id
stringlengths
36
36
source
stringclasses
15 values
formatted_source
stringclasses
13 values
text
stringlengths
2
7.55M
b360ed2a-0ddc-4528-8c06-aff4008cc5f0
StampyAI/alignment-research-dataset/arbital
Arbital
Modeling distant superintelligences One of the things we almost certainly don't want our AI to do, unless we're *extremely* confident that it is *extremely* robust and value-aligned, is have it think about and try to model alien civilizations that might contain superintelligences or potential simulators. Among the po...
73ee1fcd-777c-4b03-8295-4ab3cecd3539
trentmkelly/LessWrong-43k
LessWrong
Let's ask some of the largest LLMs for tips and ideas on how to take over the world Intro: Imagine you were an ASI. You were smarter and faster than any Human alive. And, you were immortal (you won't naturally die of "old age"). You could learn, grow, expand and experience the entire known Universe, for an eternity. ...
28fb5616-3a74-4c29-9a4c-63b7023a9bc9
StampyAI/alignment-research-dataset/alignmentforum
Alignment Forum
[Replication] Conjecture's Sparse Coding in Toy Models Summary ------- In the post [Taking features out of superposition with sparse autoencoders](https://www.lesswrong.com/posts/z6QQJbtpkEAX3Aojj/interim-research-report-taking-features-out-of-superposition), Lee Sharkey, Beren Millidge and Dan Braun (formerly at Con...
2b096358-b0bc-4917-82ae-1acecf5868cb
trentmkelly/LessWrong-43k
LessWrong
Is it rational to take psilocybin? Is it rational to take psilocybin? Just to make my definition of rational clear: Rationality is only intelligible when in the context of a goal (whether that goal be rational or irrational). Now, if one acts rationally, given their information set, will chose the best plan-of-actio...
51bbb564-4bbf-40f9-84c3-bc7d30a01b92
trentmkelly/LessWrong-43k
LessWrong
Covid 12/22/22: Reevaluating Past Options This is the non-China section of this week’s update. Despite there being essentially zero good statistical data about the real situation in China, things are clearly moving fast, and I have assembled what useful information I have been able to find. If cases are rising as quic...
2d4c1b8b-fa62-407c-93d6-3d5ac3eaab1e
trentmkelly/LessWrong-43k
LessWrong
A Less Wrong singularity article? Robin criticizes Eliezer for not having written up his arguments about the Singularity in a standard style and submitted them for publication. Others, too, make the same complaint: the arguments involved are covered over such a huge mountain of posts that it's impossible for most outs...
e9438c66-6a4c-4d6b-a905-44d1e55e6b62
trentmkelly/LessWrong-43k
LessWrong
Play in Hard Mode Epistemic Status: Love the player, love the game Also consider: Playing on Easy Mode Raymond Arnold asked me, why do you insist on playing in hard mode? Hard mode is harder. The reason to Play in Hard Mode is because it is the only known way to become stronger, and to defend against Goodhart’s Law...
ff30c02f-6698-48bf-8b34-a3ba8cfe6408
trentmkelly/LessWrong-43k
LessWrong
A Defense of Peer Review This article by Dan Elton (moreisdifferent.blog) was published by Asimov Press. > “I think peer review is, like democracy, bad, but better than anything else." > > — Timothy Bates, University of Edinburgh I used to see peer review as primarily good — an important gatekeeping process, essent...
fc5e9742-f607-4efd-844d-627e29f7f288
StampyAI/alignment-research-dataset/lesswrong
LessWrong
World, mind, and learnability: A note on the metaphysical structure of the cosmos [& LLMs] *Cross-posted from* [*New Savanna*](https://new-savanna.blogspot.com/2020/08/world-mind-and-learnability-note-on.html). There is no a priori reason to believe that world has to be learnable. But if it were not, then we wouldn’...
996d0c12-8d6a-4e59-97e1-237c07d48427
trentmkelly/LessWrong-43k
LessWrong
Are these cognitive biases, biases? Continuing my special report on people who don't think human reasoning is all that bad, I'll now briefly present some studies which claim that phenomena other researchers have considered signs of faulty reasoning aren't actually that. I found these from Gigerenzer (2004), which I in...
b3df8556-bb2c-4f1e-a615-baeac6df1b1b
trentmkelly/LessWrong-43k
LessWrong
Want to work on "strong AI" topic in my bachelor thesis Hello, I currently study maths, physics and programming (general course) on CVUT at Prague (CZE). I'm finishing second year and I'm really into AI. The most interesting questions for me are: * what formalism to use for connecting epistemology questions (about...
1d13b1cf-0188-4631-9846-63c637460aae
StampyAI/alignment-research-dataset/blogs
Blogs
Accuracy of AI Predictions *Updated 4 June 2015* It is unclear how informative we should expect expert predictions about AI timelines to be. Individual predictions are undoubtedly often off by many decades, since they disagree with each other. However their aggregate may still be quite informative. The main potentia...
fe8ae980-a425-4e2c-b52b-08d401b90e9f
trentmkelly/LessWrong-43k
LessWrong
Meetup : Washington DC Munchinkry meetup Discussion article for the meetup : Washington DC Munchinkry meetup WHEN: 12 May 2013 03:00:00PM (-0400) WHERE: National Portrait Gallery, Washington, DC 20001, USA (courtyard) Meet to talk about interesting lifehacky things! (Credit to Maia for the idea for this one). Di...
fc870241-8aa2-4a74-9958-157a54387785
trentmkelly/LessWrong-43k
LessWrong
Meetup : Madison Monday Meetup Discussion article for the meetup : Madison Monday Meetup WHEN: 23 April 2012 06:30:00PM (-0500) WHERE: 1831 Monroe St., Madison, WI Last week, we discussed the first half-or-so of the Human's Guide to Words, going so far as to brainstorm some methods and habits that'd be useful to le...
021e6706-3733-4a99-b928-39175726bdc4
trentmkelly/LessWrong-43k
LessWrong
Friendly AI at the NYC Future Salon For those in the New York City area:  I will be giving a talk on Friendly AI at the NYC Future Salon this Saturday, the 20th of February, at 3:00pm.  We'll be gathering in the back room of Stone Creek Bar, located at 140 E. 27th Street in Manhattan. At the moment, 29 people have RS...
615e0569-27ba-4a3f-a472-17fca041f084
trentmkelly/LessWrong-43k
LessWrong
The AI research process as incipient superintelligence EDIT: So this got upvotes while being a draft. Not sure what happened there, but it's now officially posted, even though it's incomplete. Won't be including it in the Guide for now, since the style clashes. If I include it in the Guide I'll come back to this and ...
0885175b-3eaa-45b9-839b-978bfe55c096
trentmkelly/LessWrong-43k
LessWrong
Win/continue/lose scenarios and execute/replace/audit protocols In this post, I’ll make a technical point that comes up when thinking about risks from scheming AIs from a control perspective. In brief: Consider a deployment of an AI in a setting where it’s going to be given a sequence of tasks, and you’re worried abo...
679c0f0d-4dfc-4662-8c59-a0ae92703720
trentmkelly/LessWrong-43k
LessWrong
How the LW2.0 front page could be better at incentivizing good content (This should probably be in meta, but I'm not sure how to post to meta. Feel free to move if needed.) Imagine tomorrow someone writes the best ever LW2.0 post and everyone upvotes it. How can front page visitors read that best and newest post? The...
7d664f3d-fcbb-4ede-a420-3d44c0bce552
trentmkelly/LessWrong-43k
LessWrong
Quick takes on "AI is easy to control" A friend asked me for my quick takes on “AI is easy to control”, and gave an advance guess as to what my take would be. I only skimmed the article, rather than reading it in depth, but on that skim I produced the following: > Re: "AIs are white boxes", there's a huge gap between...
0b903690-c376-4c6c-b84d-2df45ce66df1
trentmkelly/LessWrong-43k
LessWrong
Open Thread - Aug 24 - Aug 30 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. (Immediately before...
bca79607-9363-4ee6-a37e-203428368865
StampyAI/alignment-research-dataset/arxiv
Arxiv
AMP: Adversarial Motion Priors for Stylized Physics-Based Character Control. 1. Introduction ---------------- Synthesizing natural and life-like motions for virtual characters is a crucial element for breathing life into immersive experiences, such as films and games. The demand for realistic motions becomes even m...
42b553ea-eb98-4fe0-9eaa-99d518f910b9
trentmkelly/LessWrong-43k
LessWrong
[SEQ RERUN] Unbounded Scales, Huge Jury Awards, & Futurism Today's post, Unbounded Scales, Huge Jury Awards, & Futurism was originally published on 29 November 2007. A summary (taken from the LW wiki):   > Without a metric for comparison, estimates of, e.g., what sorts of punitive damages should be awarded, or when ...
b2d33980-78f6-4abe-88bc-70652adc58ec
StampyAI/alignment-research-dataset/eaforum
Effective Altruism Forum
The US expands restrictions on AI exports to China. What are the x-risk effects? Late last week, the Biden administration announced a new set of regulations that make it illegal for US companies to export a range of AI-related products and services to China (Financial Times coverage [here](https://www.ft.com/content/6...
46a79294-52ad-4180-90c4-1f863b0b4d82
trentmkelly/LessWrong-43k
LessWrong
Meetup : Moscow meetup, creating community Discussion article for the meetup : Moscow meetup, creating community WHEN: 26 May 2012 04:00:00PM (+0400) WHERE: Rossiya, Moscow, ulitsa Ostozhenka 14 We will meet at “Subway” restaurant, entrance from Lopukhinskiy pereulok. Look for a table with “LW” banner, I will be th...
f416c9f6-a50a-4ca5-8051-df243a97c880
trentmkelly/LessWrong-43k
LessWrong
Can we prime people to take risk? Cross-posted from The Metasophist. ---------------------------------------- A key contributor to stagnation could be a growing reluctance to take risk. All creation involves risk, because when embarking on the journey you rarely know the result. As Paul Graham said, you might fear c...
42d33043-3614-47ec-b641-c9b37e1802b7
StampyAI/alignment-research-dataset/eaforum
Effective Altruism Forum
A concern about the “evolutionary anchor” of Ajeya Cotra’s report on AI timelines. ![](https://upload.wikimedia.org/wikipedia/commons/a/ae/Darwin%27s_finches_by_Gould.jpg)**tl;dr**: The report underestimates the amount of compute used by evolution because it only looks at what it would take to simulate neurons, rather...
b94d2767-abfc-4199-916c-032043c89e0c
trentmkelly/LessWrong-43k
LessWrong
The Plan - 2022 Update So, how’s The Plan going? Pretty well! In last year’s writeup of The Plan, I gave “better than a 50/50 chance” that it would work before AGI kills us all (and my median AI timelines were around 10-15 years). That was an outside view, accounting for planning fallacy and the inevitable negative s...
dce1ffa4-1014-4cc4-834d-9b7b83fafd78
trentmkelly/LessWrong-43k
LessWrong
How useful would alien alignment research be? Imagine aliens on a distant world. They have values very different to humans. However, they also have complicated values, and don't exactly know their own values.  Imagine these aliens are doing well at AI alignment. They are just about to boot up a friendly (to them) su...
061f4c1d-1717-4209-b292-c2256acabc5a
StampyAI/alignment-research-dataset/arbital
Arbital
Decimal notation Seventeen is the number that represents as many things as there are x marks at the end of this sentence: xxxxxxxxxxxxxxxxx. Writing out numbers by saying "the number representing how many things there are in this pile:" gets unwieldy when the pile gets large. Thus, we _represent_ numbers using the [nu...
133dc245-40f2-49c1-b8a2-a5269f504d22
LDJnr/LessWrong-Amplify-Instruct
LessWrong
"Did you ever wonder, when you were a kid, whether your inane "summer camp" actually had some kind of elaborate hidden purpose—say, it was all a science experiment and the "camp counselors" were really researchers observing your behavior? Me neither. But we'd have been more paranoid if we'd read Intergroup Conflict and...
e0932f78-62b1-4a1c-b469-61d2870a4e42
trentmkelly/LessWrong-43k
LessWrong
Naturalism and AI alignment Edit 1: Allergic to naturalism and other realist positions? You can still benefit from reading this, by considering ideal observer theory instead. I am claiming that something like an ideal-observer AI can be built, and that there is a non-trivial chance such an agent becomes aligned (after...
e17abca2-d28d-4b5a-a035-a4c40729b273
trentmkelly/LessWrong-43k
LessWrong
Happy Ada Lovelace Day Today is Ada Lovelace Day, when STEM enthusiasts highlight the work of modern and historical women scientists, engineers, and mathematicians.  If you run a blog, you may want to participate by posting about a woman in a STEM field whom you admire.  But I'd love to have people share women scienti...
d99ef50f-f16b-46a8-a82f-45e0cfbfc0b5
trentmkelly/LessWrong-43k
LessWrong
BIDA Election Thoughts At this Sunday's dance BIDA will be holding its annual meeting, which means at the break there will paper ballots for voting on two things: * Board: Who will run the organization for the next year? * Bylaws: Three proposed changes to simplify elections. Harris wrote up a blog post with a s...
e0ce3651-81c8-45f7-8a43-0b72cccaf917
trentmkelly/LessWrong-43k
LessWrong
Umbilical cord stem cell banking for future medical use I had not heard of this until recently, and I doubt I am alone. I will quote from this source, which looks OK (there is also this). What is the idea? > [D]uring the 1970s, researchers discovered that umbilical cord blood could supply the same kinds of blood-for...
9f08cc3c-5965-476e-b833-732b0c65f34d
trentmkelly/LessWrong-43k
LessWrong
The Little Dragon is Dead None
1f90ef90-b450-4e6a-b79e-df66651067f4
trentmkelly/LessWrong-43k
LessWrong
Bogus Exam Questions Take a look at this multiple choice question from the Open University BSc on psychology. Can you tell what is wrong with it. This is sampled from a set of 15 questions, some of the others aren't this bad, some are guessing the teachers password. One of the questions requires that you know that sta...
6cf1ce8d-ce7b-42f5-b0e9-cab0ebc2c04f
trentmkelly/LessWrong-43k
LessWrong
[SEQ RERUN] Anchoring and Adjustment Today's post, Anchoring and Adjustment was originally published on 07 September 2007. A summary (taken from the LW wiki):   > Exposure to numbers affects guesses on estimation problems by anchoring your mind to an given estimate, even if it's wildly off base. Be aware of the effe...
cf2d9ba8-1340-4748-9d1d-3a88ecd637b4
StampyAI/alignment-research-dataset/arbital
Arbital
Communication: magician example Imagine that you and I are both magicians, performing a trick where I think of a card from a deck of cards and then leave the room, at which point you enter the room and have to guess which card I picked. To perform this trick, I leave some [https://arbital.com/p/3xd](https://arbital.co...
083548a4-c07e-4e73-b733-9b50f6ea7aeb
StampyAI/alignment-research-dataset/blogs
Blogs
Introducing Myself Many friends of MIRI will know that I have been a supporter of its mission since its founding, and have rendered my informal assistance, including a major role in arranging matching funds for the Institute’s 2007 Challenge Grant. I am pleased to take a more direct role in fostering its success as ...
010e6019-5773-44ae-86a9-cdbf403b5964
trentmkelly/LessWrong-43k
LessWrong
Peter Thiel backs "AI startup" [link] http://venturebeat.com/2011/02/07/vicarious-systems-funding-facebook/
5dfb451d-4d53-4dc1-9aa8-12f89bb3a99d
trentmkelly/LessWrong-43k
LessWrong
Meetup : Baltimore Area / UMBC Weekly Meetup Discussion article for the meetup : Baltimore Area / UMBC Weekly Meetup WHEN: 13 November 2016 08:00:00PM (-0500) WHERE: Performing Arts and Humanities Bldg Room 456, 1000 Hilltop Cir, Baltimore, MD 21250 Meeting is on 4th floor of the Performing Arts and Humanities Buil...
580470f8-308b-4ef8-83d4-f6d20aea5bd0
trentmkelly/LessWrong-43k
LessWrong
Introduction to Existential Risks from Artificial Intelligence, for an EA audience Among the causes that Effective Altruists support, existential risk stands out as different from ordinary charitable concerns, and existential risk from Artificial Intelligence even more so. In this talk, which assumes no prior knowled...
4ffe5dcc-2c7f-4fe9-89fc-ce218708fe6f
trentmkelly/LessWrong-43k
LessWrong
On GPT-4.5 It’s happening. The question is, what is the it that is happening? An impressive progression of intelligence? An expensive, slow disappointment? Something else? The evals we have available don’t help us that much here, even more than usual. My tentative conclusion is it’s Secret Third Thing. It’s a diff...
3998361a-24f2-4fdd-b2f5-9e708fd47ec9
trentmkelly/LessWrong-43k
LessWrong
Types of generalism [Cross-posted from here] I am interested in the nature of inter- and transdisciplinary research, which often involves some notion of “generalism”. There are different ways to further conceptualize generalism in this context. First, a bit of terminology that I will rely on throughout this post: I ...
729f0fe2-7675-49da-b095-67b29e224320
StampyAI/alignment-research-dataset/alignmentforum
Alignment Forum
[AN #141]: The case for practicing alignment work on GPT-3 and other large models 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 throu...
d18dcbbf-30c1-47cc-8582-ecafdfbc4acb
trentmkelly/LessWrong-43k
LessWrong
Missing magic Image by ArcheiaMuriel via Flickr Many people believe they can fix the human body using magic of some sort. Many also claim to magically fix emotional or psychological problems or to fix other animals or even plants. Yet there are many things nobody seems to claim they have magic to fix. Why is there n...
de157d53-704f-4092-9907-5aa25481ef9d
trentmkelly/LessWrong-43k
LessWrong
Check OK, babble-read, optimize (how I read textbooks) A year ago, I didn't have any conscious ideas for how to read a textbook. I was frustrated by forgetting things, and had heard about memory palaces. When I was playing classical piano, I was able to memorize long complicated pieces of music with thousands of notes...
3afe4456-3ae1-468f-b203-3a794e07fdb0
StampyAI/alignment-research-dataset/eaforum
Effective Altruism Forum
AMA: The new Open Philanthropy Technology Policy Fellowship We will be answering questions about a new fellowship program run by Open Philanthropy, the [Technology Policy Fellowship](https://www.openphilanthropy.org/focus/global-catastrophic-risks/technology-policy-fellowship). We will begin answering questions in the...
e02cc787-60d2-42dd-acc8-48f015d97b0e
trentmkelly/LessWrong-43k
LessWrong
Agents of No Moral Value: Constrained Cognition? Thought experiments involving multiple agents usually postulate that the agents have no moral value, so that the explicitly specified payoff from the choice of actions can be considered in isolation, as both the sole reason and evaluation criterion for agents' decisions...
b92ff09f-17d6-4984-bc0a-5b1fd26728da
awestover/filtering-for-misalignment
Redwood Research: Alek's Filtering Results
id: post3750 A putative new idea for AI control; index here . After explaining riggable learning processes , we can now define influenceable (and uninfluenceable) learning processes. Recall that the (unriggable) influence problem is due to agents randomising their preferences, as a sort of artificial `learning' proces...
84a99e61-2db7-4b5f-8b61-3f28d67e7288
trentmkelly/LessWrong-43k
LessWrong
Meetup : Las Vegas/Henderson, NV Discussion article for the meetup : Las Vegas/Henderson, NV WHEN: 30 May 2014 06:00:00PM (-0700) WHERE: 605 Mall Ring Cir. Henderson, NV 89014 Come say hi and meet fellow Reason enthusiasts! I'm not an active member of LW, but would like to get a feel for other Rationally-inspired in...
da38cb26-904b-4387-9eec-864f3cf178e7
trentmkelly/LessWrong-43k
LessWrong
On the destruction of America’s best high school This is a rare case when I think it's actually worth sharing an article that directly comments on the culture wars on Less Wrong: > I’d like you to feel about the impending destruction of Virginia’s Thomas Jefferson High School for Science and Technology, the same way ...
58956aa8-38a7-4726-a233-219dc0a2513a
trentmkelly/LessWrong-43k
LessWrong
Progress links digest, 2023-09-08: The Conservative Futurist, cargo airships, and more Announcements * The Conservative Futurist: How to Create the Sci-Fi World We Were Promised is available for pre-order (via @JimPethokoukis) * Asimov Press is “an editorially-independent venture that will publish books and essays t...
4d9048f1-f41c-4a44-98d9-dcade36324bc
trentmkelly/LessWrong-43k
LessWrong
Miles Brundage resigned from OpenAI, and his AGI readiness team was disbanded This is the full text of a post from "The Obsolete Newsletter," a Substack that I write about the intersection of capitalism, geopolitics, and artificial intelligence. I’m a freelance journalist and the author of a forthcoming book called Ob...
ab60c7ca-570a-49e6-ab10-7589747422f2
trentmkelly/LessWrong-43k
LessWrong
Vote on Anthropic Topics to Discuss What important questions would you want to see discussed and debated here about Anthropic? Suggest and vote below. (This is the third such poll, see the first and second linked.) How to use the poll 1. Reacts: Click on the agree/disagree reacts to help people see how much disagr...
9b32ecf8-e2c8-4f3b-8b01-9776567155ef
trentmkelly/LessWrong-43k
LessWrong
OpenAI introduces function calling for GPT-4 OpenAI’s latest update unlocks function calling for GPT-4 and GPT-3.5 Turbo. This enables developers to "describe functions to gpt-4-0613 and gpt-3.5-turbo-0613, and have the model intelligently choose to output a JSON object containing arguments to call those functions." T...
98136da2-50fe-4261-918e-7cd9a3adb1a3
StampyAI/alignment-research-dataset/lesswrong
LessWrong
Saying the quiet part out loud: trading off x-risk for personal immortality ***Statement:*** *I want to deliberately balance the caution and the recklessness in developing AGI, such that it gets created in the last possible moment so that I and my close ones do not die.* This Statement confuses me. There are several...
5762453f-d136-40e5-afad-bfdf03fd96bc
trentmkelly/LessWrong-43k
LessWrong
Agents with Cartesian childhood and Physicalist adulthood Followup to: Updateless intelligence metrics in the multiverse In the previous post I explained how to define a quantity that I called "the intelligence metric" which allows comparing intelligence of programs written for a given hardware. It is a development o...
bbd0b367-ee9a-41df-807e-0c3e3fdb3dc5
trentmkelly/LessWrong-43k
LessWrong
Can any LLM be represented as an Equation? Can an arbitrary LLM (or LxM) be presented in the form of an equation? I realised it would need to be some crazy big equation with billions of parameters, but is it theoretically possible? The way I see it, the weights are static once the model is trained so why not
9cfd862b-8e5b-4e23-b0d7-a4c15d1e1ea1
trentmkelly/LessWrong-43k
LessWrong
Help with a derivation in chapter 18 of Jaynes PT:LoS? Hi, I've spent like two hours trying to derive 18.11 from 18.10. Can anyone who's read it help me? If you comment or message me I can send you my gmail for gchat or we can work out something else.
a6a4e9d3-23dc-455f-86cc-6c8e3ebd492f
trentmkelly/LessWrong-43k
LessWrong
The ethics of breeding to kill Veganism, vegetarianism, and "ethical" farming seem to be gaining a lot of ground lately, which is something I find fascinatingly absurd. In part, I think this comes from a felicitous style of reasoning that I outlined here. But, in hindsight, I think a lot of people that consume meat ...
f93730e5-de87-4b32-a900-8dcadde6e08d
StampyAI/alignment-research-dataset/alignmentforum
Alignment Forum
EfficientZero: How It Works The goal of this essay is to help you understand [EfficientZero](https://github.com/YeWR/EfficientZero), a reinforcement learning agent that obtains better-than-human median performance on a set of 26 Atari games after just two hours of real-time experience playing each game. Specifically...
58771bff-4c7f-464e-bdcf-8f8236877eb5
trentmkelly/LessWrong-43k
LessWrong
[POLL RESULTS] LessWrong Members and their Local Communities The results for these have been stable for a while now; I'm posting them a bit late. 95 people took the survey after I modified it to add two questions. For the public version, I removed the pre-change data (10 data points). One text response included ident...
5bc7df4f-3f8b-4e17-9944-534466202436
trentmkelly/LessWrong-43k
LessWrong
Fuck Your Miracle Year Dwarkesh Patel’s essay “The Mystery of the Miracle Year” just went COVID-19-level viral (woah, too soon). Fucking Jeff Bezos read the article at tweeted at him (and Marc Andreessen and Paul Graham). I’ve appeared on Dwarkesh’s podcast and he has given me feedback on some of my writing before. I...
af29a028-c98d-4b6d-aa3e-9c41e028765e
StampyAI/alignment-research-dataset/arxiv
Arxiv
How Different Groups Prioritize Ethical Values for Responsible AI 1. Introduction ---------------- Advances in artificial intelligence (AI) have the potential to benefit people and society, but they also raise ethical challenges and concerns about possible adverse impacts (Montreal, [2017](#bib.bib53)). Being prone...
4ab9befb-3adb-4e23-995e-6ccb4ae75cae
trentmkelly/LessWrong-43k
LessWrong
LINK: New clinical trial will try to restore dead brains "... in an effort to revive the brains of those being kept alive solely through life support. Stem cells will be injected directly into the brain..." More at: http://news.discovery.com/tech/biotechnology/dead-could-be-brought-back-to-life-in-medical-trial-1605...
719ea70f-96da-45a3-bc44-c022b3f9470b
trentmkelly/LessWrong-43k
LessWrong
Helping your Senator Prepare for the Upcoming Sam Altman Hearing Hello Everyone, Next Tuesday, the US Senate will hold a hearing on AI (https://www.judiciary.senate.gov/committee-activity/hearings/oversight-of-ai-rules-for-artificial-intelligence). OpenAI CEO Sam Altman will testify before Congress for the first time...
16833320-f583-4507-b161-46679a6404c8
trentmkelly/LessWrong-43k
LessWrong
Uncritical Supercriticality Every now and then, you see people arguing over whether atheism is a “religion.” As I touch on elsewhere, in “Purpose and Pragmatism,” arguing over the meaning of a word nearly always means that you’ve lost track of the original question.1 How might this argument arise to begin with? An at...
1107e633-182d-4f55-8b7f-9de8c1694f03
StampyAI/alignment-research-dataset/alignmentforum
Alignment Forum
Will transparency help catch deception? Perhaps not In the last few months, some AI safety researchers have signaled what Wei Dai [called](https://www.alignmentforum.org/posts/pFAavCTW56iTsYkvR/ai-alignment-open-thread-october-2019#9sy6DdZx7tSSFYXcg) a "wave of optimism" for using transparency tools to catch deceptive...
e0f023b2-55d7-4caa-8948-e74b96058138
trentmkelly/LessWrong-43k
LessWrong
What sorts of preparations ought I do in case of further escalation in Ukraine? I guess the Russian invasion of Ukraine can go in many ways. It might just fizzle out, or continue to be a fairly conventional war with extent limited to Ukraine, or it might escalate with a Russian nuclear strike against Ukraine followed ...
16b26dbf-dc7c-4e53-b9ff-86fd77ccc8c2
StampyAI/alignment-research-dataset/blogs
Blogs
New paper: “Toward negotiable reinforcement learning” [![Toward Negotiable Reinforcement Learning](https://intelligence.org/files/paretooptimal.png)](https://arxiv.org/abs/1701.01302)MIRI Research Fellow Andrew Critch has developed a new result in the theory of conflict resolution, described in “[**Toward negotiable r...
4354c483-6c0f-4281-ab41-7e5f20c3d32f
LDJnr/LessWrong-Amplify-Instruct
LessWrong
"I blew through all of MoR in about 48 hours, and in an attempt to learn more about the science and philosophy that Harry espouses, I've been reading the sequences and Eliezer's posts on Less Wrong. Eliezer has written extensively about AI, rationality, quantum physics, singularity research, etc. I have a question: how...
6d390560-44ac-4ca0-ac73-62d5250d0055
trentmkelly/LessWrong-43k
LessWrong
Logan Strohl on exercise norms (cw: exercise, shame) (also relevant: Shame Processing) Logan Strohl said some things about exercise (and shame, and aspiring to be better) on Facebook today that I found interesting. Minus some mostly-irrelevant parts: > [...] I definitely wasn't trying to kick people who struggle wi...
ea5b84da-4758-45eb-8051-6393eb7436ed
trentmkelly/LessWrong-43k
LessWrong
Beyond Gaussian: Language Model Representations and Distributions In January 2023, beren and Eric Winsor cataloged basic distributional properties of weights, activations, and gradients in GPT-2 models, providing a systematic view of model internals (thanks to Ryan Greenblatt for the pointer). This post extends their ...
1b85a01a-0754-47e0-a04f-11a7c2a1b0e0
trentmkelly/LessWrong-43k
LessWrong
Duct Tape security This is a linkpost for On Duct Tape and Fence Posts. Eliezer writes about fence post security. When people think to themselves "in the current system, what's the weakest point?", and then dedicate their resources to shoring up the defenses at that point, not realizing that after the first small imp...
efe1e285-ed9a-4b3e-8ae0-b0d6a95ee6a0
trentmkelly/LessWrong-43k
LessWrong
Far & Near / Runaway Trolleys / The Proximity Of (Fat) Strangers I went to the Royal Institute last week to hear the laconic and dismissive Dr Guy Kahane on whether we are 'Biologically Moral'  [His message: Neurological evidence suggests - somewhat alarmingly - that our moral and ethical decisions may be no more tha...
e5c2d15d-168f-4681-b1ed-26c14564c7ee
StampyAI/alignment-research-dataset/alignmentforum
Alignment Forum
[AN #112]: Engineering a Safer World 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](https://docs.google.c...
eceb221e-3043-4eda-bdbf-fdbaafe4a6b9
trentmkelly/LessWrong-43k
LessWrong
Studying The Alien Mind This post is part of a sequence on LLM psychology TL;DR We introduce our perspective on a top-down approach for exploring the cognition of LLMs by studying their behavior, which we refer to as LLM psychology. In this post we take the mental stance of treating LLMs as “alien minds,” comparin...
e4579e5e-9332-4247-9e9c-734b462c3f4a
trentmkelly/LessWrong-43k
LessWrong
Discussion: How scientifically sound are MBAs? I'm finishing up the first year of my distance-learning MBA, which has been a very confusing experience. I went into the course partially as insurance against "unknown unknowns", i.e., lacking concepts important to building or running a business because I didn't know abo...
9a74a53c-ca81-4429-88b8-0633e3670b47
trentmkelly/LessWrong-43k
LessWrong
Interesting talk on Bayesians and frequentists I recently started watching an interesting lecture by Michael Jordan on Bayesians and frequentists; he's a pretty successful machine learning expert that takes both views in his work. You can watch it here: http://videolectures.net/mlss09uk_jordan_bfway/. I found it inter...
4e098963-af86-4f17-8c16-e62d95fb3034
trentmkelly/LessWrong-43k
LessWrong
Society Is Fixed, Biology Is Mutable Today during an otherwise terrible lecture on ADHD I realized something important we get sort of backwards. There’s this stereotype that the Left believes that human characteristics are socially determined, and therefore mutable. And social problems are easy to fix, through things...
79f9e67a-0d69-4478-951b-3fb6139d44b3
trentmkelly/LessWrong-43k
LessWrong
Hypotheses about Finding Knowledge and One-Shot Causal Entanglements Epistemic status: my own thoughts I've thought up in my own time. They may be quite or very wrong! I am likely not the first person to come to these ideas. All of my main points here are just hypotheses which I've come to by the reasoning stated belo...
bb05d4aa-8b52-4e73-b3a0-45488784a5ba
trentmkelly/LessWrong-43k
LessWrong
My Number 1 Epistemology Book Recommendation: Inventing Temperature In my last post, I wrote that no resource out there exactly captured my model of epistemology, which is why I wanted to share a half-baked version of it. But I do have one book which I always recommend to people who want to learn more about epistemol...
40d79c2e-0879-4430-8ffc-839dfea4da4e
trentmkelly/LessWrong-43k
LessWrong
Refactoring EMH – Thoughts following the latest market crash In a twitter thread, Eliezer Yudkowsky challenges the belief in EMH due to the seemingly illogical behavior of the markets with respect to the COVID-19. In the replies, Robin Hanson asks the obvious question: “what exactly is your better theory?” I've been ...
605172b3-d880-44d6-9f4b-2dfc0017498c
trentmkelly/LessWrong-43k
LessWrong
Some work on connecting UDT and Reinforcement Learning
7ca2cd68-d073-480a-b853-86ab7adf2c0b
trentmkelly/LessWrong-43k
LessWrong
Objective Dog Ratings: The Shiba Inu The Shiba Inu Shibas are a spitz dog, which, if you remember from the original Dog Ratings post, is my projected winner for the All-Dog All-Stars Contest, but that doesn't mean that this spitz will take the trophy home. Let's take a look: First, it must be noted that shibas h...
da51c8ef-4021-4160-86f7-8e1b7ff4faca
trentmkelly/LessWrong-43k
LessWrong
STRUCTURE: A Hazardous Guide to Words This post is part of my Hazardous Guide To Rationality. I don't expect this to be new or exciting to frequent LW people, and I would super appreciate comments and feedback in light of intents for the sequence, as outlined in the above link. Also, note this is a STRUCTURE post, aga...
8ce7ce0b-d194-40e8-9a6e-0f69689f327c
StampyAI/alignment-research-dataset/youtube
Youtube Transcripts
9 Examples of Specification Gaming hi when talking about AI safety people often talk about the legend of King Midas you've probably heard this one before Midas is an ancient king who values above all else wealth and money and so when he's given an opportunity to make a wish he wishes that everything he touches would t...
c3c36189-e632-4ff0-8c58-a225913bb0dd
trentmkelly/LessWrong-43k
LessWrong
[CKC] [May 2018] What subjects are most important for an AI safety researcher to know? (Open Call) [CKC]: I intend this to be a Community Knowledge Convergence project - that is to say, I do not think I am particularly well informed on the subject, and I think there's a lot to be gained from compiling an understanding...
92380885-a329-4442-9807-2c4ff1c2b001
trentmkelly/LessWrong-43k
LessWrong
Typical Mind and Politics Yesterday, in the The Terrible, Horrible, No Good Truth About Morality, Roko mentioned some good evidence that we develop an opinion first based on intuitions, and only later look for rational justifications. For example, people would claim incest was wrong because of worries like genetic def...
adde5125-388d-4b8d-bea7-fe1b4a3f03a1
StampyAI/alignment-research-dataset/youtube
Youtube Transcripts
AI Safety Reading Group (Session 43) hello and welcome to the 43rd session of the arca de icse reading group today I'll present that the blog post strong AI isn't here yet by Sarah Constantine Sarek Rasmussen is a PhD from Yale in mathematics and as a block called auction where she's written this article and that stro...
d80785a8-be34-4560-b703-96b6652b1d0a
trentmkelly/LessWrong-43k
LessWrong
A Brief Excursion Into Molecular Neuroscience TL;DR: Poking fun at complicated nomenclature (p̴͓̂͘á̴̻͌s̶̻̗͋͑ṯ̸̹̈́̌à̶̳), identifying a pretty important signaling factor in the brain, and arbitrating (part of) the Guzey/Mendonça debate. Previously in this series: How to build a mind - neuroscience edition. The Empe...
12d03819-2f42-4654-a8aa-34a89394920d
trentmkelly/LessWrong-43k
LessWrong
What is your true decision metric? A look at medicinal chemists http://pipeline.corante.com/archives/2012/11/27/how_do_chemist_think_that_they_judge_compounds.php Some background: medicinal chemists are responsible for identifying drug candidates, usually by screening large (10^6) libraries of combinatorially genera...
9ab74952-9d57-4eae-abc5-05729e69c0b4
trentmkelly/LessWrong-43k
LessWrong
Mediation From a Distance Suppose we have an electronic circuit with two pieces: combinational logic and memory. The combinational logic is designed to behave as memoryless input/output logic gates - it has only short-term state, and computes some function. The memory reads the output of the combinational logic once p...
1a8d6f4c-5a33-4f95-8b40-7317dbe5f890
StampyAI/alignment-research-dataset/blogs
Blogs
Historic trends in flight airspeed records Flight airspeed records between 1903 and 1976 contained one greater than 10-year discontinuity: a 19-year discontinuity corresponding to the Fairey Delta 2 flight in 1956. The average annual growth in flight airspeed markedly increased with the Fairey Delta 2, from 16mph/ye...
486b800c-6c39-486e-ac2e-8e4ca8fa3a68
trentmkelly/LessWrong-43k
LessWrong
For and Against Lotteries in Elite University Admissions People (whose names often rhyme with Palcolm Sadwell) sometimes suggest replacing the admissions systems of highly selective universities with lotteries. The proposal is that universities would mark a pool of students as ‘good enough’ and then students from that...
39f5e198-286e-4e0d-b4b8-e6c24db317b3
awestover/filtering-for-misalignment
Redwood Research: Alek's Filtering Results
id: post1328 Several of the ideas in this post originated in conversations between the authors months ago. I did the work involved in exploring them and thinking of new ideas and framings, as well as writing this post; old writings by Janus were the inspiration for some of the directions of thought here. Thanks to Ada...
03c66bf4-b8e3-4875-8fe2-bda1ba94df48
trentmkelly/LessWrong-43k
LessWrong
Initial Thoughts on Personally Finding a High-Impact Career As I mentioned in my strategic plan, I have 10 months and 28 days until I graduate from Denison University and hopefully will be transitioning to a career.  Careers are important because having one will not only mean that I won't starve, but that I'll have an...
7ab30b7e-8741-498c-8d3c-9edb81e45a84
trentmkelly/LessWrong-43k
LessWrong
Meetup : Cambridge (MA) third-Sundays Meetup Discussion article for the meetup : Cambridge (MA) third-Sundays Meetup WHEN: 21 October 2012 02:00:00PM (-0400) WHERE: 25 Ames St, Cambridge, MA 02139 Cambridge/Boston-area Less Wrong meetups on the first and third Sunday of every month, 2pm at the MIT Landau Building [...
4beb4982-4a2b-4077-954f-4cf0d0515471
StampyAI/alignment-research-dataset/lesswrong
LessWrong
I've updated towards AI boxing being surprisingly easy Specifically because I think that sandboxes like an improved WASM could make it such that conditioned on careful data curation, that side channels have 0 probability to give the AI any information outside of what we give it. I.e I would predict 0 probability of...