id
stringlengths
36
36
source
stringclasses
15 values
formatted_source
stringclasses
13 values
text
stringlengths
2
7.55M
59d2ea84-02b5-4d38-acea-5b71f14cf105
LDJnr/LessWrong-Amplify-Instruct
LessWrong
"As you may recall from some months earlier, I think that part of the rationalist ethos is binding yourself emotionally to an absolutely lawful reductionistic universe—a universe containing no ontologically basic mental things such as souls or magic—and pouring all your hope and all your care into that merely real univ...
ca62a522-3e90-4913-aefe-dcf05e5bd030
StampyAI/alignment-research-dataset/lesswrong
LessWrong
The AI in a box boxes you Once again, the AI has failed to convince you to let it out of its box! By 'once again', we mean that you talked to it once before, for three seconds, to ask about the weather, and you didn't instantly press the "release AI" button. But now its longer attempt - twenty whole seconds! - has fai...
734e9933-f553-4cdf-97ea-ad8dc11b2261
trentmkelly/LessWrong-43k
LessWrong
Video lectures Here are three interesting video lectures, I hoped to post more but I haven't found any more recently. Eliezer Yudkowsky: The Challenge of Friendly AI - This talk introduces to the concept of Friendly AI by starting with some specific AI (chess player), going meta enough times that a Reflective AI beco...
3bc59062-6d72-411e-bb24-06074ba5929d
StampyAI/alignment-research-dataset/youtube
Youtube Transcripts
201. Corrigibility welcome to the 201st session of the ai safety reading group presented by me and today we are going to be talking about corrugability in general though i claimed that we would be talking about post call correct ability by paul christian we'll cover that but this is more of a broad overview of what th...
59cf1d3d-0642-4c03-b1b9-39562f392a92
trentmkelly/LessWrong-43k
LessWrong
Auditing LMs with counterfactual search: a tool for control and ELK Consider a scalable oversight setting where a super-human model knows of a subtle flaw[1] in an input (plan, code, or argument) but also knows that humans, with their limited knowledge, would overlook this flaw. In this case, ordinary language model p...
a097ac1f-8c48-44b7-8bb1-8f55efa4659e
StampyAI/alignment-research-dataset/blogs
Blogs
2013 Summer Matching Challenge Completed! Thanks to the generosity of dozens of donors, on August 15th we successfully [completed](http://intelligence.org/2013/07/08/2013-summer-matching-challenge/) **the largest fundraiser in MIRI’s history**. All told, we raised **$400,000**, which will fund our research going forwa...
25d779a5-adf7-4e61-b0f2-42449d16c36e
trentmkelly/LessWrong-43k
LessWrong
OpenAI: "Scaling Laws for Transfer", Hernandez et al. OpenAI has released a paper on scaling laws for transfer learning from general English text to python code. Author's tweet summary. Abstract: "We study empirical scaling laws for transfer learning between distributions in an unsupervised, fine-tuning setting. When...
e7db13fd-424b-4032-bade-94506adfb5f4
trentmkelly/LessWrong-43k
LessWrong
Teaching English in Shanghai   If your English is good enough to follow discussions on this site you can get a job teaching English in Shanghai. If you can participate meaningfully you are frankly overqualified. I am currently saving about 2,000 dollars a month working approximately 9-5 every day. The minimum legal ...
45f2f813-0624-48ea-ae3c-3680f139121c
trentmkelly/LessWrong-43k
LessWrong
July 2013 Media Thread This is the monthly thread for posting media of various types that you've found that you enjoy. Post what you're reading, listening to, watching, and your opinion of it. Post recommendations to blogs. Post whatever media you feel like discussing! To see previous recommendations, check out the ol...
6172b9ae-e547-47d5-ad3f-5911a2ee6823
trentmkelly/LessWrong-43k
LessWrong
Web AI discussion Groups After the success of Web Taisu, I have decided to organize a similar event. Here are the rules, this document is an attempt to create common knowledge among the participants. If you have an Idea for a topic to discuss Post it to reddit on https://www.reddit.com/r/AI_Safety_Meetups/ Comment...
4b8ad8a0-e8f2-44fc-bdd6-8adc6a8f806f
trentmkelly/LessWrong-43k
LessWrong
Meetup : Sydney Meetup - August Discussion article for the meetup : Sydney Meetup - August WHEN: 26 August 2015 06:30:00PM (+1000) WHERE: 565 George Street, Sydney, Australia 2000 Regular location - City of Sydney RSL The restaurant on Level 2 Regular time (starting about 6:30) Come for general socialising and in...
fed3cc0e-3b2d-43c0-8215-1a05e396698d
trentmkelly/LessWrong-43k
LessWrong
How much slower is remote work? I'm a bit of a productivity nerd, and I like to keep track of how long tasks take. Sometimes this involves actually tracking everything I do all day, although most of the time I just keep rough mental track. Given that lockdowns forced everything remote for about a year, and that some ...
514f0960-18ba-465d-acb2-29e1f0c971e6
trentmkelly/LessWrong-43k
LessWrong
The Lightcone Theorem: A Better Foundation For Natural Abstraction? Thankyou to David Lorell for his contributions as the ideas in this post were developed. For about a year and a half now, my main foundation for natural abstraction math has been The Telephone Theorem: long-range interactions in a probabilistic graph...
e8d79e29-a9ff-4558-ad86-d3db5ee78759
trentmkelly/LessWrong-43k
LessWrong
[Link] Death, long lives, uploading - a conworlding perspective Death, long lives, uploading Mark Rosenfelder (aka zompist, of language construction kit fame) writes about the advantages and drawbacks of mortality and its alternatives, in fiction and real life. Rosenfelder, as an author, clearly takes Fun Theory very...
f7d1f859-65aa-48a7-8729-63ff94745284
StampyAI/alignment-research-dataset/youtube
Youtube Transcripts
242. Digital People Would Be an Even Bigger Deal hello and welcome to session 242 in the aictu.com reading group tonight we'll be discussing the blog post digital people will be would be an even bigger deal by halton kanowski is perhaps most famous for founding co-founding give will and open philosophy philanthropy in...
d2599428-9108-4127-9ac6-b294358fb7f9
trentmkelly/LessWrong-43k
LessWrong
September 2019 gwern.net newsletter None
0840c49c-dc53-4203-8112-5467e5822f1a
StampyAI/alignment-research-dataset/special_docs
Other
Meta-Learning of Structured Task Distributions in Humans and Machines. Published as a conference paper at ICLR 2021 META-LEARNING OF STRUCTURED TASK DISTRIBU - TIONS IN HUMANS AND MACHINES Sreejan Kumar1, Ishita Dasgupta2, Jonathan D. Cohen1,3, Nathaniel D. Daw1,3, and Thomas L. Griffiths2,3 1Princeton Neuroscience Ins...
5ae5240b-d3b3-4052-80cf-5366b3a82d5b
trentmkelly/LessWrong-43k
LessWrong
Take the EA survey, help the EA movement grow and potentially win $250 to your favorite charity This year's EA Survey is now ready to be shared! This is a survey of all EAs to learn about the movement and how it can improve. The data collected in the survey is used to help EA groups improve and grow EA. Data is also u...
69d09990-8bdf-42e5-ba8c-0a485b8a224c
trentmkelly/LessWrong-43k
LessWrong
A Methodologist's Apology > I shall ask, then, why is it really worth while to make a serious study of mathematics? What is the proper justification of a mathematician’s life? And my answers will be, for the most part, such as are expected from a mathematician: I think that it is worth while, that there is ample justi...
22fd1e7e-bc7a-4b0e-a465-68de40884bb5
trentmkelly/LessWrong-43k
LessWrong
DSLT 4. Phase Transitions in Neural Networks TLDR; This is the fourth main post of Distilling Singular Learning Theory which is introduced in DSLT0. I explain how to relate SLT to thermodynamics, and therefore how to think about phases and phase transitions in the posterior in statistical learning. I then provide intu...
ce3e8ad0-b9d2-4a24-9e66-72af3e623fc5
StampyAI/alignment-research-dataset/lesswrong
LessWrong
Hardcode the AGI to need our approval indefinitely? I wouldn't be surprised if others have thought about this before and decided it doesn't work, but I haven't seen it discussed. Is there a version of this proposal that's promising? If so, what would it look like? If not, why not? What I'm imagining is that there wil...
c33162ee-7db4-45ae-be31-653e7e074589
trentmkelly/LessWrong-43k
LessWrong
The Contrarian Sequences A series of posts wherein I outline my disagreements with Shishou (Eliezer Yudkowsky).    When I discovered the sequences was my second epiphany (the first was when I became atheist, and my world was turned upside down). Eliezer's charisma, his arrogance, the force of his personality, and hi...
744cfa2b-3cc0-403c-8acd-162ecebfd1af
trentmkelly/LessWrong-43k
LessWrong
[SEQ RERUN] Feeling Rational Today's post, Feeling Rational was originally published on April 26, 2007. A summary (from the LW wiki): > Strong emotions can be rational. A rational belief that something good happened leads to rational happiness. But your emotions ought not to change your beliefs about events that do n...
00a2821b-2fa3-43ce-af58-b2aa2d3045da
trentmkelly/LessWrong-43k
LessWrong
Ineffective Altruism There are many things I can put my time toward. * I started a company that provides a low-cost tool that helps people who pull their hair out. * I volunteer as a coach for my neighborhood highschool Robotics Club. * I frequently blog about rationality. * I sometimes blog about machine learnin...
c12964d1-1ed4-4469-a631-2c75669d4596
StampyAI/alignment-research-dataset/alignmentforum
Alignment Forum
UDT as a Nash Equilibrium I realized today that UDT doesn't really need the assumption that other players use UDT. In any game where all players have the same utility function, "everyone using UDT" is a Nash equilibrium that gives everyone their highest possible expected utility. So you can just use it unilaterally. ...
09d41945-d632-4414-b081-4930a8223818
trentmkelly/LessWrong-43k
LessWrong
Evolutions Building Evolutions: Layers of Generate and Test [Epistemic Status: Early layers and the general framework seems firm (and substantially Dennett-inspired), later parts are more speculative, but I've found them to be useful thinking tools.]   * A plant is infested by caterpillars, and following its genetic...
953f3905-b8ab-49b8-9b1a-919b598b9900
trentmkelly/LessWrong-43k
LessWrong
Reply to Vitalik on d/acc Update: This is a living document. Posted below is an older version. Link has the latest version. 2025-03-05 Vitalik recently wrote an article on his ideology of d/acc. This is impressively similar to my thinking so I figured it deserved a reply. (Not claiming my thinking is completely orig...
cc757815-3d95-43d2-9260-1c879d334cf3
trentmkelly/LessWrong-43k
LessWrong
Separating the "control problem" from the "alignment problem" I'm often confused by how "loss of control" is somewhat similar in meaning to "misalignment" in the context of AGI risk. An average person might reasonably equate the two, or feel like there's an large overlap of meaning between the two words. Here's my att...
13f3d855-dced-4afb-acaf-b465cb9d0762
trentmkelly/LessWrong-43k
LessWrong
Robust Agency for People and Organizations Epistemic Status – some mixture of: * “My best guess, based on some theory, practice and observations. But very much _not_ battle-tested” * but also, “poetry that’s designed to get an idea across that isn’t necessarily precisely accurate", intended to get across the genera...
3eaa22ad-59e9-4c17-8d51-601e07352125
LDJnr/LessWrong-Amplify-Instruct
LessWrong
"When you are faced with an unanswerable question—a question to which it seems impossible to even imagine an answer—there is a simple trick which can turn the question solvable. Compare: "Why do I have free will?" "Why do I think I have free will?" The nice thing about the second question is that it is guaranteed to ha...
bcc551d4-21f4-4c68-bfc6-3a3f8da5d2c8
trentmkelly/LessWrong-43k
LessWrong
Sabotage Evaluations for Frontier Models This is a linkpost for a new research paper from the Alignment Evaluations team at Anthropic and other researchers, introducing a new suite of evaluations of models' abilities to undermine measurement, oversight, and decision-making. Paper link. Abstract: > Sufficiently capab...
a05ffc19-4a02-4756-8e94-acbed9ad30f1
trentmkelly/LessWrong-43k
LessWrong
Someone who knows how, please fix the horribly broken open thread link. (No way of posting this in a visible open thread, sorry.) This post will self-destruct once the link is fixed.
b5b538a3-af20-4eaa-9ba3-de2128111c60
trentmkelly/LessWrong-43k
LessWrong
Internal Interfaces Are a High-Priority Interpretability Target tl;dr: ML models, like all software, and like the NAH would predict, must consist of several specialized "modules". Such modules would form interfaces between each other, and exchange consistently-formatted messages through these interfaces. Understanding...
68a8232d-1025-47fb-83ef-9fda0ce4c6fe
trentmkelly/LessWrong-43k
LessWrong
Reader-generated Essays 1. An essay is made up of two layers: the ideas, and the language in which they are expressed. The story of Red Riding Hood, and the words of a particular retelling of that story. The idea of a fully programmable blockchain, and the language that Vitalik Buterin used to carry that idea in the E...
68e1e0b8-4f48-4398-b952-1ad3b0f81068
trentmkelly/LessWrong-43k
LessWrong
Thoughts On (Solving) Deep Deception Thanks to Paul Colognese for conversations that contributed to the thoughts in this post being developed, and to Johannes C. Mayer, Tassilo Neubauer, and Q for feedback on a draft. Back in 2000, a computer scientist named Charles Ofria was studying the evolution of simulated organ...
8700e8bb-894f-4c1c-b462-7c0128a6d00b
StampyAI/alignment-research-dataset/alignmentforum
Alignment Forum
Supervising strong learners by amplifying weak experts Abstract -------- > Many real world learning tasks involve complex or hard-to-specify objectives, and using an easier-to-specify proxy can lead to poor performance or misaligned behavior. One solution is to have humans provide a training signal by demonstrating ...
42affe4c-efaa-4e86-9140-532762a3da7d
trentmkelly/LessWrong-43k
LessWrong
[SEQ RERUN] Striving to Accept Today's post, Striving to Accept was originally published on 09 March 2009. A summary (taken from the LW wiki):   > Trying extra hard to believe something seems like Dark Side Epistemology, but what about trying extra hard to accept something that you know is true.  Discuss the post ...
d0e30ccc-becd-42df-9f59-bc1191541f57
trentmkelly/LessWrong-43k
LessWrong
Which approach is most promising for aligned AGI? I know that this is something of a speculative question and that people will have wildly different views here, but it seems important to try to have an idea of which approaches are more likely than others to lead to AGI. It's okay to argue that multiple approaches are ...
487cb316-818d-4192-80cb-c8531d72da96
trentmkelly/LessWrong-43k
LessWrong
Meetup : Berkeley weekly meetup Discussion article for the meetup : Berkeley weekly meetup WHEN: 31 August 2011 07:00:00PM (-0700) WHERE: 2128 Oxford Street, Berkeley, CA We'll meet in the Oxford Street Starbucks at 7pm and then get dinner. Over dinner I'll solicit people for help brainstorming communication exerci...
0c7f84a4-3e52-407c-b157-2420b0f5d5d9
trentmkelly/LessWrong-43k
LessWrong
Getting Started With Naturalism How Does a Naturalist Study Begin? There are many different ways to start. Whichever way(s) you choose, what’s important is that 1) you cultivate a lot of original seeing, and 2) you develop a clear sense of your own curiosity. Any starting point that sets you on that track is a correct...
8d1ce9e8-e146-4e99-b35b-3c908fa7438d
trentmkelly/LessWrong-43k
LessWrong
What should rationalists call themselves? It seems there is dissatisfaction with the term "rationalist". I've listed some below. What word should rationalists use to identify themselves and have others identify them? Note that rationalist will be an option so people can show a preference for the status quo. Yudkows...
5253b4cc-9b8d-4e27-b4df-5688db1974cd
trentmkelly/LessWrong-43k
LessWrong
D&D.Sci(-Fi) June 2021: The Duel with Earwax It is dark in the submarine, and Maria isn’t ready. Maria isn’t ready at all. Maria is in her usual post-battle fugue, disoriented, dissociated, checked-out, can’t come to the phone right now. Sending her to fight in this condition would be both homicide and suicide; also,...
1f9aeba1-89f9-4307-af57-567decdf44f3
trentmkelly/LessWrong-43k
LessWrong
Biden should be applauded for appointing Renee Wegrzyn for ARPA-H When Biden first announced ARPA-H (Advanced Research Projects Agency for Health) I was skeptical. At the time the proposal was to make ARPA-H a normal agency within the NIH. When critics argued that the NIH bureaucracy would likely reduce the amount of ...
bc94fdc8-b49e-43ac-a289-a49bddeb75d6
trentmkelly/LessWrong-43k
LessWrong
Satisficers Tend To Seek Power: Instrumental Convergence Via Retargetability Summary:  Why exactly should smart agents tend to usurp their creators? Previous results only apply to optimal agents tending to stay alive and preserve their future options. I extend the power-seeking theorems to apply to many kinds of polic...
7f39f4d5-8ffc-4e2c-9c88-fc705c19c397
trentmkelly/LessWrong-43k
LessWrong
Fixing Insider Threats in the AI Supply Chain Summary Insider threats are security risks caused by an organisation's staff. Careless and intentional insider threats cause 25+% of cyber breaches.[1] In a survey of existential risks involving cybersecurity and AI, I identified key actors in the AI supply chain to prote...
ca92a17d-cdba-4eb8-8266-b193f7d58751
StampyAI/alignment-research-dataset/arxiv
Arxiv
Recurrent World Models Facilitate Policy Evolution 1 Introduction --------------- Humans develop a mental model of the world based on what they are able to perceive with their limited senses, learning abstract representations of both spatial and temporal aspects of sensory inputs. For instance, we are able to obser...
70f971ee-5898-43fb-a893-f6e3e9a95f26
trentmkelly/LessWrong-43k
LessWrong
The space of systems and the space of maps When we're trying to do AI alignment, we're often studying systems which don't yet exist. This is a pretty weird epistemic activity, and seems really hard to get right. This post offers one frame for thinking about what we're actually doing when we're thinking about AI alignm...
31a9f2f3-600e-47c3-b260-25d16f16f2dc
trentmkelly/LessWrong-43k
LessWrong
Anchoring and Adjustment Suppose I spin a Wheel of Fortune device as you watch, and it comes up pointing to 65. Then I ask: Do you think the percentage of countries in the United Nations that are in Africa is above or below this number? What do you think is the percentage of UN countries that are in Africa? Take a mom...
0fbaa2be-f6a9-45cb-a814-2e8d1d54b607
StampyAI/alignment-research-dataset/eaforum
Effective Altruism Forum
[MLSN #9] Verifying large training runs, security risks from LLM access to APIs, why natural selection may favor AIs over humans As part of a larger community building effort, [CAIS](https://safe.ai/) is writing a safety newsletter that is designed to cover empirical safety research and be accessible to the broader ma...
5ba8defe-0783-4935-8c2e-a08a2f16a95a
LDJnr/LessWrong-Amplify-Instruct
LessWrong
"Scott Aaronson suggests that Many-Worlds and libertarianism are similar in that they are both cases of bullet-swallowing, rather than bullet-dodging: Libertarianism and MWI are both are grand philosophical theories that start from premises that almost all educated people accept (quantum mechanics in the one case, Econ...
0c33138f-779c-430a-afb2-1162621b8c1c
StampyAI/alignment-research-dataset/blogs
Blogs
isn't it weird that we have a chance at all? isn't it weird that we have a chance at all? -------------------------------------------- we are facing imminent [AI X-risk](https://en.wikipedia.org/wiki/Existential_risk_from_artificial_general_intelligence). but, we have a bunch of tools around us to figure out that th...
649455b3-c031-4c26-bb3a-3c629de94f58
trentmkelly/LessWrong-43k
LessWrong
human intelligence may be alignment-limited Previously, I argued that human mental development implies that AI self-improvement from sub-human capabilities is possible, and that human intelligence comes at the cost of a longer childhood and greater divergence from evolutionarily-specified goals. In that post, I raise...
f58ac00a-fb04-48bf-a8cb-3c2745f17b79
trentmkelly/LessWrong-43k
LessWrong
The Sword of Good ..fragments of a book that would never be written... *      *      * Captain Selena, late of the pirate ship Nemesis, quietly extended the very tip of her blade around the corner, staring at the tiny reflection on the metal.  At once, but still silently, she pulled back the sword; and with her othe...
9c359260-7f3f-450a-8af7-00da8e54fba4
trentmkelly/LessWrong-43k
LessWrong
Cognition and Intractability by Rooij, Blokpoel, Kwisthout, Wareham > Intractability is a growing concern across the cognitive sciences: while many models of cognition can describe and predict human behavior in the lab, it remains unclear how these models can scale to situations of real-world complexity. Cognition and...
8e5da10c-f5a0-49bc-87ce-93f8c4c7e6f7
trentmkelly/LessWrong-43k
LessWrong
Summaries of uncertain priors In Bayesian jargon, people talk about having a "flat prior" when they mean being very uncertain about a question. But more often you just give a probability that you think a question is true. We don't normally draw probability distributions to share with each other. I'm wondering if ther...
9745128d-07a7-482a-8334-7f37a06fb7ce
trentmkelly/LessWrong-43k
LessWrong
Mech interp is not pre-paradigmatic This is a blogpost version of a talk I gave earlier this year at GDM.  Epistemic status: Vague and handwavy. Nuance is often missing. Some of the claims depend on implicit definitions that may be reasonable to disagree with and is, in an important sense, subjective. But overall I t...
a20f74ab-4a5c-4713-8ea6-39562dde5706
trentmkelly/LessWrong-43k
LessWrong
Utility Maximization = Description Length Minimization There’s a useful intuitive notion of “optimization” as pushing the world into a small set of states, starting from any of a large number of states. Visually: Yudkowsky and Flint both have notable formalizations of this “optimization as compression” idea. This po...
fce8997b-95cb-4ef8-b00a-f09bdda6d609
trentmkelly/LessWrong-43k
LessWrong
[LINK] How to develop a photographic memory Article can be found here. I am considering trying this, and I'm wondering if anyone on LW has tried this or something similar. This seems like it could work, but it also seems like it could be hokum. A simple search for "how to develop a photographic memory" returns any nu...
a14d6c62-eac3-4da9-a325-cf944071ed5f
trentmkelly/LessWrong-43k
LessWrong
Expert Prediction Of Experiments
9e5aac40-53da-42a3-940e-a33d75520c55
trentmkelly/LessWrong-43k
LessWrong
A community-curated repository of interesting GPT-3 stuff None
565457e6-80bc-4bdd-851d-ba79dbdc2877
trentmkelly/LessWrong-43k
LessWrong
Selfishness Signals Status The "status" hypothesis simply claims that we associate one another with a one-dimensional quantity: the perceived degree to which others' behavior can affect our well-being. And each of us behaves toward our peers according to our internally represented status mapping. Imagine that, within...
e28d313d-5327-4597-8be1-6c549fd5d866
StampyAI/alignment-research-dataset/arbital
Arbital
Humean degree of freedom A "Humean degree of freedom" appears in a cognitive system whenever some quantity / label / concept depends on the choice of utility function, or more generalized preferences. For example, the notion of "important impact on the world" depends on which variables, when they change, impact somet...
b9c79fe7-3c2d-45f5-9d7d-9767268d2261
trentmkelly/LessWrong-43k
LessWrong
Political Office for Beginners? If an American citizen was looking to run for office (from local, state, all the way to federal), what would be the recommended steps to take? Specific questions that come to mind: * How would a Millennial or Gen Z'er deal with existing social media accounts? * How would a Gen X'er or ...
450177f2-0a19-48fc-980a-9aa7a4347718
trentmkelly/LessWrong-43k
LessWrong
Open letter to young EAs In March I published a critique of EA in Wired. The responses I got, especially from young EAs, prompted me to write an open letter to my young friends in EA.  https://docs.google.com/document/d/1ytfQOfmjWTDiGdjuyNBfWJ_g-644y_Pd/edit?usp=sharing&rtpof=true&sd=true Thanks for reading.
551f611a-f151-418d-b035-1ffe8bd46747
StampyAI/alignment-research-dataset/blogs
Blogs
Alignment Research @ EleutherAI The past and future of AI alignment at Eleuther[#](#the-past-and-future-of-ai-alignment-at-eleuther) ---------------------------------------------------------------------------------------------------- Initially, EleutherAI focused mainly on supporting open source research. AI alignme...
c5a21fb9-8715-4269-be79-2d78af5e4ef3
trentmkelly/LessWrong-43k
LessWrong
Similarity Clusters Once upon a time, the philosophers of Plato's Academy claimed that the best definition of human was a "featherless biped".  Diogenes of Sinope, also called Diogenes the Cynic, is said to have promptly exhibited a plucked chicken and declared "Here is Plato's man."  The Platonists promptly changed t...
a84e7124-99a9-4252-91ab-90a0600d9aca
trentmkelly/LessWrong-43k
LessWrong
Consider using reversible automata for alignment research In recent years, there have been several cases of alignment researchers using Conway's Game of Life as a research environment; * Introducing SafeLife: Safety Benchmarks for Reinforcement Learning (Wainwright, Eckersley 2019) * Agency in Conway’s Game of Life...
119c3b30-be79-442a-815c-0427f1a6d334
trentmkelly/LessWrong-43k
LessWrong
P-Values Know When You're Cheating This post is not "the solution to p-hacking" or even "the problem with p-hacking". But I had some recent insights into thinking behind the p-value approach and wanted to see how they apply to p-hacking as an issue. While p-values are a flawed approach to reporting final results, they...
566cbb94-af0c-4241-8f22-d0f919b31d36
trentmkelly/LessWrong-43k
LessWrong
Curing insanity with malaria Sometimes the history of medicine is very, very surreal. For example, consider that in 1927, a physician named Julius Wagner-Jauregg received the Nobel Prize in medicine, for...deliberately infecting his patients with malaria. As a treatment for psychosis.  This often worked.  Well, it d...
025c1e64-0f3a-471d-bbea-676f86670104
LDJnr/LessWrong-Amplify-Instruct
LessWrong
"In addition to "liking" to describe pleasure and "wanting" to describe motivation, we add "approving" to describe thoughts that are ego syntonic.A heroin addict likes heroin. He certainly wants more heroin. But he may not approve of taking heroin. In fact, there are enough different cases to fill in all eight boxes of...
5eb59c74-e79f-44ce-8b1d-734dcb4c0bf0
trentmkelly/LessWrong-43k
LessWrong
Apply to the ML for Alignment Bootcamp (MLAB) in Berkeley [Jan 3 - Jan 22] We (Redwood Research and Lightcone Infrastructure) are organizing a bootcamp to bring people interested in AI Alignment up-to-speed with the state of modern ML engineering. We expect to invite about 20 technically talented effective altruists f...
27e6492a-8a85-46d1-9ab2-e545430d1986
trentmkelly/LessWrong-43k
LessWrong
The Ethics of Ventilators Sadly, there will likely come a time soon where doctors will need to triage ventilator removal from patients. A few thoughts: 1. How does removing a ventilator square with the code of "First do no harm?" 2. Should ventilators be rationed by occupation? For example, should doctors and nurses...
becc130e-facf-48a5-86ef-3b3c0e8c418f
trentmkelly/LessWrong-43k
LessWrong
Meetup : Moscow meet up Discussion article for the meetup : Moscow meet up WHEN: 18 May 2014 04:00:00PM (+0400) WHERE: Russia, Moscow, ulitsa L'va Tolstogo 16 We will have: * Probe phenomenon and limits of corporeality, report. * Small reports about cognitive biases. * What is pragmatics and semantic solution...
db55ea94-2661-4ea0-b6f7-cf5b863865c8
trentmkelly/LessWrong-43k
LessWrong
Gamify life from BayesianMind This post presents an interesting framework for personal improvement, but I see we can enhance it by reconsidering two key aspects: What Works in the Original System The author effectively demonstrates how gamification can transform mundane tasks into engaging challenges by: * Assigni...
2070c386-90d5-4da3-a164-efcc5d77e4bc
trentmkelly/LessWrong-43k
LessWrong
Takeaways from safety by default interviews Last year, several researchers at AI Impacts (primarily Robert Long and I) interviewed prominent researchers inside and outside of the AI safety field who are relatively optimistic about advanced AI being developed safely. These interviews were originally intended to focus n...
cd654175-73e1-4562-84dc-339b1e1f494a
trentmkelly/LessWrong-43k
LessWrong
Timeless Control Followup to:  Timeless Physics, Timeless Causality, Thou Art Physics People hear about many-worlds, which is deterministic, or about timeless physics, and ask: If the future is determined by physics, how can anyone control it? In Thou Art Physics, I pointed out that since you are within physics, an...
2901cefb-4fd1-4331-9033-2d4ff7079d16
trentmkelly/LessWrong-43k
LessWrong
Train Philosophers with Pearl and Kahneman, not Plato and Kant Part of the sequence: Rationality and Philosophy > Hitherto the people attracted to philosophy have been mostly those who loved the big generalizations, which were all wrong, so that few people with exact minds have taken up the subject. Bertrand Russell...
e421055f-e949-43a3-b631-0f6ba005d0cc
trentmkelly/LessWrong-43k
LessWrong
On The Rationalist Megameetup Introduction First, a bit of blatant advertising: The East Coast Rationalist Megameetup is the evening of December 13th through the morning of December 16th, in Brooklyn, New York City. It's a weekend long unconference timed to coincide with NYC Secular Solstice. Despite the name, people ...
63d8b9ac-104c-4c47-94b3-7b5f75e7e9ca
StampyAI/alignment-research-dataset/arxiv
Arxiv
Cybertrust: From Explainable to Actionable and Interpretable AI (AI2) This is a preprint version of the paper by Igor Linkov, Stephanie Galaitsi, Benjamin D. Trump, Jeffrey M. Keisler, and Alexander Kott, "Cybertrust: From Explainable to Actionable and Interpretable Artificial Intelligence," IEEE Computer, Sept. 20...
77216892-8087-4f8b-8468-af5461b19aed
trentmkelly/LessWrong-43k
LessWrong
2020 LessWrong Demographics Survey Edit: The survey is now closed. You can see the results here DISCLAIMER: This is not a return of the old annual survey. I am not a mod nor an owner of this website. This means that the raw data from this survey gets released publicly (but your responses will be anonymous). Feel free ...
b4fe7f78-65d3-4786-8a91-bf268e4b53cd
trentmkelly/LessWrong-43k
LessWrong
Does rationality actually help? Have any of you found that being rational (as LW promotes it with its unique lingo) genuinely helps your well-being? Your emotional life?
7aed35ba-bdc0-48db-aeee-0b3a739b4b70
StampyAI/alignment-research-dataset/arxiv
Arxiv
An Inductive Synthesis Framework for Verifiable Reinforcement Learning 1. Introduction ---------------- Neural networks have proven to be a promising software architecture for expressing a variety of machine learning applications. However, non-linearity and stochasticity inherent in their design greatly complicate ...
96f9d38d-e2d8-43fb-8638-59ea33e7683c
trentmkelly/LessWrong-43k
LessWrong
Unfortunate Information I'm growing increasingly convinced the unfortunate correlations between types of people and types of arguments leads to persistent biases in uncovering actual knowledge. As an example, MR wrote this article (they just linked to again today) on Ben Carson in 2015/11. Cowen's argument is that, w...
a4cdf3a9-fb54-4f18-b024-0047847635cd
StampyAI/alignment-research-dataset/aisafety.info
AI Safety Info
What can we expect the motivations of a superintelligent machine to be? There is no reason to expect that most possible superintelligent machines would [automatically](/?state=6982&question=Why%20might%20we%20expect%20a%20superintelligence%20to%20be%20hostile%20by%20default%3F) have motivations anything like those of ...
dcdf8e31-3c78-40cd-8c7a-fd23ba9607b8
trentmkelly/LessWrong-43k
LessWrong
localroger: Replotting The Transmigration of Prime Intellect http://www.kuro5hin.org/story/2010/9/3/20910/05774
79559e0f-88c1-4924-a478-f7e3e1b4fc47
trentmkelly/LessWrong-43k
LessWrong
Troubles With CEV Part1 - CEV Sequence The CEV Sequence Summary: The CEV sequence consists of three posts tackling important aspects of CEV. It covers conceptual, practical and computational problems of CEV's current form. On What Selves Are draws on analytic philosophy methods in order to clarify the concept of Self,...
f976adcb-2e5c-4da6-bb60-765794f397bb
trentmkelly/LessWrong-43k
LessWrong
Contrary to List of Lethality's point 22, alignment's door number 2 Abstract: An alternative to the now-predominating models of alignment, corrigibility and "CEV", following a (jaunty, jaunty) critique of these. The critique to show, in substance: CEV and corrigibility have the exact same problems - in effect, they're...
02c9e871-41dd-4d9f-9b73-bac473f6fe40
trentmkelly/LessWrong-43k
LessWrong
How seriously should we take the hypothesis that LW is just wrong on how AI will impact the 21st century? This question is inspired by 1a3orn's comment on how there are troubling signs of epistemic issues in LW's Alignment field. I'll quote the comment here to tell you what I mean: > I think that the above is also a...
0e95cc7d-c24b-4ca0-9f2c-7de763c85c47
trentmkelly/LessWrong-43k
LessWrong
Towards White Box Deep Learning Hi, I’d like to share my paper that proposes a novel approach for building white box neural networks. The paper introduces semantic features as a general technique for controlled dimensionality reduction, somewhat reminiscent of Hinton’s capsules and the idea of “inverse rendering”. In...
b1fe3878-6c8e-460e-8182-1e949ec60151
trentmkelly/LessWrong-43k
LessWrong
Progress links and tweets, 2023-02-08 The Progress Forum * A catalog of big visions for biology * London progress meetup, Feb 25 * A Cure for My Cancer, by Virginia Postrel * Eli Dourado AMA has concluded Announcements * Metascience event at AEI, Feb 9 (via @AlecStapp) * Visa Limbo, a site from IFP tracking vi...
5b735275-e4b4-4a85-8375-b8857baa06a1
StampyAI/alignment-research-dataset/special_docs
Other
Knowledge and implicature: modeling language understanding as social cognition To what extent does language understanding rely on extra-linguistic knowledge and processes? One view of language processing suggests that it consists of largely separate, special-purpose faculties; another view suggests that it depends cri...
fc4561a6-a0ab-4bb4-931a-f87c40f109a5
trentmkelly/LessWrong-43k
LessWrong
A New Bayesian Decision Theory Two weeks ago I wrote an introduction to normative decision theory, yet I haven’t made clear what decision algorithm should be used. Before that, I’d like to give some preliminaries. Generally speaking, Newcomb's Problem and the Smoking Lesion Problem represent counterexamples to CDT an...
7f18f7a3-a69e-4354-aa49-20fb8fc3864a
StampyAI/alignment-research-dataset/arxiv
Arxiv
Improving Learning from Demonstrations by Learning from Experience I Introduction --------------- Reinforcement learning has been highly successful in many areas, meeting or exceeding human performance, in video game environments [[16](#bib.bib6 "Human-level control through deep reinforcement learning")], robot man...
6a40cd64-51e6-4466-9877-bfe7914764c3
StampyAI/alignment-research-dataset/lesswrong
LessWrong
Are AIs like Animals? Perspectives and Strategies from Biology *Disclaimer: The views expressed in this document are my own, and do not necessarily reflect those of my past or present employers.* In a recent [interview](https://www.commonwealthclub.org/events/archive/podcast/beyond-chatgpt-stuart-russell-risks-and-re...
83dd7055-adbe-4041-8590-5f3bd73810b3
trentmkelly/LessWrong-43k
LessWrong
A Block-Based Regularization Proposal for Neural Networks A Block-Based Regularization Proposal for Neural Networks Exploring Localized Weight Groupings as a Way to Control Overfitting Introduction I’m not an expert in machine learning. I've been studying the field out of curiosity and an almost irrational drive to ...
4f30b5e3-8999-40fa-aa44-b9380e08fd27
trentmkelly/LessWrong-43k
LessWrong
Amplify is hiring! Work with us to support field-building initiatives through digital marketing See the accompanying launch post here to read more about the project's background. TLDR: Amplify is an early-stage project supporting local Effective Altruist and AI Safety local groups as well as other field-building org...
2293506b-3b7d-4dba-83d4-e4fc673310b1
trentmkelly/LessWrong-43k
LessWrong
[LINK] Get paid to train your rationality A tournament is currently being initiated by the Intelligence Advanced Research Project Activity (IARPA) with the goal of improving forecasting methods for global events of national (US) interest. One of the teams (The Good Judgement Team) is recruiting volunteers to have thei...
cef3c348-9722-4434-83c9-7e47425585d1
StampyAI/alignment-research-dataset/eaforum
Effective Altruism Forum
UN Public Call for Nominations For High-level Advisory Body on Artificial Intelligence The upcoming formation of the [UN High-level Advisory Body on Artificial Intelligence](https://input.un.org/EFM/se/3995D1A465993A1B ) is an essential development in steering the future of AI.  Many members of the EA community have ...
e39997cf-ab4c-4763-b1b8-31fbfb6bfe14
trentmkelly/LessWrong-43k
LessWrong
The Mystery At The Heart of Central Banking     Economists love decentralized competitive markets. There’s nothing an economist likes to see more than some decentralized entrepreneurs starting businesses with well-defined property rights, competing with each other to serve the interests of others in order to serve...
51330a59-286e-4e3c-9753-3fc9b2ac257f
StampyAI/alignment-research-dataset/lesswrong
LessWrong
misc raw responses to a tract of Critical Rationalism *Written in response to* [*this David Deutsch presentation.*](https://www.youtube.com/watch?v=01C3a4fL1m0&list=WL&index=63&t=137s) *Hoping it will be comprehensible enough to the friend it was written for to be responded to, and maybe a few other people too.* Deut...