hands and accept that half the community is just going to run a deficit.Here's a different take: status in the sense of worth. Here's a set of things we like, or here's a set of problems for you ...
41961f9c-d43e-45c9-8899-03aae78415ef
trentmkelly/LessWrong-43k
LessWrong
Sam Altman at SlateStarCodex Online Meetup Our online meetup on  November 8, at  10:30 AM Pacific Standard Time, 18:30 UTC, 20:30 Israel Standard Time, will feature Sam Altman, CEO of OpenAI, who will be answering questions. We will then have open video-chat rooms as usual. We hope to focus the Q&A on AI, AGI, GPT-N ...
1462436c-b1f5-4ab2-9ed3-d6b66994140a
trentmkelly/LessWrong-43k
LessWrong
Weekly LW Meetups: Houston, Atlanta, Eugene OR, San Diego, Fort Collins CO, Waterloo, Melbourne There are upcoming irregularly scheduled Less Wrong meetups in: * Houston Meetup: Saturday, 11/19: 19 November 2011 02:25AM * Atlanta Less Wrong Meetup: 19 November 2011 07:00PM * First Eugene Meetup: 20 November 2011 0...
53b2e6e0-2dbc-4220-8437-e66026034698
trentmkelly/LessWrong-43k
LessWrong
Book Launch: The Engines of Cognition The Engines of Cognition is a newly published collection of essays by the LessWrong Community. They're currently available with Prime at Amazon US, where they'll arrive in time for Christmas, and you can read on to learn more about them. About In The Engines of Cognition, the wr...
ca7643a6-d228-46f7-838a-c82ba9ce597a
trentmkelly/LessWrong-43k
LessWrong
The Handbook of Rationality (2021, MIT press) is now open access Of particular interest here is section 4.1 on subjective accounts of probability and section 8.4 which explicitly addresses Knightian uncertainty and its challenges for Bayesian accounts. https://direct.mit.edu/books/oa-edited-volume/5525/The-Handbook-o...
90ea9b1f-e5c5-4860-b32d-a16a3cb30b20
trentmkelly/LessWrong-43k
LessWrong
To first order, moral realism and moral anti-realism are the same thing I've taken a somewhat caricaturist view of moral realism[1], describing it, essentially, as the random walk of a process defined by its "stopping" properties. In this view, people start improving their morality according to certain criteria (self...
85b97544-5d4b-441c-9cbf-845534194b2e
trentmkelly/LessWrong-43k
LessWrong
Start taking drugs randomly I wanted to try a drug. I also wanted to be able to distinguish the direct effects of the drug from the placebo effects of the drug. Most importantly, I wanted to isolate placebo side effects. It seemed there was a chance of real side effects, and that meant that it was basically guaranteed...
a4bed85a-2172-436f-a92f-916f78154b67
trentmkelly/LessWrong-43k
LessWrong
The Tragedy of Group Selectionism Before 1966, it was not unusual to see serious biologists advocating evolutionary hypotheses that we would now regard as magical thinking. These muddled notions played an important historical role in the development of later evolutionary theory, error calling forth correction; like th...
5899b806-2b95-4bc7-a70f-329e3208fc09
StampyAI/alignment-research-dataset/eaforum
Effective Altruism Forum
Update from Campaign for AI Safety Hi! *I am starting to share weekly updates from the Campaign for AI Safety on this forum. Please sign up on the* [*campaign website*](https://www.campaignforaisafety.org/) *to receive these by email.* --- 🌐 The big global news this week is that Rishi Sunak [promised to look int...
1ee05ff4-fb66-4ee2-ae3b-a140f15df708
trentmkelly/LessWrong-43k
LessWrong
Ways that China is surpassing the US [Edit: Changed the post title from the original article title to something more meaningful.] I came across this article today, and I have to agree with it strongly based on my own recent trip to China. The update it triggered for me is the realization that China is genuinely doing...
0aae1624-882a-4b0e-805d-538577dd9cb4
trentmkelly/LessWrong-43k
LessWrong
Morality is Scary I'm worried that many AI alignment researchers and other LWers have a view of how human morality works, that really only applies to a small fraction of all humans (notably moral philosophers and themselves). In this view, people know or at least suspect that they are confused about morality, and are ...
625a82dc-f210-4ddf-9fde-ffd2c5cd3385
trentmkelly/LessWrong-43k
LessWrong
Intuitions about goal-directed behavior One broad argument for AI risk is the Misspecified Goal argument: > The Misspecified Goal Argument for AI Risk: Very intelligent AI systems will be able to make long-term plans in order to achieve their goals, and if their goals are even slightly misspecified then the AI system...
144026a5-cc61-4b38-b1d7-d982c6443896
trentmkelly/LessWrong-43k
LessWrong
Ethical Intuitionism, part 1 I read Michael Huemer’s book Ethical Intuitionism at Bryan Caplan’s suggestion. Before reading it I thought that Ethical Intuitionism seemed an unlikely position and that Bryan and Michael seemed like smart guys, so I hoped I might be persuaded to significantly change my mind. I still beli...
d0356774-d603-4fa2-8484-0618592cf5e6
trentmkelly/LessWrong-43k
LessWrong
Can You Give Support or Feedback for My Program to Alleviate Poverty? Hi LessWrong, Two years ago, when I travelled to Belize, I came up with an idea for a self-sufficient scalable program to address poverty. I saw how many people in Belize were unemployed or getting paid very low wages, but I also saw how skilled th...
ecc4fb2d-be5e-4d62-82e6-5abede975976
StampyAI/alignment-research-dataset/lesswrong
LessWrong
Military AI as a Convergent Goal of Self-Improving AI This is our accepted chapter in the edited volume Forthcoming as a chapter in Artificial Safety And Security ([Roman V. Yampolskiy](https://www.facebook.com/roman.yampolskiy?fref=gs&hc_location=group) , ed.), CRC Press. >Abstract Better instruments to predict the ...
8b19f5ac-ff01-43d2-82a3-f737e95e36eb
trentmkelly/LessWrong-43k
LessWrong
Ambitious Altruistic Software Engineering Efforts: Opportunities and Benefits LessWrong Note I recently published this on the EA Forum, then was recommended to also cross-post here. It's written a bit more for the EA community, but I think of course much of it applies and would be relevant for the communities here too...
495fefee-5e59-436d-a711-539ae64c9b9f
trentmkelly/LessWrong-43k
LessWrong
Accuracy of arguments that are seen as ridiculous and intuitively false but don't have good counter-arguments I think one of the biggest reasons I am worried about AI doom is that there doesn't seem to be very good counter-arguments. Most of them are just bad, and the ones that aren't a bit vague (usually something al...
b38268c1-9d32-40b4-852a-6e7365dfc203
trentmkelly/LessWrong-43k
LessWrong
Developing AI Safety: Bridging the Power-Ethics Gap (Introducing New Concepts) TLDR * One of the most significant challenges facing humanity is the widening gap between our rapidly increasing power and our slower progress (if there is any) in ethical understanding and application. I term this the power-ethics gap, an...
86f86cb9-af95-41c9-88fe-d3dd0bf6a211
StampyAI/alignment-research-dataset/special_docs
Other
How to Explain Individual Classification Decisions Journalof Machine LearningResearch 11(2010)1803-1831 Su bmitted 4/09;Revised 11/09;Published6/10 How toExplain Individual Classification Decisions David Baehrens∗BAEHRENS @CS.TU-BERLIN.DE Timon Schroeter∗TIMON@CS.TU-BERLIN.DE Technische Universit ¨at Berlin Franklinstr...
9f9a8a9f-b871-427a-bf14-cc43eaffb446
trentmkelly/LessWrong-43k
LessWrong
State of play of AI progress (and related brakes on an intelligence explosion) [Linkpost] This time around, I'm sharing a post on Interconnects on why he doesn't believe that the AI 2027 scenario by @Daniel Kokotajlo and many others will come true, and he has 4 sections on this plus a bonus section below: 1. How labs...
1e4dbf4a-55ab-4895-b5fb-fe5b2c7f6830
trentmkelly/LessWrong-43k
LessWrong
Why personal experimentation? Cross posted from Overcoming Bias. Comments there. *** I experiment with many things, as do those around me. Some of this is randomization and explicit records, more is just trying different things, muttering ‘VOI‘ and repeating what felt good. I refer here to everything between a cauti...
7cc648cb-ef85-4ac2-9a0f-61c9a96c8cdd
trentmkelly/LessWrong-43k
LessWrong
A policy guaranteed to increase AI timelines The number of years until the creation of powerful AI is a major input to our thinking about risk from AI and which approaches are most promising for mitigating that risk. While there are downsides to transformative AI arriving many years from now, rather than few years fro...
a2b4b2bf-9a8f-43d5-9d02-a8e595ac0037
StampyAI/alignment-research-dataset/lesswrong
LessWrong
Superintelligence 27: Pathways and enablers *This is part of a weekly reading group on [Nick Bostrom](http://www.nickbostrom.com/)'s book, [Superintelligence](http://www.amazon.com/Superintelligence-Dangers-Strategies-Nick-Bostrom/dp/0199678111). For more information about the group, and an index of posts so far see t...
7006ef93-0172-470b-b419-67c473a1047c
StampyAI/alignment-research-dataset/alignmentforum
Alignment Forum
Factored Cognition *Note: This post (originally published [here](https://ought.org/presentations/factored-cognition-2018-05)) is the transcript of a presentation about a project worked on at the non-profit Ought. It is included in the sequence because it contains a very clear explanation of some of the key ideas behin...
6f6223b3-a453-4772-8d3e-54e1f9c38a37
trentmkelly/LessWrong-43k
LessWrong
Machines vs. Memes 2: Memetically-Motivated Model Extensions [This is post 2 of 3 in a series written as the output for the 2022 AI Safety Camp, for the ‘Impact of Memetics on Alignment’ team, coached by Daniel Kokotajlo and comprising Harriet Farlow, Nate Rush and Claudio Ceruti. You can view the first post in this s...
d1cff7f0-9465-4a7f-952e-04d8d0d78b7e
trentmkelly/LessWrong-43k
LessWrong
Bell's Theorem: No EPR "Reality" Previously in series:  Entangled Photons (Note:  So that this post can be read by people who haven't followed the whole series, I shall temporarily adopt some more standard and less accurate terms; for example, talking about "many worlds" instead of "decoherent blobs of amplitude".) ...
2630b616-ddcb-448e-b705-88594db7d254
trentmkelly/LessWrong-43k
LessWrong
The case for "Generous Tit for Tat" as the ultimate game theory strategy I wish I could title this How “Pacifism” wins at life, since it’d immediately make everything much easier (especially writing this post), but as most well-meaning kids swiftly learn on the playground - the world is not gentle. It’s always bother...
6ed16e40-0d0d-4c5d-ad99-daebe01a1010
trentmkelly/LessWrong-43k
LessWrong
The AI Revolution in Biology An expert in the field (Amelie Schreiber) predicts the convergence of AI models for biology to the state where designing drugs, targeted genetic mutations (which then could be enacted through genetic therapy), or viruses for specific purposes becomes really easy in a few years. (Note: appa...
3c3dc66d-0f7e-4581-96f6-fd976cefb48b
trentmkelly/LessWrong-43k
LessWrong
Alignment from equivariance  Epistemic status: research direction Equivariance is a way of enforcing a user-specified sense of consistency across inputs for a NN. It's a concept from computer vision which I think could be used to align AIs, or rather for building an interface/framework for aligning them. So far it h...
54cdb7cd-ad48-4d5b-a303-5e5f4085a2a7
trentmkelly/LessWrong-43k
LessWrong
Hufflepuff Cynicism on Hypocrisy Epistemic Status: I've been thinking about this for a number of years, looking for steelmen of the position against hypocrisy. I haven't found anything satisfying yet, but maybe you can tell me why hypocrisy is actually bad? Barring that, I'm rather confident in the view expressed here...
137cce10-cce9-457a-b5b2-58ca581adcaa
trentmkelly/LessWrong-43k
LessWrong
Keltham's Lectures in Project Lawful (Not an April fools' joke) If anyone wants to return to one of Keltham's lectures, here's a list with links to all the lectures. In principle, one could also read these without having read the rest of Project Lawful aka Planecrash, I am not sure how easy it would be to follow thou...
3311a001-559f-4dde-9e00-5f6e2c22949a
trentmkelly/LessWrong-43k
LessWrong
Looking and the no-self (This post is a reply to Kaj's "My attempt to explain Looking" and Valentine's Kenshō. The target audience is people who feel like they have a good grasp on the underlying concepts but of course the post might also be interesting for other people) I have the feeling that both you and Valentine...
b044e9b3-8d2d-4963-b90e-292688b82a93
StampyAI/alignment-research-dataset/youtube
Youtube Transcripts
Governing Transformative AI hello my name is mauricio and i'm excited to welcome you to the talk governing transformative ai i'm thrilled to introduce you to marcus andreyan andrew young is head of operations and policy engagement focusing on growing the centre for the governance of ai at the future of humanity instit...
f03a4914-c807-42f7-95dd-879a3f2ae7bf
trentmkelly/LessWrong-43k
LessWrong
Meetup : London Social - Exposure to Direct Sunlight - June 23rd Discussion article for the meetup : London Social - Exposure to Direct Sunlight - June 23rd WHEN: 23 June 2013 02:00:00PM (+0100) WHERE: St James Park, London, SW1H   EDIT: morning of the 23rd, and the weather is looking good enough to chance it. Head...
d8702d6b-83e1-4274-ae41-29295faca256
trentmkelly/LessWrong-43k
LessWrong
Takeoff speeds, the chimps analogy, and the Cultural Intelligence Hypothesis In debates about AI takeoff speeds, the most common empirical example/analogy is that of the transition between chimpanzees and humans.  For example, in this conversation between Paul Christiano and Eliezer Yudkowsky on takeoff speeds, the ch...
232984f3-ad46-4ae8-be62-31f97fedd123
trentmkelly/LessWrong-43k
LessWrong
AI #108: Straight Line on a Graph The x-axis of the graph is time. The y-axis of the graph is the log of ‘how long a software engineering task can AIs reliably succeed at doing.’ The straight line says the answer doubles roughly every 7 months. Yikes. Upcoming: The comment period on America’s AI strategy is over, so...
d9997ba6-e8be-4911-8c42-f89de3ca906d
trentmkelly/LessWrong-43k
LessWrong
Announcing: Progress Studies Reading Group. This is cross-posted from the EA forum. https://forum.effectivealtruism.org/posts/JBuWGZzCb9TMHNgHw/announcing-progress-studies-reading-group Kris Gulati recently began to organize a virtual Progress Studies reading group. Everyone is welcome to attend. The group is organiz...
b199a506-db05-4ac6-ac3c-aacd06a4292a
awestover/filtering-for-misalignment
Redwood Research: Alek's Filtering Results
id: post240 The AI should talk like a team of many AI. Each AI only uses the word "I" when referring to itself, and calls other AI in the team by their name. I argue that this may massively reduce Self-Allegiance by making it far more coherent for one AI to whistleblow or fight another AI which is unethical or dangero...
77b1f7b1-f6e4-48ef-81a2-2620f9534701
StampyAI/alignment-research-dataset/youtube
Youtube Transcripts
Experts' Predictions about the Future of AI hi there's a lot of disagreement about the future of AI but there's also a lot of disagreement about what the experts think about the future of AI I sometimes hear people saying that all of this concern about AI risk just comes from watching too much sci-fi and the actual AI...
96ec094c-bbfb-4b38-aac9-01e638230800
StampyAI/alignment-research-dataset/aisafety.info
AI Safety Info
I’d like to get deeper into the AI alignment literature. Where should I look? The [AGI Safety Fundamentals Course](https://www.eacambridge.org/technical-alignment-curriculum) is arguably the best way to get up to speed on alignment; you can apply to go through it with many other people studying and mentorship, but not...
5b2d5f7c-7dad-4339-bab7-ce8d49dfbfcd
trentmkelly/LessWrong-43k
LessWrong
Goal completion: noise, errors, bias, prejudice, preference and complexity A putative new idea for AI control; index here. This is a preliminary look at how an AI might assess and deal with various types of errors and uncertainties, when estimating true human preferences. I'll be using the circular rocket model to il...
4dc7de3b-a2c6-4cfa-bb7b-1f53ed33f815
trentmkelly/LessWrong-43k
LessWrong
Housing Roundup #6 The listings will continue until people can afford houses in places they want to live. NEW CITY BY THE BAY, WHO DIS They want to live in San Francisco, but San Francisco has no interest in building homes. What to do? The city’s answer is to spend massively to support its homeless population, causi...
0eb105c0-d51a-41ad-9782-9577528c0a2a
trentmkelly/LessWrong-43k
LessWrong
Some Intuitions Around Short AI Timelines Based on Recent Progress tldr: I give some informal evidence and intuitions that point toward AGI coming soon. These include thinking about how crazy the last year has been, beliefs from those in major AI labs, and progress on MMLU.  Intro This post is intended to be a low-e...
80ab45ff-e6ec-4717-8caa-dc05f557ed1b
StampyAI/alignment-research-dataset/alignmentforum
Alignment Forum
Acausal trade: universal utility, or selling non-existence insurance too late .mjx-chtml {display: inline-block; line-height: 0; text-indent: 0; text-align: left; text-transform: none; font-style: normal; font-weight: normal; font-size: 100%; font-size-adjust: none; letter-spacing: normal; word-wrap: normal; word-spac...
d7951272-4532-4d9f-8426-5ff1e9711867
StampyAI/alignment-research-dataset/arxiv
Arxiv
AI and the Everything in the Whole Wide World Benchmark 1 Introduction --------------- In the 1974 Sesame Street children’s storybook *Grover and the Everything in the Whole Wide World Museum* (Stiles and Wilcox, [1974](#bib.bib91)), the Muppet monster Grover visits a museum claiming to showcase “everything in the ...
7ae6ffe4-4e7b-4c9f-bbed-49000e4d9d45
trentmkelly/LessWrong-43k
LessWrong
The type of AI humanity has chosen to create so far is unsafe, for soft social reasons and not technical ones. I saw this image on X: and I not only endorse this lady's comment, but would like to briefly expand on it in terms of AI safety. How many people here were attracted to the concept of AI safety by Eliezer Yu...
d1804236-ab49-4734-a263-75f173dff5d7
StampyAI/alignment-research-dataset/blogs
Blogs
November 2021 Newsletter #### MIRI updates * MIRI won’t be running a formal fundraiser this year, though we’ll still be participating in Giving Tuesday and other matching opportunities. Visit [intelligence.org/donate](https://intelligence.org/donate) to donate and to get information on tax-advantaged donations, empl...
cfbcc16d-104c-4d8f-85c2-fa08bfaa8761
StampyAI/alignment-research-dataset/eaforum
Effective Altruism Forum
Critiques of non-existent AI safety labs: Yours *"Starting a company is like chewing glass. Eventually, you start to like the taste of your own blood."* Building a new organization is extremely hard. It's hard when you've done it before, even several times. It's even harder the first time. Some new organizat...
0a79d604-21b9-4f73-be71-593d91af41ae
trentmkelly/LessWrong-43k
LessWrong
What makes counterfactuals comparable? l was attempting to write a reference post on the concept of comparability in decision theory problems, but I realised that I don't yet have a strong enough grasp on the various positions that one could adopt to write a post worthy of being a reference. I'll quote my draft quite ...
2594ef68-a035-4998-a220-b6358052f2fd
trentmkelly/LessWrong-43k
LessWrong
Mapping downside risks and information hazards This post was written for Convergence Analysis. Many altruistic actions have downside risks; they might turn out to have negative effects, or to even be negative overall. Perhaps, for example, the biosecurity research you might do could pose information hazards, the arti...
d31c13de-75d6-43f7-a6fa-95ab11559dda
trentmkelly/LessWrong-43k
LessWrong
You Are Not So Smart (Pop-Rationality Book) Journalist David McRaney has very recently published a popular book on human rationality. The book, You Are Not So Smart, is currently the 3rd best selling book in Nonfiction/Philosophy on Amazon.com after less than a week on the market. (Eighth best selling book in Nonfict...
1927158f-8556-462a-a066-3a452b590b9b
trentmkelly/LessWrong-43k
LessWrong
Non-theist cinema? There isn't much in the way of explicitly atheist cinema* -- that is, movies that contain the explicit or implicit message that religion is nothing but superstition, and where this point itself is a central part of the story. The only popular films that jump to mind here are The Invention of Lying, ...
bb5d82f0-e481-4c69-a8d0-a5b39c7bdc64
trentmkelly/LessWrong-43k
LessWrong
Work in the Mainstream CS community: Communicating Meaning Under Uncertainty
41788660-69ae-463e-b628-4275f690e28c
trentmkelly/LessWrong-43k
LessWrong
Sam Altman's sister claims Sam sexually abused her -- Part 5: Timeline, continued Previous posts (which you should read first) This post is the 5th post in a series of 11 posts about the claims of Sam Altman's sister, Annie Altman. Annie has claimed that Sam sexually abused her for about 9 years as a child, and that s...
48b3901f-871e-42e2-995f-f5c9a17aee39
LDJnr/LessWrong-Amplify-Instruct
LessWrong
"It commonly acknowledged here that current decision theories have deficiencies that show up in the form of various paradoxes. Since there seems to be little hope that Eliezer will publish his Timeless Decision Theory any time soon, I decided to try to synthesize some of the ideas discussed in this forum, along with a ...
913ef59d-f9ca-4606-ab6b-fe37bbcd2f99
trentmkelly/LessWrong-43k
LessWrong
What do you do when you find out you have inconsistent probabilities? I've recently been reading about a rationalist blogger who converted to Catholicism. She may have assigned subjective probabilities like: P(Objective Morality)=0.9P(God)=0.05 Then she may have introspected and come up with: P(Objective Morality | ...
d943dedf-72d5-4c13-ac34-6508ec93417e
trentmkelly/LessWrong-43k
LessWrong
Responsible Deployment in 20XX The purpose of this post is to briefly talk about two categories of alignment and how I think AI safety evaluations should use them. I only consider responsible deployment in 20XX because I often find discussions about 2XXX to be distracting and unproductive. Also, for the remainder of t...
bac3438d-9821-48b3-84a5-8eea09673ec9
trentmkelly/LessWrong-43k
LessWrong
Exploration-Exploitation problems Original post: http://bearlamp.com.au/exploration-exploitation-problems/ I have been working on the assumption that exploration-exploitation knowledge was just common.  Unfortunately I did the smart thing of learning about them from a mathematician at a dojo in Melbourne, which means...
1ce683d2-950b-4190-bac1-5634bf133af5
trentmkelly/LessWrong-43k
LessWrong
Does GPT-4's ability to compress text in a way that it can actually decompress indicate self-awareness? My model of a GPT-4 without a self-model which was told to create a compression technique for text is that it would copy something from its training data which related to the string 'compression,' but which it could...
9d70e183-f5fa-4fe7-b7ef-a8f3e037d58d
trentmkelly/LessWrong-43k
LessWrong
Skills from a year of Purposeful Rationality Practice A year ago, I started trying to deliberate practice skills that would "help people figure out the answers to confusing, important questions." I experimented with Thinking Physics questions, GPQA questions, Puzzle Games , Strategy Games, and a stupid twitchy reflex ...
f7145e17-a96c-480a-9c54-9d09b45989c7
trentmkelly/LessWrong-43k
LessWrong
Goals of model vs. goals of simulacra? In classical examples of misaligned AIs (like the famous paperclip maximiser ones), we tend to view the gradient descent training as the step in which an AI gets imbued with some overarching goal that it might then pursue to its bitter conclusion. However, the currently dominatin...
3c315b0b-f2d1-4979-999b-1abb65d2fc62
trentmkelly/LessWrong-43k
LessWrong
Announcing AlignmentForum.org Beta We've just launched the beta for AlignmentForum.org. Much of the value of LessWrong has come from the development of technical research on AI Alignment. In particular, having those discussions be in an accessible place has allowed newcomers to get up to speed and involved. But the a...
560bf52e-1b5f-4ace-9e72-92311842c566
trentmkelly/LessWrong-43k
LessWrong
The Zeroth Skillset Related: 23 Cognitive Mistakes that make People Play Bad Poker Followed by: Situational Awareness And You If epistemic rationality is the art of updating one's beliefs based on new evidence to better correspond with reality, the zeroth skillset of epistemic rationality-- the one that enables all ...
68e1c31e-0690-486f-be83-b05de5dfcad5
trentmkelly/LessWrong-43k
LessWrong
Massive Activations and why <bos> is important in Tokenized SAE Unigrams This article is a footnote to the (excellent) Tokenized SAEs. To briefly summarize: the authors (Thomas Dooms and Daniel Wilhelm) start by proposing the idea of a token "unigram".  Let Ut be the unigram for a specific token t. Conceptually there...
ac3ed03a-1d50-4964-a8e5-e9637c4d75e3
StampyAI/alignment-research-dataset/lesswrong
LessWrong
Walkthrough of 'Formalizing Convergent Instrumental Goals' Introduction ------------ I found *[Formalizing Convergent Instrumental Goals](https://intelligence.org/files/FormalizingConvergentGoals.pdf)* (Benson-Tilsen and Soares) to be quite readable. I was surprised that the instrumental convergence hypothesis had be...
363a52ac-f2ed-4b66-89ac-9b0b523ad984
trentmkelly/LessWrong-43k
LessWrong
Superhuman Coders in AI 2027 - Not So Fast Thank you @elifland for reviewing this post. He and AI Futures are planning to publish updates to the AI 2027 Timeline Forecast soon. AI 2027 (also launched in this LW post) forecasts an R&D-based AI takeoff starting with the development of Superhuman Coders[1] within a fron...
bbba79b0-cb4e-431c-8dd6-f2e12f85f555
trentmkelly/LessWrong-43k
LessWrong
Group Rationality Diary, Oct. 6-18, 2015 This is the public group rationality diary for October 6-18, 2015. 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 * De...
a0efeeed-48f5-42d0-ba68-cea24c3dd5d2
trentmkelly/LessWrong-43k
LessWrong
are IQ tests a good measure of intelligence? I define "intelligence" as "having a strong ability to hit narrow targets in a large search space" Also, is there a one "official" IQ test? or does any random internet thing that calls itself an "iq test" work? which ones are real? if they are not good measures of intelli...
38566772-72e3-470a-a73f-8a2cf3c55d37
trentmkelly/LessWrong-43k
LessWrong
What LessWrong/Rationality/EA chat-servers exist that newcomers can join? A few times over the past few years newcomers to the community have asked me how they can start engaging with the community, and one of the lowest-cost ways to engage has historically been chatrooms. I know that there has been a pretty signific...
b2cfb217-2bd3-4144-ae97-bb83c2a2a471
trentmkelly/LessWrong-43k
LessWrong
Against blanket arguments against interpretability On blanket criticism and refutation In his long post on the subject, Charbel-Raphaël argues against theories of impacts of interpretability. I think it's a largely a good, well-argued post, and if the only thing you get out of it is reading that post, I'll be contribu...
867efef2-5ccc-4955-8d1b-7ed27a13a2fd
trentmkelly/LessWrong-43k
LessWrong
Interest in Leetcode, but for Rationality? the following is motivated by: * Self-Improvement or Shiny Distraction: Why Less Wrong is anti-Instrumental Rationality * The Martial Art of Rationality I've been a long time lurker on Less Wrong and I've noticed the recurring criticism that despite its focus on rationali...
abf58998-4775-4fbb-a82f-c615f9b2507a
trentmkelly/LessWrong-43k
LessWrong
Meetup : "What is the Intelligence Explosion?" (and more...) LW Melbourne dojo Discussion article for the meetup : "What is the Intelligence Explosion?" (and more...) LW Melbourne dojo WHEN: 02 April 2016 02:00:00PM (+1100) WHERE: Bull and Bear Tavern, Melbourne Less Wrong Melbourne's Rationality Dojos are self-imp...
a3389e86-3e3b-4b14-85f0-155367b2cc9b
trentmkelly/LessWrong-43k
LessWrong
Towards cause priotisation estimates for child abuse       Closest community background reading: http://www.givewell.org/labs/causes/criminal-justice-reform   Scale   prevalence   Back of the envelope estimate of the number of abused excluding those who are emotionaly abused and neglected (because those stats...
0d15484f-d335-4041-a80b-d2135c5f6446
StampyAI/alignment-research-dataset/youtube
Youtube Transcripts
AI safety needs social scientists | Amanda Askell | EA Global: London 2018 Hi everyone. Can you all hear me okay? I'm going to assume that was a yes. Okay. So, here's an overview of what I'm going to be talking about today. So first, I'm going to talk a little bit about why learning human values is difficult for AI sy...
5ed435af-fe30-4729-9e68-c7e894fa4e68
trentmkelly/LessWrong-43k
LessWrong
When and why should you use the Kelly criterion? This is a dialogue made during the online dialogues party. Phil, River, and I (Garrett) talked about the Kelly criterion. I was under the mistaken impression that it would come up in finite games, even without discount factors. This turned out to be wrong! In such betti...
f12e68f2-bcdc-4271-946b-8c2b0333eb53
StampyAI/alignment-research-dataset/blogs
Blogs
Michie and overoptimism *By Katja Grace, 12 January 2015* We recently wrote about [Donald Michie’s](http://en.wikipedia.org/wiki/Donald_Michie) [survey](http://aiimpacts.wpengine.com/michie-survey/ "Michie Survey") on timelines to human-level AI. Michie’s survey is especially interesting because it was taken in 1972...
50cadd34-0aa9-477f-99c5-9f411ac94220
trentmkelly/LessWrong-43k
LessWrong
TAPs for Tutoring Introduction Scott Alexander: > Architects. Teachers. Teachers of teachers, but the art of teaching teaching is much the same as the art of teaching. Three levels is enough. Though the levels have to mix. The teacher who trains the next architect must be a master both of teaching and of architecture...
8d803171-ce9f-496b-acd5-c7c65ca80320
trentmkelly/LessWrong-43k
LessWrong
Monthly Shorts 10/22 India? Map of Indian poverty, raw data on page 136. For how important poverty/wealth in those states are, you can look at That data was drawn from Wikipedia, while the population map distortion was done with this program, which I suspect I will use myself one of these days. Comparing the two i...
6fdff81c-26f2-4af8-9cc0-4bba86afd466
trentmkelly/LessWrong-43k
LessWrong
Oracles, Informers, and Controllers Epistemic Status: I think the axes are straightforward enough. I'm not particularly sure about the names, happy to hear other sorts of suggestions. Scholarly Status: Push/pull systems have been discussed before. I'm sure there's other naming for this sort of thing in the engineerin...
fb40f832-8bd8-4f84-aea8-946ae414bec0
StampyAI/alignment-research-dataset/blogs
Blogs
you can't simulate the universe from the beginning? you can't simulate the universe from the beginning? --------------------------------------------------- [QACI](qaci.html) and plausibly [PreDCA](predca.html) rely on a [true name](https://www.lesswrong.com/posts/FWvzwCDRgcjb9sigb/why-agent-foundations-an-overly-abs...
394a5cbf-83c4-4b53-bb28-c26d30d516e5
trentmkelly/LessWrong-43k
LessWrong
Interlude: Agents as Automobiles I ended up writing a long rant about agency in my review of Joe Carlsmith’s report on x-risk from power-seeking AI. I’ve polished the rant a bit and posted it into this sequence. The central analogy between APS-AI and self-propelled machines (“Auto-mobiles”) is a fun one, and I suspect...
a5dd6c33-d7b0-4293-bfdb-10a4fcf46b1b
trentmkelly/LessWrong-43k
LessWrong
Does functionalism imply dualism? This post follows on from Personal research update, and is followed by State your physical explanation of experienced color. In a recent post, I claimed that functionalism about consciousness implies dualism. Since most functionalists think their philosophy is an alternative to dual...