id
stringlengths
36
36
source
stringclasses
15 values
formatted_source
stringclasses
13 values
text
stringlengths
2
7.55M
38e53665-f3d6-468e-8f32-d1d99c89ae78
trentmkelly/LessWrong-43k
LessWrong
Asymptotic Logical Uncertainty: The Benford Test This is the second post in a series on Asymptotic Logical Uncertainty. All Turing Machines we discuss will have no input, and will output an infinite (or finite) sequence of bits. We let M(n) denote the nth bit output by the Turing machine M. Let A(n)=n↑nn be the nth ...
2b70a62e-3649-42fc-91be-102241fd1afb
trentmkelly/LessWrong-43k
LessWrong
Goal-thinking vs desire-thinking [Adapted from an old post on my personal blog] There's a lot of long-running arguments on the internet that basically consist of people arguing past each other due to differing basic assumptions that they don't know how to make explicit, preventing them from noticing the fundamental d...
9a68f4ff-33e9-43cd-a8b4-a74c429bd40d
trentmkelly/LessWrong-43k
LessWrong
Meetup: Philadelphia-- Taleb discussion When: Sunday, October 20, 1 PM Where: Nam Phuong at 11 St. and Washington St. Nam Phuong is a Vietnamese restaurant notable for good food, low prices, and very little background noise. Discussion topic: Taleb on how to think about small probabilities. Meetup discussion group.
2c908b6a-bcad-4310-81db-171de7b7abe6
trentmkelly/LessWrong-43k
LessWrong
Language Learning and the Dark Arts. I've been wanting to learn Mandarin Chinese for years now and just recently I wrote a small website to help me practise.1 All of the exercises are gap sentences that require you to type the correct answer before you can move on. I chose this kind of exercise because of the convinci...
e2b891a0-40d7-4691-917b-2f498dafd7b0
StampyAI/alignment-research-dataset/arxiv
Arxiv
Uni[MASK]: Unified Inference in Sequential Decision Problems. . 1 Introduction --------------- Masked language modeling (Devlin et al., [2019](#bib.bib11)) is a key technique in natural language processing (NLP). Under this paradigm, models are trained to predict randomly-masked subsets of tokens in a sequence. For...
f777f738-90e3-4f42-adc0-05ce3ae9e834
trentmkelly/LessWrong-43k
LessWrong
NYC Rationalist Community For those who don't yet know, there has been a thriving rationalist community in NYC since April 2009.  We've been holding weekly meetups for the past several months now, and often have game nights, focused discussions, etc.  For those of you who live in the area, and not yet involved, I high...
ba701702-b9ee-4c23-8234-0d0b27bf0090
trentmkelly/LessWrong-43k
LessWrong
[SEQ RERUN] Incremental Progress and the Valley Today's post, was originally published on 04 April 2009. A summary (taken from the LW wiki):   > The optimality theorems for probability theory and decision theory, are for perfect probability theory and decision theory. There is no theorem that incremental changes tow...
f6266bdb-2a75-4a61-9e7c-16955c2cd04c
trentmkelly/LessWrong-43k
LessWrong
Cassette Tape Thoughts > “The packaging of intellectual positions and views is one of the most active enterprises of some of the best minds of our day. The viewer of television, the listener to radio, the reader of magazines, is presented with a whole complex of elements-all the way from ingenious rhetoric to carefull...
579e0b38-9677-4762-8e3f-4b44fdbf0a10
trentmkelly/LessWrong-43k
LessWrong
Best *organization* red-pill books and posts? Most books about orgs are written for people in an org who need to cope. (Art of Possibility by the Zanders is a genuinely great one in this genre.) I am currently orgless and don't need to cope with anything orgwise. Hence, I would love to read some brutal insightful take...
7f977d65-45c6-4805-9a36-cd950424a0cf
trentmkelly/LessWrong-43k
LessWrong
Scoring forecasts from the 2016 “Expert Survey on Progress in AI” Summary This document looks at the predictions made by AI experts in The 2016 Expert Survey on Progress in AI, analyses the predictions on ‘Narrow tasks’, and gives a Brier score to the median of the experts’ predictions.  My analysis suggests that the...
1df8c866-5197-4719-b2d7-031c7ba3f850
trentmkelly/LessWrong-43k
LessWrong
An accidental experiment in location memory I bought a plastic mat to put underneath my desk chair, to protect the wooden floor from having bits of stone ground into it by the chair wheels. But it kept sliding when I stepped onto it, nearly sending me stumbling into my large, expensive, and fragile monitor. I decided ...
726592c5-71fe-421f-a9f5-fcf7279bbde1
trentmkelly/LessWrong-43k
LessWrong
Amplifying the Computational No-Coincidence Conjecture Introduction Recently, the Computational No-Coincidende Conjecture[1] was proposed, presented as an assumption that might be needed to develop methods to explain neural nets in the current approach from the Alignment Research Center. In this post, I will prove tha...
1c488cf9-0cb5-45b2-9af1-fa4b6e905b17
StampyAI/alignment-research-dataset/alignmentforum
Alignment Forum
More Recent Progress in the Theory of Neural Networks *Thanks to Dan Roberts and Sho Yaida for comments on a draft of this post.* In this post, I would like to draw attention to the book [Principles of Deep Learning Theory](https://deeplearningtheory.com/) (PDLT), which I think represents a significant advance in our...
d5dee5db-2b1d-4388-9a62-b7d6c8bcef07
trentmkelly/LessWrong-43k
LessWrong
Is there a 'time series forecasting' equivalent of AIXI? In the way that AIXI is an abstracted mathematical formalism for (very roughly) "a program that maximizes the expected total rewards received from the environment", what is the equivalent formalism for an abstracted next token predictor? Does this exist in the ...
53d90308-0cc3-4d04-ba1b-170cf6ecb157
trentmkelly/LessWrong-43k
LessWrong
In-depth description of a quite strict, quite successful government program against teen substance abuse, spreading from Iceland
bc62fe8b-3537-4e1e-9be6-0d9ef199ef44
trentmkelly/LessWrong-43k
LessWrong
Review and Summary of 'Moral Uncertainty' I just read the recently published 'Moral Uncertainty' by Will MacAskill, Krister Bykvist, and Toby Ord. It's an excellent book with significant practical implications. As such, I originally intended to write a brief review — which ballooned into this monstrosity. In any case,...
c9e421d7-e1be-4bab-971e-1af7cc877629
trentmkelly/LessWrong-43k
LessWrong
A Small Vacation (note: I have been thinking about this essay for several months; I did not realize it would turn out to be so timely). Many people in the rationality community and adjacent corners of the intellectual world are sympathetic to the idea of Open Borders. This concept can be explained easily. People in t...
868e1ed8-36de-4a95-97b8-0ad9a46d9afe
trentmkelly/LessWrong-43k
LessWrong
Reflections on Connect Developers I was working on creating a projects page for my personal website, https://adamzerner.bearblog.dev/. It hit me how many of my projects have been failures, and that I should reflect on them to see what I can learn. Perhaps I'll reflect on other ones in a future blog post, but here I wa...
5da77f13-c8ec-4ab0-8584-13513fba017b
trentmkelly/LessWrong-43k
LessWrong
Is Greed Stupid? I just finished reading a fantastic Wait But Why post: How Tesla Will Change The World. One of the things that was noted is that the people in the Auto and Oil industries are trying to delay the introduction of Electric Vehicles (EVs) so they could make more money. The post also explains how importan...
f22befd1-5d3e-43fb-9b2f-05adb0d28f6c
trentmkelly/LessWrong-43k
LessWrong
Is the ChatGPT-simulated Linux virtual machine real? Context: Building A Virtual Machine inside ChatGPT It very much triggers my "roll to disbelieve" reflex, but people seem to be talking about it as if it's real and not that surprising. I'd try it myself, but OpenAI is saying ChatGPT is overloaded (and I don't have ...
003275ed-2417-426e-958c-ad86d8b17bda
trentmkelly/LessWrong-43k
LessWrong
How can a layman contribute to AI Alignment efforts, given shorter timeline/doomier scenarios? Versions of this question have probably been asked before, but MIRI at least is getting more pessimistic and I’ve seen Eliezer express multiple times that he doesn’t know what actually useful advice to give people who aren’t...
c79b819a-f611-490f-9862-3e6226a87696
trentmkelly/LessWrong-43k
LessWrong
LDL 2: Nonconvex Optimization Edit 10/23: by request I'm going back and putting the images in. It's a pain that you can't place images in line! Also I edited the dumb probability calculation to make more sense since I may not be able to write another post today. My favorite “party theorem” is Gauss-Bonnet. It’s the t...
0d873ff6-ef0c-4ab8-a349-101a9cc7da89
trentmkelly/LessWrong-43k
LessWrong
Advocating for factual advocacy In this post I present a Hansonian view of morality, and tease out its consequences for advocacy. Beware, all of this is the result of armchair speculation. The simplified model goes something like this: * Human morality is mostly a justification mechanism, trying to give coherence ...
c690001d-4441-4e5a-b115-c9822fbb45a0
StampyAI/alignment-research-dataset/alignmentforum
Alignment Forum
AXRP Episode 7 - Side Effects with Victoria Krakovna [Google Podcasts link](https://podcasts.google.com/feed/aHR0cHM6Ly9heHJwb2RjYXN0LmxpYnN5bi5jb20vcnNz/episode/ZTQ1NjdlMWEtMmQzNC00Y2FlLThlZDMtYzgwZWJmNzRjMWFj) This podcast is called AXRP, pronounced axe-urp and short for the AI X-risk Research Podcast. Here, I ([D...
4b3a34bb-ceff-448d-8f33-80dd414d84f8
trentmkelly/LessWrong-43k
LessWrong
Prize for Alignment Research Tasks Can AI systems substantially help with alignment research before transformative AI? People disagree. Ought is collecting a dataset of alignment research tasks so that we can: 1. Make progress on the disagreement 2. Guide AI research towards helping with alignment We’re offering ...
969e58f3-0254-4fa2-a66c-b13e0b3b7d54
StampyAI/alignment-research-dataset/arxiv
Arxiv
Human Instruction-Following with Deep Reinforcement Learning via Transfer-Learning from Text 1 Introduction --------------- Developing machines that can follow natural human commands, particularly those pertaining to an environment shared by both machine and human, is a long-standing and elusive goal of AI (Winogra...
84512601-7a36-4fe2-b2bb-4a6340659cbc
trentmkelly/LessWrong-43k
LessWrong
Repugnant levels of violins This is a rambling post about the repugnant conclusion, mass-manufacturing, metis and woo, and the fragility of expert discernment — through the lens of my experiences playing, repairing, and selling violins. (And eating discount sushi.) Epistemic status: Discussing utilitarianism-related ...
22f8f033-2bff-4592-81b1-de16e480cf43
StampyAI/alignment-research-dataset/aisafety.info
AI Safety Info
What is the "control problem"? The "control problem" is the problem of preventing artificial superintelligence (ASI) from harming humanity. How do we keep a more intelligent being under control, or how do we align it with our values? If we succeed in solving this problem, intelligence vastly superior to ours can take ...
6d7607f2-14b1-4ea2-8532-a6bcf7e51967
trentmkelly/LessWrong-43k
LessWrong
Crosspost: Developing the middle ground on polarized topics Crossposted from Otherwise   I was once in a group discussion about whether wild animals might be having net negative lives. One person didn’t want to consider that possibility, essentially because “then people would want to kill all the wild animals.” Hol...
a6b19513-96a8-4211-a662-8c591c322639
trentmkelly/LessWrong-43k
LessWrong
[link] SMBC on utilitarianism and vegatarianism. http://www.smbc-comics.com/index.php?db=comics&id=2393#comic
b7bafa09-24c5-4c55-bf34-8730a5117336
trentmkelly/LessWrong-43k
LessWrong
Robin Hanson's Grabby Aliens model explained - part 1 This article is the script of the video linked above. It differs slightly from the video due to some phrasing changes in the narration  [now it's exactly the narration]. Here, I aim to explain the rationale for Robin Hanson's Grabby Aliens model. In short, humanity...
2a24152d-89c2-4eab-a52d-abece88b5e77
trentmkelly/LessWrong-43k
LessWrong
Automated Researchers Can Subtly Sandbag tl;dr When prompted, current models can sandbag ML experiments and research decisions without being detected by zero-shot prompted monitors. Claude 3.5 Sonnet (new) can only sandbag effectively when seeing a one-shot example, while Claude 3.7 Sonnet can do this without an examp...
b08489ed-599e-4bee-bca4-3abea32c22a5
trentmkelly/LessWrong-43k
LessWrong
newcomb's altruism Hello lesswrong community. I want to play a game. *jigsaw music plays* In box B, I have placed either $1,000,000 multiplied by the number of humans on earth or $0. In box A I placed $1000 multiplied by the number of humans on earth. Every human on the planet, including you, will now be asked the fo...
67f2ff5e-4e41-4230-b68a-96d19a16b658
trentmkelly/LessWrong-43k
LessWrong
An Exercise to Build Intuitions on AGI Risk Epistemic status: confident that the underlying idea is useful; less confident about the details, though they're straightforward enough that I expect they're mostly in the right direction. TLDR: This post describes a pre-mortem-like exercise that I find useful for thinking ...
b027bfc4-dae2-4027-aa3c-8b7d7698277e
StampyAI/alignment-research-dataset/blogs
Blogs
Five Theses, Using Only Simple Words ![xkcd at desk](https://intelligence.org/wp-content/uploads/2013/09/xkcd-at-desk.png)A [recent xkcd comic](http://xkcd.com/1133/) described the Saturn V rocket using only the 1000 most frequently used words (in English). The rocket was called “up-goer five,” and the liquid hydrogen...
0d01a81e-0a75-40ac-a78f-ee407bc0fc37
trentmkelly/LessWrong-43k
LessWrong
Decision Theory (A longer text-based version of this post is also available on MIRI's blog here, and the bibliography for the whole sequence can be found here.)
b9a6d658-1042-4700-b8cb-2d3b7fd2a7e7
trentmkelly/LessWrong-43k
LessWrong
Careless talk on US-China AI competition? (and criticism of CAIS coverage) > China has made several efforts to preserve their chip access, including smuggling, buying chips that are just under the legal limit of performance, and investing in their domestic chip industry.[1] Sounds about right? hotpot.ai/art-generato...
f6deb983-0bda-4534-b748-1e15b9b825c9
StampyAI/alignment-research-dataset/blogs
Blogs
the systematic absence of libertarian thought the systematic absence of libertarian thought --------------------------------------------- to me, [libertarianism](https://en.wikipedia.org/wiki/Libertarianism) aims to give people [freedom](defining-freedom.html), to [give them self-determination over their life](core-...
db64b78a-0112-4f64-abb4-19de7d10b322
StampyAI/alignment-research-dataset/lesswrong
LessWrong
CEV-tropes As seen in other [threads](/r/discussion/lw/l0v/cev_coherence_versus_extrapolation/), people disagree on whether CEV exists, and if it does, what it might turn out to be.   It would be nice to try to categorize common speculations about CEV. 1a. CEV doesn't exist, because human preferences are too [di...
91dcf845-859c-437c-88a8-67429f5daa58
trentmkelly/LessWrong-43k
LessWrong
Metaculus Announces Forecasting Tournament to Evaluate Focused Research Organizations, in Partnership With the Federation of American Scientists You are invited to participate in Metaculus's FRO-Casting Tournament, an exciting pilot project in partnership with The Federation of American Scientists that harnesses pred...
4a681a92-4abb-4c33-bb41-1ce3943aca9c
trentmkelly/LessWrong-43k
LessWrong
[Link] Study: no big filter, we're just too early "Earth is one of the first habitable planets to form - and we're probably too early to the party to get a chance to meet future alien civilisations."   http://www.sciencealert.com/earth-was-one-of-the-first-habitable-planets-in-the-universe-and-most-are-yet-to-be-bor...
b301d515-fc38-4bfd-a1db-0a202bb354f8
trentmkelly/LessWrong-43k
LessWrong
"Alignment at Large": Bending the Arc of History Towards Life-Affirming Futures What if we expanded our lens beyond AI alignment, to grapple with the full scope of what "alignment" really means? We're facing not just the challenge of aligning artificial intelligence, but the deeper question of how to align technologic...
f533f99b-05e7-416a-b3c0-f8529e26f837
StampyAI/alignment-research-dataset/arxiv
Arxiv
On the Equilibrium Elicitation of Markov Games Through Information Design 1 Introduction --------------- Building rational multi-agent system is an important research desideratum in Artificial Intelligence. In goal-directed decision making systems, an agent’s action is controlled by its consequence [[1](#bib.bib1)]...
2860aeb0-ea8d-4da8-b5dd-519771bba194
StampyAI/alignment-research-dataset/blogs
Blogs
To contribute to AI safety, consider doing AI research Among those concerned about risks from advanced AI, I’ve encountered people who would be interested in a career in AI research, but are worried that doing so would speed up AI capability relative to safety. I think it is a mistake for AI safety proponents to avoid...
7096de63-bd37-4fde-a524-227e66f38614
StampyAI/alignment-research-dataset/blogs
Blogs
Automation of music production Most machine learning researchers expect machines will be able to create top quality music by 2036. Details ------- ### Evidence from survey data In the [2016 ESPAI](http://aiimpacts.org/2016-expert-survey-on-progress-in-ai/), participants were asked two relevant questions: [**To...
169674f7-b6d6-45c4-abf8-7d8deef99bc1
trentmkelly/LessWrong-43k
LessWrong
Are long-term investments a good way to help the future? [Epistemic status: I am very confused, possibly about very basic Econ 101 things. Probably there is more research I could do to resolve my confusion but I don't have the time or motivation for that right now.] [Update: Paul Christiano's comment resolved a reas...
fcec7881-e4b0-4160-82e8-5663bc75d1ff
trentmkelly/LessWrong-43k
LessWrong
The Aspiring Rationalist Congregation 1. Meta Note: This post has been languishing in a Google doc for many months as I've procrastinated on cleaning it up to be more coherent and polished. So... I'm posting it as is, with very little cleanup, in the hopes that it's valuable in the current state. I'm sure there are b...
32c42a71-4118-41fd-bee7-f78ba4a5224a
trentmkelly/LessWrong-43k
LessWrong
Proof Section to an Introduction to Reinforcement Learning for Understanding Infra-Bayesianism Introduction This post accompanies An Introduction to Reinforcement Learning for Understanding Infra-Bayesianism. The goal of this introduction is to provide a high-level overview of the proofs contained in this post. The p...
311c3fc3-4ec3-4b99-b575-c6b599b9acc4
trentmkelly/LessWrong-43k
LessWrong
The mystery of pain and pleasure   Some arrangements of particles feel better than others. Why? We have no general theories, only descriptive observations within the context of the vertebrate brain, about what produces pain and pleasure. It seems like there's a mystery here, a general principle to uncover. Let's tr...
8f5a10b5-8883-4e14-a00f-c93f3e5f666a
LDJnr/LessWrong-Amplify-Instruct
LessWrong
"Where the mind cuts against reality's grain, it generates wrong questions—questions that cannot possibly be answered on their own terms, but only dissolved by understanding the cognitive algorithm that generates the perception of a question. One good cue that you're dealing with a "wrong question" is when you cannot e...
147ebdfe-12ca-4e9f-95ec-5ea007a16961
trentmkelly/LessWrong-43k
LessWrong
Biological humans and the rising tide of AI The Hanson-Yudkowsky AI-Foom Debate focused on whether AI progress is winner-take-all. But even if it isn't, humans might still fare badly. Suppose Robin is right. Instead of one basement project going foom, AI progresses slowly as many organizations share ideas with each o...
f8be3ec1-5537-4cb9-945f-bc89b578ee32
trentmkelly/LessWrong-43k
LessWrong
Conversation Visualizer Conversation Visualizer Voice to text to image to reflection. ---------------------------------------- Contents * Background * Construction * Cost * Examples ---------------------------------------- Background niplav and I were discussing what makes good living room decor. It must be ...
0554e45c-7c81-4376-ad2f-393c87a042e5
StampyAI/alignment-research-dataset/blogs
Blogs
Holden Karnofsky on Transparent Research Analyses ![](http://intelligence.org/wp-content/uploads/2013/08/holden_w150.jpg)Holden Karnofsky is the co-founder of [GiveWell](http://www.givewell.org), which finds outstanding giving opportunities and publishes the full details of its analysis to help donors decide where to ...
daa4c018-1dff-4df8-afd4-d2f396f92494
trentmkelly/LessWrong-43k
LessWrong
How much should I update on the fact that my dentist is named Dennis? According to some psychology research (eg Pelham, Mirenberg, and Jones, people are slightly more likely to choose occupations that sound like their names: "people named Dennis or Denise are overrepresented among dentists". Other research (eg Simonso...
88bd1aca-d1da-4816-8047-2839f2274c61
StampyAI/alignment-research-dataset/arxiv
Arxiv
Learning to Be Cautious 1 Introduction --------------- One challenge in the field of artificial intelligence (AI) is to design agents that avoid doing harm or being destructive. This is of particular concern in complex, dynamic environments that change over time, where exhaustive training and testing are infeasible...
40d85dca-7fde-4776-a953-2cb82eedbf06
trentmkelly/LessWrong-43k
LessWrong
Reality is Fractal-Shaped Tl;dr: Much of common sense and personal insights are just first order approximations, broadly generalizing complex situations to a single supposed truth. But often the truth value of a claim depends on many hidden factors. In this post I argue that it’s important to maintain a general awaren...
9900bdbd-6d9a-49ea-8ba4-9dd85eaee110
StampyAI/alignment-research-dataset/lesswrong
LessWrong
The Genie in the Bottle: An Introduction to AI Alignment and Risk *I'm writing an article about the alignment problem and AI risk, and I'm sharing a draft with the fine folks here on LessWrong to elicit some feedback. I suspect most frequent readers here will already be familiar with the contents and arguments presen...
dc36ad3c-7dfe-435f-a509-3ce0796d50ae
trentmkelly/LessWrong-43k
LessWrong
A cheating approach to the tiling agents problem (This post resulted from a conversation with Wei Dai.) Formalizing the tiling agents problem is very delicate. In this post I'll show a toy problem and a solution to it, which arguably meets all the desiderata, but only by cheating in a new and unusual way. Here's a s...
af3dc54e-e618-47ff-afae-5ad096333b42
StampyAI/alignment-research-dataset/lesswrong
LessWrong
Fixed points and free will In his [most recent appearance](https://80000hours.org/podcast/episodes/bryan-caplan-parenting-workers-betting/) on the 80,000 Hours podcast, [Bryan Caplan](https://en.wikipedia.org/wiki/Bryan_Caplan) gave the following argument in favor of libertarian free will: > > One that I’m very te...
3d355a6d-6593-4ea5-8526-84a219cd2354
trentmkelly/LessWrong-43k
LessWrong
A Muggle Studies course At my local Harry Potter fanclub (Bogotá, Colombia) some members teach "classes" on subjects they're passionate about. We've had informal courses on history, creative writing, English, etc. But recently some other classes have appeared that have made me worry seriously: astrology, divination, a...
6d27201c-5b66-4aa8-8edb-8ad0f249ea4c
trentmkelly/LessWrong-43k
LessWrong
How to get people to produce more great exposition? Some strategies and their assumptions Some of the recent posts in the distillation & pedagogy tag here on LessWrong, such as "Call For Distillers" and the AI Safety Distillation Contest, have been bugging me, and this post is my attempt to introspect about why. Here...
e76f41ed-ec78-417f-9f44-adf2b0e10de2
trentmkelly/LessWrong-43k
LessWrong
I Converted Book I of The Sequences Into A Zoomer-Readable Format If I (a 19 year old male) texted "www.readthesequences.com" to my roommate, the probable outcome is that he would skim the site for under a minute, text back something like "seems interesting, I'll def check it out sometime", and then proceed to never r...
2b376b59-3a38-40aa-81e5-d4db5cca809d
trentmkelly/LessWrong-43k
LessWrong
Simulating Problems Apologies for the rather mathematical nature of this post, but it seems to have some implications for topics relevant to LW. Prior to posting I looked for literature on this but was unable to find any; pointers would be appreciated. In short, my question is: How can we prove that any simulation of...
72e45fbb-2cff-4251-b158-cfc380a42f5a
StampyAI/alignment-research-dataset/lesswrong
LessWrong
The Strangest Thing An AI Could Tell You Human beings are all crazy.  And if you tap on our brains just a little, we get so crazy that even other humans notice.  [Anosognosics](/lw/20/the_apologist_and_the_revolutionary/) are one of my favorite examples of this; people with right-hemisphere damage whose left arms beco...
3e0e7d6a-8cfa-43bb-b9c9-b2fc5e601adc
StampyAI/alignment-research-dataset/blogs
Blogs
Max Tegmark on the mathematical universe ![Randal A. Koene portrait](http://intelligence.org/wp-content/uploads/2014/03/Tegmark_2.jpg)Known as “Mad Max” for his unorthodox ideas and passion for adventure, his scientific interests range from precision cosmology to the ultimate nature of reality, all explored in his new...
c5a29085-1b4f-4c6e-9206-a0b6e4455ec2
StampyAI/alignment-research-dataset/eaforum
Effective Altruism Forum
Please, someone make a dataset of supposed cases of "tech panic" TL;DR: Someone should probably [write a grant to] produce a spreadsheet/dataset of past instances where people claimed a new technology would lead to societal catastrophe, but with informative variables such as “multiple people working on the tech believ...
69563ef7-d584-470b-b0c2-e9dad60d06fc
trentmkelly/LessWrong-43k
LessWrong
4. Why existing approaches to cause prioritization are not robust to unawareness We’re finally ready to see why unawareness so deeply undermines action guidance from impartial altruism. Let’s recollect the story thus far: 1. First: Under unawareness, “just take the expected value” is unmotivated. 2. Second: Likewis...
13a36c0c-b5cd-4b3f-b0a9-c213e8759383
trentmkelly/LessWrong-43k
LessWrong
Meetup : Washington, D.C.: [Postponed for snow] Discussion article for the meetup : Washington, D.C.: [Postponed for snow] WHEN: 24 January 2016 03:00:00PM (-0500) WHERE: Reynolds Center Meetup has been postponed due to heavy snow expected this Friday shutting down public transit for the weekend. This is a meetup ...
e1b7a192-a33c-49d4-b57e-cf97e78f7589
trentmkelly/LessWrong-43k
LessWrong
Meetup : Rationality Meetup Vienna Discussion article for the meetup : Rationality Meetup Vienna WHEN: 13 February 2016 03:00:00PM (+0100) WHERE: Kaisermühlenstraße, Wien Location: (often depicted wrong, come to U2 Stadlau) See picture below! Event on Facebook: https://www.facebook.com/events/1227373493945585/ (jo...
003e21b3-9ddc-4070-8c22-8f45272ea1d4
trentmkelly/LessWrong-43k
LessWrong
LessWrong opinion of Nietzsche? So, at the risk of starting controversy, I'm not exactly sure what the policy is about asking questions on philosophy.. But would you mind giving your opinion on Nietzsche? I just bought Human, All Too Human. It's a tough read for me, and I'm sort of plowing through it, though it's int...
7e5031b6-1ba8-4994-b21c-954e22b265ff
trentmkelly/LessWrong-43k
LessWrong
Percent reduction of gun-related deaths by color of gun. A thought experiment: Assume all guns were pink by law tomorrow. Would that have an impact on the number of gun-related deaths? What percentage? Would it have an impact on the number of mass shootings? This is not an actual policy suggestion, so the feasibili...
c1f0ee20-fd52-4030-800f-c8333b2661be
trentmkelly/LessWrong-43k
LessWrong
Recent site changes, Mon 4th July We've pushed out some new changes, including a front page. Promoted articles are now available at http://lesswrong.com/promoted/ and from the MAIN link in the top nav. Most of the recent changes were requested here or here. This is the place to comment on the new changes… but please...
42e058c2-e6b4-4d28-aa1b-12bd6e40bfe9
trentmkelly/LessWrong-43k
LessWrong
Forcing Freedom Wretched Worlds You are a freshly recruited cadet in the Intergalactic Liberation Corps (ILC). With a handful of valiant companions, you will travel across the galaxy, helping to free oppressed Beta Species from the claws of their Alpha masters, and fighting slavery whenever it dares to rear its ugly h...
54e16310-1929-42a3-a29b-d5e650957109
trentmkelly/LessWrong-43k
LessWrong
If someone with sufficient capabilities decides to deliberately create an Unaligned AGI, is there anything anyone can actually DO, to stop them? The scenario is simple, some unassuming programmer creates a DAO on a Blockchain that is the seed AI, with the single purpose gaining political, economic, and military power ...
066754d7-842b-4f72-8698-6a45ed23f77d
trentmkelly/LessWrong-43k
LessWrong
The U.S. is becoming less stable We focus so much on arguing over who is at fault in this country that I think sometimes we fail to alert on what's actually happening. I would just like to point out, without attempting to assign blame, that American political institutions appear to be losing common knowledge of their ...
00ac9014-3b15-4759-ac29-4c4cec72079b
StampyAI/alignment-research-dataset/lesswrong
LessWrong
Solving Mysteries - Somewhat related? I've been wondering about whether or not GPT-4 can solve fictional mysteries (or provide a best guess). But I guess I'm also interested to see if fictional mysteries are actually typically written so that they'd be solvable, and if so, I'm assuming the pattern of assigning blame i...
a8585ec3-688b-421b-8532-aeb948ed4c15
trentmkelly/LessWrong-43k
LessWrong
Reality versus Human Expectations Given two options: 1. Alter reality so it more closely aligns with human expectations. 2. Alter humans so their expectations more closely align with reality. If it cost the same to implement either option, which would you prefer and why?
8f11b671-9302-48e1-b221-e2719614757d
trentmkelly/LessWrong-43k
LessWrong
Basic model of Sending a Message (Communication 101) My hope in writing this out is that you go "well duh! Of course" and then pretend like you knew this all already. Whether or not you did is up to you. ---------------------------------------- With communication there is going to be a Sender and a Receiver. These c...
e6f564c1-21eb-4ff6-88ab-60c1e0f0940b
trentmkelly/LessWrong-43k
LessWrong
We Change Our Minds Less Often Than We Think > Over the past few years, we have discreetly approached colleagues faced with a choice between job offers, and asked them to estimate the probability that they will choose one job over another. The average confidence in the predicted choice was a modest 66%, but only 1 of ...
6ce23b6c-ebd2-4702-8370-bb741ee9e57f
StampyAI/alignment-research-dataset/lesswrong
LessWrong
Is keeping AI "in the box" during training enough? Many current ML architectures do offline training on a fixed training set ahead of time. For example, GPT-3 is quite successful with this approach. These are, so-to-speak, "in the box" during training: all they can do is match the given text completion more or less we...
07d72e48-4215-4f1f-93d8-78e08eb11066
trentmkelly/LessWrong-43k
LessWrong
AI Safety in China: Part 2 I've given things a lot more thought and wanted to make an edited summary of my views regarding AI Safety in China. 1. AI Safety is almost nonexistent in China. The only toe-hold it has is rooted in the expat EA community, which is almost exclusively doing field building. This hasn't been ...
8d64a002-d579-4136-8fb0-01bf9947186c
trentmkelly/LessWrong-43k
LessWrong
Alignment Newsletter #49 Find all Alignment Newsletter resources here. In particular, you can sign up, or look through this spreadsheet of all summaries that have ever been in the newsletter. Highlights Exploring Neural Networks with Activation Atlases (Shan Carter et al): Previous work by this group of people inclu...
42356087-bd70-4c3d-a6ee-aac4ba780cd0
trentmkelly/LessWrong-43k
LessWrong
Forecast 2025 With Vox's Future Perfect Team — $2,500 Prize Pool Vox's Future Perfect highlights long-term thinking and evidence-based ways to improve human and animal welfare. Each year, Future Perfect writers like Dylan Matthews, Kelsey Piper, and Sigal Samuel forecast the year to come and evaluate their previous pr...
3364b982-ee65-4521-8df0-d2f2fbbf2cd3
trentmkelly/LessWrong-43k
LessWrong
A ghost (Cross-posted from Hands and Cities) This post describes a type of thought experiment I sometimes perform in thinking about what to do. I find it a helpful tool for stepping back from what’s immediately salient to me. It’s mostly just a somewhat hokey variant on a classic type of thought, and I’m not sure how...
120e0712-a982-4d0c-8df3-c605164f8ce4
StampyAI/alignment-research-dataset/lesswrong
LessWrong
Timeless Decision Theory: Problems I Can't Solve Suppose you're out in the desert, running out of water, and soon to die - when someone in a motor vehicle drives up next to you.  Furthermore, the driver of the motor vehicle is a perfectly selfish ideal game-theoretic agent, and even further, so are you; and what's mor...
e4c63b11-e69a-48a8-bd58-a30bb092ff89
StampyAI/alignment-research-dataset/alignmentforum
Alignment Forum
ELK prize results From January - February the Alignment Research Center [offered prizes](https://www.lesswrong.com/posts/QEYWkRoCn4fZxXQAY/prizes-for-elk-proposals) for proposed algorithms for [eliciting latent knowledge](https://www.lesswrong.com/posts/qHCDysDnvhteW7kRd/arc-s-first-technical-report-eliciting-latent-k...
3902396a-c796-4d39-a74f-8674c4f69381
StampyAI/alignment-research-dataset/eaforum
Effective Altruism Forum
Jade Leung: Why companies should be leading on AI governance --- *Are companies better-suited than governments to solve collective action problems around artificial intelligence? Do they have the right incentives to do so in a prosocial way? In this talk, Jade Leung argues that the answer to both questions is "yes".*...
84f37f09-aa24-4f87-895f-7d0a83e4d071
StampyAI/alignment-research-dataset/alignmentforum
Alignment Forum
Empirical Observations of Objective Robustness Failures [Inner alignment](https://arxiv.org/abs/1906.01820) and [objective](https://www.alignmentforum.org/posts/2mhFMgtAjFJesaSYR/2-d-robustness) [robustness](https://www.alignmentforum.org/posts/SzecSPYxqRa5GCaSF/clarifying-inner-alignment-terminology) have been freque...
fbc1cfa0-d5be-4b42-a8d9-954da7407f56
trentmkelly/LessWrong-43k
LessWrong
[SEQ RERUN] Beyond the Reach of God Today's post, Beyond the Reach of God was originally published on 04 October 2008. A summary (taken from the LW wiki):   > Compare the world in which there is a God, who will intervene at some threshold, against a world in which everything happens as a result of physical laws. Whi...
fdcadd5f-284b-43f4-b2cf-f4a7368eee10
trentmkelly/LessWrong-43k
LessWrong
Book Review: Open Borders Edit: I now have a comment response post for this post on my personal website.  This is a (significantly updated) cross-post from my personal website. All sources are from the book unless otherwise mentioned. Bryan Caplan is an economist at George Mason University known for his wacky libert...
c59df8b5-9d5c-4a1b-8b65-419f9958e0ba
trentmkelly/LessWrong-43k
LessWrong
Framing Practicum: Turnover Time This is a framing practicum post. We’ll talk about what turnover time is, how to recognize when turnover time is relevant in the wild, and what questions to ask when you find it. Then, we’ll have a challenge to apply the idea. Today’s challenge: come up with 3 examples of approximatel...
a722e950-512c-4e9f-ad30-9fbc901afb22
trentmkelly/LessWrong-43k
LessWrong
Thoughts on 'List of Lethalities' I read through Eliezer’s “AGI Ruin: A List of Lethalities” post, and wrote down my reactions* as I went. I tried to track my internal responses, *before* adjusting for my perception of the overall thoughts of the field. In order to get this written and posted, I aimed to write the re...
457b2f54-b396-4e37-aad5-73820465f8b4
trentmkelly/LessWrong-43k
LessWrong
Using GPT-3 for preventing conflict during messaging — a pitch for an app I'm going to pitch an idea for an app, which I would love to see happen, but don't have time to build myself. My hope is that other people will steal and improve this idea because I believe it has the potential to help a lot of people including ...
d0ed8e27-5a31-475f-bca3-f83a38919ebb
trentmkelly/LessWrong-43k
LessWrong
What Bayesianism taught me David Chapman criticizes "pop Bayesianism" as just common-sense rationality dressed up as intimidating math[1]: > Bayesianism boils down to “don’t be so sure of your beliefs; be less sure when you see contradictory evidence.” > Now that is just common sense. Why does anyone need to be told ...
731aeb68-fbdc-423a-9baa-f87aeac2fc0f
trentmkelly/LessWrong-43k
LessWrong
Singularity Summit 2010 Less Wrong Meetup It's been a whole year (almost) and the Singularity Summit is upon us again. This time, instead of having a Meetup for only a few hours, we're going to have a continuous meetup lasting the entire length of the Summit. We will have a room adjacent to the lobby with a sign procl...
66e84331-fdf7-43e1-88d5-9e39afe998e3
trentmkelly/LessWrong-43k
LessWrong
Ranking Cultures by Rationality I have been interested in different cultures for a while, probably because I find so many deficiencies in my own culture. Which countries/cultures do you find to be generally the most rationally based? Japan and Sweden especially seem to have quite a secular populace mostly immune to si...
343014ba-e406-4d2c-ad4b-1e8b8837d0cd
trentmkelly/LessWrong-43k
LessWrong
[Link] The Typical Mind Fallacy, Illustrated If you've ever wanted a quick way to explain the Typical Mind Fallacy, this cartoon by Winston Rowntree will do the trick nicely. (Panel 3 seems of particular relevance to LessWrongians, and the last two panels also remind me of Hufflepuff/Slytherin themes in HP:MOR.) Enj...
670021e4-b623-449c-b98a-3d5b4ec277ce
StampyAI/alignment-research-dataset/arbital
Arbital
Data capacity The data capacity of an object is defined to be the [https://arbital.com/p/-3nd](https://arbital.com/p/-3nd) of the number of different distinguishable states the object can be placed into. For example, a coin that can be placed heads or tails has a data capacity of $\log(2)$ units of [https://arbital.co...
8c009c76-3ded-43d0-b34e-43d57c30145e
trentmkelly/LessWrong-43k
LessWrong
Attempting to Deconstruct "Real" I see some current contentions which have a relatively invisible crux with respect to the concept of real. I'm increasingly doubtful that the concept means very much. I think ‘real’ or ‘reality’ is probably a mostly needless reification over several different and not fully compatible d...
9a62dc20-0635-413a-bc3e-02d6ee48dea8
trentmkelly/LessWrong-43k
LessWrong
Eight claims about multi-agent AGI safety There are quite a few arguments about how interactions between multiple AGIs affect risks from AGI development. I’ve identified at least eight distinct but closely-related claims which it seems worthwhile to disambiguate. I’ve split them up into four claims about the process o...