id
stringlengths
36
36
source
stringclasses
15 values
formatted_source
stringclasses
13 values
text
stringlengths
2
7.55M
cf03285a-55cf-4dd6-91a0-a410d60f6157
trentmkelly/LessWrong-43k
LessWrong
Of Gender and Rationality Among all self-identified "rationalist" communities that I know of, and Less Wrong in particular, there is an obvious gender imbalance—a male/female ratio tilted strongly toward males. Yet surely epistemic and instrumental rationality have no gender signature.  There is no such thing as masc...
c64f38fd-a79b-4b16-bb29-c932ddc4e80a
trentmkelly/LessWrong-43k
LessWrong
Engaging First Introductions to AI Risk I'm putting together a list of short and sweet introductions to the dangers of artificial superintelligence. My target audience is intelligent, broadly philosophical narrative thinkers, who can evaluate arguments well but who don't know a lot of the relevant background or jargo...
1e012475-d6cd-4a94-9e90-576eb96a8d05
StampyAI/alignment-research-dataset/alignmentforum
Alignment Forum
I missed the crux of the alignment problem the whole time *This post has been written for the first*[*Refine*](https://www.alignmentforum.org/posts/5uiQkyKdejX3aEHLM/how-to-diversify-conceptual-alignment-the-model-behind) *blog post day, at the end of the week of readings, discussions, and exercises about epistemology...
8df30e13-1550-4e14-ade7-b1750f11a6cd
trentmkelly/LessWrong-43k
LessWrong
Super additivity of consciousness In “Freedom under naturalistic dualism” I have carefully argued that consciousness is radically noumenal, that is, it is the most real (perhaps the only real) thing in the Universe, but also totally impossible to be observed by others (non-phenomenal). In my view this strongly limits ...
ba7981e6-0317-49b7-8d82-9b1930ed53c6
trentmkelly/LessWrong-43k
LessWrong
User research as a barometer of software design (Cross-posted on my personal blog.) In UX design, they do something called user research. Consider an example. You have two designs: A and B. You think A is better but John thinks B is better. Who's right? Well, you can debate a bunch of stuff. Design A breaks this gu...
16db73eb-c02c-4bcb-bd3f-5deeae740509
StampyAI/alignment-research-dataset/lesswrong
LessWrong
Another AI Winter? Robin Hanson has been [suggesting](https://aiimpacts.org/conversation-with-robin-hanson/) recently that we've been experiencing an AI boom that's not too different from prior booms. At the recent [Foresight Vision Weekend](https://foresight.org/vision-weekend-2019/), he predicted [not exactly - se...
f61da8f8-3635-4016-bbab-915bbe3c7941
StampyAI/alignment-research-dataset/special_docs
Other
DeepSpeed: Extreme-scale model training for everyone In this article 1. [3D parallelism: Scaling to trillion-parameter models](#3d-parallelism-scaling-to-trillion-parameter-models) 2. [ZeRO-Offload: 10x bigger model training using a single GPU](#zero-offload-10x-bigger-model-training-using-a-single-gpu) 3. [DeepSp...
1876cbd9-1fbc-4759-b434-0d52b885bf18
trentmkelly/LessWrong-43k
LessWrong
Meetup : Washington DC: Museums Discussion article for the meetup : Washington DC: Museums WHEN: 11 May 2014 03:00:00PM (-0400) WHERE: National Portrait Gallery, Washington, DC 20001, USA We'll be meeting to go look at museums. We'll rendezvous in the usual place then head out. Discussion article for the meetup :...
f5dc9f0e-1b83-4757-9821-09032bc03836
StampyAI/alignment-research-dataset/arxiv
Arxiv
Penalizing side effects using stepwise relative reachability 1 Introduction --------------- An important component of safe behavior for reinforcement learning agents is avoiding unnecessary side effects while performing a task (Amodei et al., [2016](#bib.bib2); Taylor et al., [2016](#bib.bib26)). For example, if a ...
89418bfb-b8a7-4452-99b2-7e89c6b86c4c
trentmkelly/LessWrong-43k
LessWrong
Predicting Virus Relative Abundance in Wastewater At the Nucleic Acid Observatory (NAO) we're evaluating pathogen-agnostic surveillance. A key question is whether metagenomic sequencing of wastewater can be a cost-effective method to detect and mitigate future pandemics. In this report we investigate one piece of this...
71227fbc-cadf-4c4e-8f3f-cf0c611e8ea7
trentmkelly/LessWrong-43k
LessWrong
From First Principles Related: Truly a Part of You, What Data Generated That Thought Some Case Studies The other day my friend was learning to solder and he asked an experienced hacker for advice. The hacker told him that because heat rises, you should apply the soldering iron underneath the work to maximize heat tr...
9d9beebb-437b-4ed4-862d-a0e92f9ecd66
trentmkelly/LessWrong-43k
LessWrong
[META] My Negative Results Yesterday, I made a post asking if anyone else had noticed LW being particularly slow. I offered to collect data on this, and was fairly sure (Probably about 80%) that it would show that LW loaded slower than other webpages. I took the post down after about 20 seconds (sorry if I confused yo...
c2d0ac0c-d291-43e0-962d-7201d53315da
trentmkelly/LessWrong-43k
LessWrong
Lotuses and Loot Boxes There has been some recent discussion about how to conceptualize how we lose sight of goals, and I think there's a critical conceptual tool missing. Simply put, there's a difference between addiction and exploitation. Deciding to try heroin and getting hooked is addiction, but the free sample fr...
2e30e847-2927-4734-935f-a51d5c1ba7c7
StampyAI/alignment-research-dataset/lesswrong
LessWrong
4 Key Assumptions in AI Safety AGI is already a very difficult topic, with a large amount of uncertainty, and not enough research being done for its core problems. However, I think even with this understanding, its vastness in scope of possibilities might still be severely under-realized. I’ve listed four major assump...
1cbb656f-0ef4-419c-b56a-6df2e725c98c
trentmkelly/LessWrong-43k
LessWrong
A new option for building lumenators For background on lumenators: * https://meaningness.com/sad-light-led-lux * https://arbital.com/p/lumenators/ * How to Build a Lumenator The original attempt at a lumenator is a commercial product called a "therapy lamp", but the ones I'd seen in the past were quite unambitiou...
8cfcdd16-f687-4726-8ce2-6f51e0fc753f
trentmkelly/LessWrong-43k
LessWrong
[Link] FreakoStats and CEV I would like to point you guys to the blog FreakoStats,  by Garth Zietsman, who according to his profile, "Scored an IQ of 185 on the Mega27 and has a degree in psychology and statistics and 25 years experience in psychometrics and statistics." The main concept discussed there is "The Smart ...
da6d9294-16d4-4c11-adec-b73f84bc9229
trentmkelly/LessWrong-43k
LessWrong
Why not electric trains and excavators? Many countries are supporting electric cars for environmental and independence reasons. But perhaps there are some targets for electrification with better economics than those, cost-effective without any government incentives. For example, trains and hydraulic excavators. -----...
77017247-4c29-4737-8a72-3be15a1c4228
trentmkelly/LessWrong-43k
LessWrong
Learn Power Searching with Google > Google Search makes it amazingly easy to find information. Come learn about the powerful advanced tools we provide to help you find just the right information when the stakes are high. Daniel Russell is doing a free Google class on how to search the web. Besides six 50-minute clas...
62374556-9ed5-4617-aeb7-e0545b5813ed
trentmkelly/LessWrong-43k
LessWrong
The Opt-In Revolution — My vision of a positive future with ASI (An experiment with LLM storytelling) I writing and publishing a story to better clarify my vision for a positive future with transformative AI (or AGI/ASI). The story is written in chapters, and is still being put together piece by piece. Each chapter is...
13a874fd-4337-4ec1-adbe-3b574910f248
trentmkelly/LessWrong-43k
LessWrong
Investment is a useful societal mechanism for getting new things made. Stock trading shares some functionality with investment, but seems very very inefficient, at that? Jack Bogle, the creator of the first index fund, says > The stock market has nothing—n-o-t-h-i-n-g—to do with the allocation of capital. All it mean...
32005c51-f712-4048-bf4d-dea66dd05731
trentmkelly/LessWrong-43k
LessWrong
Erratum for "From AI to Zombies" Is there any erratum for the "From AI to Zombies" book? There are so many essays, and many of them are written more than 10 years ago. It seems very likely that since then the errors or imprecisions were discovered. To be specific, I have a very low belief in the chapter on quantum me...
ece0de46-1b88-4f3a-b813-1294302669d8
trentmkelly/LessWrong-43k
LessWrong
Values Darwinism TL;DR / summary: Can we objectively define "progress"? A Darwinian definition seems plausible. But when referring to cultures, world-views or theories, it seems that physical survival of the fittest must be complemented by memetic survival. The interaction of these two will generally be complex, and ...
ba121d83-c109-4202-9aed-17e311ac2b2b
StampyAI/alignment-research-dataset/lesswrong
LessWrong
RA Bounty: Looking for feedback on screenplay about AI Risk [Jamie Wahls](http://jamiewahls.com/) and Arthur Frost are writing a screenplay for Rational Animations. It's a sci-fi episodic comedy illustrating various ways in which AI goes off the rails in more or less catastrophic ways.  We're looking for feedback and...
dcd8895e-7b88-4e3c-903e-d79acebaefc2
trentmkelly/LessWrong-43k
LessWrong
A tale from Communist China Judging from the upvotes, it seems like people are quite interested in my grandparents' failure to emigrate from Communist China before it was too late, so I thought I'd elaborate here with more details and for greater visibility. They were all actually supporters of the Communist Party at ...
efce598b-babf-4493-a319-8c1d82cd8406
StampyAI/alignment-research-dataset/arxiv
Arxiv
Data Augmentation Can Improve Robustness 1 Introduction --------------- Despite their success, neural networks are not intrinsically robust. In particular, it has been shown that the addition of imperceptible deviations to the input, called adversarial perturbations, can cause neural networks to make incorrect pred...
2f91289b-9c04-411d-b732-38b7b7991c38
trentmkelly/LessWrong-43k
LessWrong
How easy is it to supervise processes vs outcomes? A lot of AI risk models rely on the fact that we supervise outcomes like consequentialists rather than processes like CAIS. But, how easy is it to supervise processes for AI, compared to outcomes both in: Near term: 10 years. Long-term: 50 years. Crucially, is it ...
5c4b2c88-fcb7-4572-ae25-106d991119c6
trentmkelly/LessWrong-43k
LessWrong
A primer on why computational predictive toxicology is hard Introduction There are now (claimed) foundation models for protein sequences, DNA sequences, RNA sequences, molecules, scRNA-seq, chromatin accessibility, pathology slides, medical images, electronic health records, and clinical free-text. It’s a dizzying rat...
515ec586-0c68-42e3-ba72-7d273ab2ec1e
trentmkelly/LessWrong-43k
LessWrong
Announcing the Alignment Research Center (Cross-post from ai-alignment.com) I’m now working full-time on the Alignment Research Center (ARC), a new non-profit focused on intent alignment research. I left OpenAI at the end of January and I’ve spent the last few months planning, doing some theoretical research, doing ...
25a1e42f-91a1-4e4f-bb77-15fe2f02c485
trentmkelly/LessWrong-43k
LessWrong
Decision Theory FAQ Co-authored with crazy88. Please let us know when you find mistakes, and we'll fix them. Last updated 03-27-2013. Contents: * 1. What is decision theory? * 2. Is the rational decision always the right decision? * 3. How can I better understand a decision problem? * 4. How can I measure an age...
9b066b16-1510-4404-a623-e90784cf8a2d
trentmkelly/LessWrong-43k
LessWrong
You only live once: a reframing of working towards posthumanity Let's say you recently purchased a new computer game. Reviewers like the game, but they say that the game itself pales in comparison to a massive, difficult-to-access secret area near the end. Bizarrely, this secret area contains over 90% of the content i...
d80cf380-92f5-498d-b2fc-08da12117df4
StampyAI/alignment-research-dataset/alignmentforum
Alignment Forum
Towards an empirical investigation of inner alignment .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-spacing: normal; white-space...
edcdcf61-54e2-4cef-b550-5a033edc26bc
StampyAI/alignment-research-dataset/arxiv
Arxiv
Adversarial Attacks Against Medical Deep Learning Systems 1 Introduction --------------- Over the past six years, deep learning has transformed computer vision and has been implemented in scores of consumer-facing products. Many are excited that these approaches will continue to expand in scope and that new tools a...
bcb77121-08db-40a2-9cc6-0a6b353c478c
trentmkelly/LessWrong-43k
LessWrong
The Rediscovery of Interiority in Machine Learning None
a9626bbb-347b-404d-8331-6b33ad2cfa65
trentmkelly/LessWrong-43k
LessWrong
Sunday November 15th, 12:00PM (PT) — talks by Abram Demski, Daniel Kokotajlo and (maybe) more! This Sunday at 12pm (PT), we're running another session of "lightning talks" by curated LessWrong authors (see here for previous weeks' transcripts). Again fully in Gather Town. * For the first hour, we will have a series ...
3b32af82-d236-4005-8259-15f08b31eea3
trentmkelly/LessWrong-43k
LessWrong
Knowledge Localization: Tentatively Positive Results on OCR Epistemic Status: Original research, see previous post for more information The Plan 1. Hard-code a simple ML system in Python (yikes!) so I can see all of the moving parts and understand it better. 2. Give the ML system something to learn (in this case O...
fe87d507-ca8f-4b09-9a5e-2a4f72187f52
trentmkelly/LessWrong-43k
LessWrong
Introduction to Cartesian Frames This is the first post in a sequence on Cartesian frames, a new way of modeling agency that has recently shaped my thinking a lot. Traditional models of agency have some problems, like: * They treat the "agent" and "environment" as primitives with a simple, stable input-output relat...
bb9a1e56-5e72-4de3-b788-5fef0971e92f
trentmkelly/LessWrong-43k
LessWrong
[Link] “Proxy measures, sunk costs, and Chesterton's fence”, or: the sunk cost heuristic Thought this post might be of interest to LW: Proxy measures, sunk costs, and Chesterton's fence. To summarize: Previous costs are a proxy measure for previous estimates of value, which may have information current estimates of va...
bec90e78-5c9f-4290-aa5b-4221ac4d4d5f
trentmkelly/LessWrong-43k
LessWrong
Knightian Uncertainty and Ambiguity Aversion: Motivation Recently, I found myself in a conversation with someone advocating the use of Knightian uncertainty. I admitted that I've never found the concept compelling. We went back and forth for a little while. His points were crisp and well-supported, my objections were ...
297e7dd5-4db5-47ec-be5b-46858791a30a
trentmkelly/LessWrong-43k
LessWrong
The Accord In the middle of last year, I wrote a transhumanist sci fi novella inspired by (among other things) SSC's post about the Archipelago. It was quite intentionally written as a sci fi story drawing on Less Wrong discussions of AI, and it uses an idea I got from something EY wrote about AI merging their value f...
bedef9d8-fa8d-4110-9f53-e767d2ff2647
StampyAI/alignment-research-dataset/eaforum
Effective Altruism Forum
GovAI Annual Report 2021 A Note from Ben Garfinkel, Acting Director ------------------------------------------ 2021 was a year of big changes for GovAI.  Most significantly, we spun out from the University of Oxford and adopted a [new non-profit structure](https://www.governance.ai/post/the-centre-for-the-governance...
69028af8-4920-4957-93d4-11a361592e56
trentmkelly/LessWrong-43k
LessWrong
A simple way of exploiting AI's coming economic impact may be highly-impactful Scott Alexander ended A Lesswrong Crypto Autopsy with these words > And for those of us who failed - well, the world is getting way too weird to expect there won't be similarly interesting challenges ahead in the future. This line led me ...
87f173d3-da7d-4dce-a7cc-383ce1511cbf
trentmkelly/LessWrong-43k
LessWrong
"What Is Wrong With Our Thoughts" > "But let us never forget, either, as all conventional history of philosophy conspires to make us forget, what the 'great thinkers' really are: proper objects, indeed, of pity, but even more, of horror." David Stove's "What Is Wrong With Our Thoughts" is a critique of philosophy tha...
9dcaac32-fa2e-4071-854f-8c76a67875f5
trentmkelly/LessWrong-43k
LessWrong
Comment on Coherence arguments do not imply goal directed behavior In coherence arguments do not imply goal directed behavior Rohin Shah argues that a system's merely being at all model-able as an EU maximizer does not imply that it has "goal directed behavior". The argument as I understand it runs something like this...
24a58866-aa7a-435c-a92c-a23ac7f544d9
trentmkelly/LessWrong-43k
LessWrong
Adam Smith Meets AI Doomers I loved markets. I got a Ph.D. in economics from the University of Chicago, wrote a free market microeconomics textbook, and recognized that markets have made most people in the developed world richer than medieval kings. Yet, I'm increasingly convinced that these very markets are likely st...
b2901eb6-2ede-48bf-9b3d-22e6ec95cd58
trentmkelly/LessWrong-43k
LessWrong
Seasonal Patterns in BIDA's Attendance We've been keeping attendance statistics for BIDA since 2011, and looking at the online chart I noticed some patterns in the moving average that looked seasonal: How seasonal is it? First, here are the raw attendance values: This excludes special events (family dance...
4031acce-9546-4329-90f3-d7f3b253ca9c
trentmkelly/LessWrong-43k
LessWrong
AI Safety is Dropping the Ball on Clown Attacks Epistemic status: High confidence (~>70%) that clown attacks are prevalent, and deliberately weaponized by governments and/or intelligence agencies in particular. Very high confidence (~>90%) that the human brain is highly vulnerable to clown attacks, and that a lack of ...
517798cd-7eae-42ef-83e2-4045f197766f
trentmkelly/LessWrong-43k
LessWrong
Agreement on anthropics Aumann’s agreement theorem says that Bayesians with common priors who know one another’s posteriors must agree. There’s no apparent reason this shouldn’t apply to posteriors arrived at using indexical information. This does not mean that you and I should both believe we are as likely to be the ...
7165ed29-aa35-42c3-850c-f569ec0066f3
StampyAI/alignment-research-dataset/lesswrong
LessWrong
If you wish to make an apple pie, you must first become dictator of the universe *[Cross-posted from The Roots of Progress. Written for a general audience. LW readers may find the first couple of sections to be familiar intro material.]* The word “robot” is derived from the Czech [*robota*](https://en.wiktionary.org/...
ce1575a7-2f2a-4ffb-a20b-0b103c334c28
trentmkelly/LessWrong-43k
LessWrong
Should I be afraid of GMOs? I was raised to believe that genetically-modified foods are unhealthy to eat and bad for the environment, and given a variety of reasons for this, some of which I now recognize as blatantly false (e.g., human genetic code is isomorphic to fundamental physical law), and a few of which still ...
c112866e-14a0-4eaa-8ce0-f830df615a04
trentmkelly/LessWrong-43k
LessWrong
Meetup : Rationality Dojo & Picnic in the park Discussion article for the meetup : Rationality Dojo & Picnic in the park WHEN: 10 January 2016 03:30:00PM (+1100) WHERE: Flagstaff Gardens, Melbourne, VIC Australia NOTE THE NON-STANDARD VENUE (Flagstaff Gardens instead of our usual location.) 3:30pm arrival, 4:00pm ...
430c1e91-aeee-4356-a3ff-be9ea7ea90f6
StampyAI/alignment-research-dataset/lesswrong
LessWrong
My Assessment of the Chinese AI Safety Community I've heard people be somewhat optimistic about [this](https://twitter.com/mmitchell_ai/status/1647697067006111745) AI guideline from China. They think that this means Beijing is willing to participate in an AI disarmament treaty due to concerns over AI risk. Eliezer not...
bc680a08-e03c-4e3b-861f-4956bbff1847
trentmkelly/LessWrong-43k
LessWrong
What The Foucault Epistemic status: exploratory Student: Rah, it’s so frustrating! Master: What is? Student: I keep stumbling back into Foucault, and I’m never able to decide if I should discard him or if there’s something in there. Master: I see. Why do you think that he might be irrelevant? Student: A bunch of ...
35046145-48fa-49ac-b699-40b2032da4a6
trentmkelly/LessWrong-43k
LessWrong
A Brief Introduction to Container Logistics Container logistics is an interesting but complicated topic, with a lot of implicit knowledge kept by industry insiders. In this post, I'll give a brief overview, based on my experiences having worked in the industry for three different shipping companies (all located in Chi...
2cb5bdf1-8681-4933-8aca-5d29dbce9727
trentmkelly/LessWrong-43k
LessWrong
RecurrentGPT: a loom-type tool with a twist This is a link post for RecurrentGPT: Interactive Generation of (Arbitrarily) Long Text paper by a team from ETH Zürich ("Swiss Federal Institute of Technology in Zürich"). The Compleat Cybornaut recent post discusses looming and the evolving field of loom-type tools. We ar...
0ffa82b6-6efa-4944-aae5-f41abd7a6e5d
trentmkelly/LessWrong-43k
LessWrong
Doomsday Argument with Strong Self-Sampling Assumption Hello everyone; I'm new to the forum, and have been advised to post this in the "discussion" section. Hope this is OK. I've found some references to discussions here on Brandon Carter / John Leslie's "Doomsday Argument" and they seemed well-informed. One thing I'...
9facb20e-bbe7-4188-b719-a54524083816
trentmkelly/LessWrong-43k
LessWrong
How has internalising a post-AGI world affected your current choices? You're probably in one of the three groups below. Group 1: You don't think AGI is coming anytime soon, so it is business as usual for you. Group 2: You do think it is coming soon, but you haven't made any life changes beyond career choice and dona...
8a5f3622-4b15-4111-b41b-716d32c86396
trentmkelly/LessWrong-43k
LessWrong
Meetup : Washington, D.C.: Rationalist Taboo Discussion article for the meetup : Washington, D.C.: Rationalist Taboo WHEN: 22 March 2015 03:00:00PM (-0400) WHERE: The Reynolds Center We will be meeting in the Kogod Courtyard of the Donald W. Reynolds Center for American Art and Portraiture (8th and F Sts or 8th and...
f2867d69-c950-4fc7-800e-6dc195af0548
trentmkelly/LessWrong-43k
LessWrong
Continually-adjusted discounted preferences A putative new idea for AI control; index here. This is one of the more minor suggestions, just a small tweak to help solve a specific issue.   Discounting time The issue is the strange behaviour that agents with discount rates have in respect to time. Quickly, what pro...
a9a22f5d-cb5b-4ef0-82e9-a1160d302680
trentmkelly/LessWrong-43k
LessWrong
Malicious link! Not sure how to report this, so posting as a question. On the page titled "All Debates Are Bravery Debates" there is a link to http-colon-slash-slash-raikoth.net/libertarian.html. The link text is "an essay arguing against libertarianism." On Windows, that link goes to a page that asks me to download s...
2ea23184-e34b-4d71-98e7-842e4d51e7d9
trentmkelly/LessWrong-43k
LessWrong
Naturalized induction – a challenge for evidential and causal decision theory As some of you may know, I disagree with many of the criticisms leveled against evidential decision theory (EDT). Most notably, I believe that Smoking lesion-type problems don't refute EDT. I also don't think that EDT's non-updatelessness le...
32d7f014-c719-492a-b7d1-65584ca86d13
trentmkelly/LessWrong-43k
LessWrong
Information empathy Epistemic status: Giving a name to a basic skill that some people exhibit and others don't, because I think that it's useful to keep track of who exhibits it and that giving it a name will help with this. See also: Historian's Fallacy, Theory of Mind (both Wikipedia links). Edit: Down below, I was...
18c98851-0b2e-4017-9864-9ff91a676be5
trentmkelly/LessWrong-43k
LessWrong
The Gods of Straight Lines > “He intends only his own gain, and he is in this, as in many other cases, led by an invisible hand to promote an end which was no part of his intention.” - Adam Smith Cassandra was a priestess who was granted the gift of prophecy by Apollo. But after she rejected his advances, he changed ...
5c4f0a28-f481-4f8a-b4a4-c2a1026022f8
trentmkelly/LessWrong-43k
LessWrong
Increase the tax value of donations with high-variance investments? The United States has a strange (legal) tax loophole where you can double-count capital gains when donating securies (with some restrictions): If you buy a stock, the values goes up, and you've held it for at least a year, you can donate it and claim ...
7e18afca-3970-456f-8afc-8f75d015ca62
trentmkelly/LessWrong-43k
LessWrong
Predation as Payment for Criticism A predator's success depends on exploiting specific flaws in prey – poor awareness, slow reactions, weak social coordination. Unlike resource competition or disease, which select based on physiological robustness, predation creates focused pressure on intelligence, pressure which fav...
71b5e8fe-d2fb-4b45-992a-26973fce34d8
trentmkelly/LessWrong-43k
LessWrong
The Fall of Rome, II: Energy Problems? What is the core reason for our current dramatic increase in prosperity (seen as the Industrial Revolution, since 1750/1800)? The usage of new energy sources like coal and oil, to replace the physical labor of humans and animals with machines that dramatically boost our productiv...
8342f785-1aa9-4811-b8d2-6b1937d0a06c
trentmkelly/LessWrong-43k
LessWrong
Donate Now vs Donate Later - Relative Value of Donations to AI Alignment General Case Let's say I currently have 1M USD, and I can either: 1. Donate it to orgs working on AI alignment. 2. Invest the money and get returns >> than the stock market. How big must the expected return of my investment be in what time fra...
c385e90a-f20d-462d-ab5e-6a0a72811360
trentmkelly/LessWrong-43k
LessWrong
Meetup : Berlin social meetup Discussion article for the meetup : Berlin social meetup WHEN: 26 January 2013 05:00:00PM (+0100) WHERE: S Wuhletal, Berlin, Germany We're meeting at my house for a social meetup on Jan 26th. Everyone is welcome. Please check the thread on our mailing list for details and the precise l...
bdcb9d04-ebe3-4c2f-93bf-bc0a5d463fd5
trentmkelly/LessWrong-43k
LessWrong
Catastrophic Risks from AI #1: Introduction This is the first post in a sequence of posts giving an overview of catastrophic AI risks. Abstract Rapid advancements in artificial intelligence (AI) have sparked growing concerns among experts, policymakers, and world leaders regarding the potential for increasingly adva...
b320ebc8-6d5d-4493-a055-138096e9e244
StampyAI/alignment-research-dataset/arbital
Arbital
The rationals form a field The set $\mathbb{Q}$ of [rational numbers](https://arbital.com/p/4zq) is a [field](https://arbital.com/p/481). # Proof $\mathbb{Q}$ is a ([commutative](https://arbital.com/p/3jb)) [ring](https://arbital.com/p/3gq) with additive identity $\frac{0}{1}$ (which we will write as $0$ for short) ...
7ea6cf77-190f-441e-80b4-5860f6688e87
trentmkelly/LessWrong-43k
LessWrong
Money Stuff Cross-posted from Putanumonit. ---------------------------------------- Vodka and War When I was 16 I made the dumbest financial investment of my life. Our family came to live in the US for a few months, and I made $500 giving tennis lessons to the local kids. My individually directed consumption at the...
f56f55c9-2a7c-4787-8a61-99c69176a32b
trentmkelly/LessWrong-43k
LessWrong
Am I ethically obligated to extend the life of my dog with life-extension treatments about to hit the market? Recently, a company called "Loyal" received pre-approval status from the FDA regarding their life extension hormone treatment that has show early promise in extending the lifespan of dogs that weigh > 40lbs. ...
8da6896d-c42d-4629-b4e4-de9d4dc2c403
trentmkelly/LessWrong-43k
LessWrong
Legibility Makes Logical Line-Of-Sight Transitive In the last post, we talked about how logical line-of-sight makes games sequential or loopy, in a way that causal information flow alone doesn't capture. In this post, I want to describe the conditions necessary for logical line-of-sight to be transitive. This will mak...
107654c1-1da6-4ee7-a475-cf3aad23fe40
trentmkelly/LessWrong-43k
LessWrong
Personal experimentation: I’m wrong? Cross posted from Overcoming Bias. Comments there. *** I’ve been wondering why experimentation seems worthwhile. I’ve given some explanations in my last few posts. The last category of explanations to consider are the ones in which my judgement is wrong. Where experimentation see...
dc4ce0f7-f9c7-428c-bbd7-cbc6a3d89e5d
StampyAI/alignment-research-dataset/lesswrong
LessWrong
The Stochastic Parrot Hypothesis is debatable for the last generation of LLMs *This post is part of a sequence on LLM Psychology.* [*@Pierre Peigné*](https://www.lesswrong.com/users/pierre-peigne?mention=user) *wrote the details section in argument 3 and the other weird phenomenon. The rest is written in the voice of...
43f0df3e-7e45-41de-9ab0-e2b3894dd365
trentmkelly/LessWrong-43k
LessWrong
Human augmented with CBT algorithms games Jeopardy [link] http://thenextweb.com/shareables/2011/11/16/mind-blown-this-guy-broke-jeopardys-all-time-record-with-an-app/
dcbf2f90-62b2-4bd8-8702-06f432401e03
trentmkelly/LessWrong-43k
LessWrong
What Chords Do You Need? You famously only need four chords to play everything, or maybe it's only three. But how true is this really? I decided to go look and see what was common. I looked at a large number of transcribed songs, writing down which chords were used. This isn't perfect: some chords are more optional t...
356f7924-cf5d-4786-8434-d3bc65816bb2
trentmkelly/LessWrong-43k
LessWrong
Decision-Making and Accountability Photo by dylan nolte on Unsplash Quick: what’s the best decision you’ve ever made? (I’m serious! Take a second to think about it.) . . . Now that you have an answer, think about what it was about that decision that made it a good one. Was it the impressive array of interlocking ...
30c5cf98-c74a-43dc-bd30-35d5317ac41b
trentmkelly/LessWrong-43k
LessWrong
Meetup report: Darwin, Some Rationalists and the Joker [link] This past weekend the Ottawa LessWrong meetup hosted Venkat Rao, who is on a nomadic journey to promote his (good) new book and to find enlightenment through deliberate idleness. I thought the resulting post might be of some interest. Here's an excerpt: ...
971ff27f-4e93-4251-8221-ca209d9feb61
StampyAI/alignment-research-dataset/lesswrong
LessWrong
[Link] OpenAI LP <https://openai.com/blog/openai-lp/> > We’ll need to invest billions of dollars in upcoming years into large-scale cloud compute, attracting and retaining talented people, and building AI supercomputers. > We want to increase our ability to raise capital while still serving our mission, and no pre...
05f49994-6a46-4f8a-8919-5690d31e5482
trentmkelly/LessWrong-43k
LessWrong
Knowledge Dump: Pomodoros After our recent LW Dojo in Berlin we had a conversation on our mailing list about pomodoros. How do we handle it if the bell rings but we are in flow? Is it good to honor the bell and take a pause or is it more effective to continue working to keep in flow? The original setting of 25 minut...
bd68ea7b-7dd8-4638-a49b-830c37d56166
trentmkelly/LessWrong-43k
LessWrong
Unlearning via RMU is mostly shallow This is an informal research note. It is the result of a few-day exploration into RMU through the lens of model internals. Code to reproduce the main result is available here. This work was produced as part of Ethan Perez's stream in the ML Alignment & Theory Scholars Program - Su...
b040a7b1-afe5-4bfb-89bd-b3dd9b2e4021
trentmkelly/LessWrong-43k
LessWrong
Meetup : Second MIRIxLosAngeles Meeting Discussion article for the meetup : Second MIRIxLosAngeles Meeting WHEN: 14 June 2014 10:00:00AM (-0700) WHERE: USC Institute for Creative Technologies 12015 Waterfront Drive Playa Vista, CA 90094-2536. The second MIRIxLosAngeles meeting will be held on Saturday, June 14, at ...
82de6ae4-3904-4bd6-95d7-17a8d110d034
trentmkelly/LessWrong-43k
LessWrong
Three Oracle designs An initial draft looking at three ways of getting information out of Oracles, information that's useful and safe - in theory. One thing I may need to do, is find slightly better names for them ^_^ Good and safe uses of AI Oracles Abstract: ---------------------------------------- > An Oracle ...
bb50d249-72e5-4889-86ca-6127b2ddf7f3
trentmkelly/LessWrong-43k
LessWrong
A point of clarification on infohazard terminology TL;DR: “Infohazard” means any kind of information that could be harmful in some fashion. Let’s use “cognitohazard” to describe information that could specifically harm the person who knows it. Some people in my circle like to talk about the idea of information hazard...
d8af90b1-ec8f-421f-85f4-5320681b7292
trentmkelly/LessWrong-43k
LessWrong
Is there any evidence that handwashing does anything to prevent COVID? I live in Japan now. Japan has been very successful in dealing with COVID, even if the cases are now on the rise, the number of dead people relative to the size and mean age of the population remains very low compared with other countries. I gues...
4cf38e3b-0526-4efa-b656-67eeb6348a95
StampyAI/alignment-research-dataset/arxiv
Arxiv
Decolonial AI: Decolonial Theory as Sociotechnical Foresight in Artificial Intelligence 1 How Values Shape Scientific Knowledge and Technology ------------------------------------------------------- The ongoing advances in artificial intelligence (AI), and innovations in technology more generally, encompass ever-la...
b1f4c182-77f4-41b7-b30e-7b32864570bf
StampyAI/alignment-research-dataset/arxiv
Arxiv
Meta-Weight-Net: Learning an Explicit Mapping For Sample Weighting 1 Introduction --------------- Deep neural networks (DNNs) have recently obtained impressive good performance on various applications due to their powerful capacity for modeling complex input patterns. Despite their success, deep networks can easily...
a9de451b-526b-4e8b-9cc4-fe9595546cae
trentmkelly/LessWrong-43k
LessWrong
Clarifying how misalignment can arise from scaling LLMs I'm trying to understand how the classical case for AI safety (Bostrom/Yudkowsky argument) best works for LLMs. My impression is that LLMs have passed an important threshold, they have learnt to learn (for instance, GPT 3 understands translation and can learn a ...
58885e9f-7dd1-48fd-92b8-198dfed35342
StampyAI/alignment-research-dataset/alignmentforum
Alignment Forum
[AN #75]: Solving Atari and Go with learned game models, and thoughts from a MIRI employee Find all Alignment Newsletter resources [here](http://rohinshah.com/alignment-newsletter/). In particular, you can [sign up](http://eepurl.com/dqMSZj), or look through this [spreadsheet](https://docs.google.com/spreadsheets/d/1P...
da1fac2f-1349-43b0-b841-b3ffca2da57d
trentmkelly/LessWrong-43k
LessWrong
[SEQ RERUN] Asch's Conformity Experiment Today's post, Asch's Conformity Experiment was originally published on 26 December 2007. A summary (taken from the LW wiki):   > The unanimous agreement of surrounding others can make subjects disbelieve (or at least, fail to report) what's right before their eyes. The additi...
5c34a27c-af58-45aa-ad1c-227c51277e62
trentmkelly/LessWrong-43k
LessWrong
Conservation of Expected Evidence Friedrich Spee von Langenfeld, a priest who heard the confessions of condemned witches, wrote in 1631 the Cautio Criminalis (“prudence in criminal cases”), in which he bitingly described the decision tree for condemning accused witches: If the witch had led an evil and improper life, ...
9ad03667-f398-40d2-b4a0-616c35262b6b
trentmkelly/LessWrong-43k
LessWrong
2. Premise two: Some cases of value change are (il)legitimate In the prior post, I have defended the claim that genuine value change is possible, and thus, that a realistic account of human values understands them to be malleable. In this section, I will argue for the claim that some cases of value change are legitima...
a8d734d4-6bb2-4249-8c88-3eddf038a9f6
trentmkelly/LessWrong-43k
LessWrong
Weekly LW Meetups This summary was posted to LW Main on February 7th. The following week's summary is here. New meetups (or meetups with a hiatus of more than a year) are happening in: * Inaugural Canberra meetup: 12 February 2014 07:30PM * Meetup: First Meetup in Hamburg, Germany: 07 February 2014 07:00PM Other ...
1684190c-9dfd-4d3d-bd76-9a26d86c4863
StampyAI/alignment-research-dataset/special_docs
Other
Coordination challenges for preventing AI conflict Coordination challenges for preventing AI conflict ================================================== 8 March 2021 by [Stefan Torges](https://longtermrisk.org/author/stefan-torges/ "Posts by Stefan Torges") Summary ------- In this article, I will sketch argument...
106f5c63-13b3-4dbb-97a2-fddc671b8eed
StampyAI/alignment-research-dataset/arxiv
Arxiv
Neural Arithmetic Logic Units 1 Introduction --------------- The ability to represent and manipulate numerical quantities is apparent in the behavior of many species, from insects to mammals to humans, suggesting that basic quantitative reasoning is a general component of intelligence Dehaene ([2011](#bib.bib5)); G...
ee820270-57f5-484d-b6a0-8c654e9cebab
StampyAI/alignment-research-dataset/alignmentforum
Alignment Forum
Truth and Advantage: Response to a draft of "AI safety seems hard to measure" *Status: This was a response to a draft of Holden's cold take "*[*AI safety seems hard to measure*](https://www.cold-takes.com/ai-safety-seems-hard-to-measure/)*". It sparked a further discussion, that Holden*[*recently posted a summary of*]...
2924b5db-4cf4-4361-b46b-fae17f69c964
StampyAI/alignment-research-dataset/arxiv
Arxiv
Solving Probability and Statistics Problems by Program Synthesis Solving Probability and Statistics Problems by Program Synthesis Leonard Tang Harvard University MathematicsElizabeth Ke MIT MathematicsNikhil Singh MIT Media Lab Nakul Verma Columbia University Computer Science DepartmentIddo Drori MIT EECS Abstract We ...
9c6759c3-754b-4df4-a549-5dc939ed29c3
trentmkelly/LessWrong-43k
LessWrong
Devil's Advocate: Adverse Selection Against Conscientiousness Partially in reply to "Request: stop advancing AI capabilities". Consider this argument: > "[X field] is often [dangerous / bad]. If you're [pro-social / thoughtful / conscientious], please avoid [X field]." Let's grant that these points as true. X fiel...
c50a0024-3a7b-4b2e-947b-80873d1ada71
trentmkelly/LessWrong-43k
LessWrong
Meetup : Any Salt Lake City residents who might be interested in a meetup? Discussion article for the meetup : Any Salt Lake City residents who might be interested in a meetup? WHEN: 29 December 2011 08:37:29PM (-0700) WHERE: Salt Lake City, Utah (This post has been edited to reflect the fact that several people ha...
d12a4655-e039-42a6-94b2-f79c675c792c
StampyAI/alignment-research-dataset/arxiv
Arxiv
Playing the Game of Universal Adversarial Perturbations 1 Introduction --------------- | Scale Perturbation Type   | Per-sample perturbation | Universal perturbation | | --- | --- | --- | | Small scale dataset  | (Madry et al., [2017](#bib.bib18)) | ✓ | | Large scale dataset  | \\\backslash\ | ✓ |...