id stringlengths 36 36 | source stringclasses 15
values | formatted_source stringclasses 13
values | text stringlengths 2 7.55M |
|---|---|---|---|
b6f58fb0-93d0-4c4d-b5b1-7529dbd1d6fc | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Shard Theory: An Overview
*Generated as part of SERI MATS, Team Shard's research, under John Wentworth.*
*Many thanks to Quintin Pope, Alex Turner, Charles Foster, Steve Byrnes, and Logan Smith for feedback, and to everyone else I've discussed this with recently! All mistakes are my own.*
 at the [Human Level AI multi-conference](https://www.hlai-conf.org/) held between the AGI, BICA, the NeSY conferences. It also featured the Future of AI track... |
db2471d6-d1d2-497b-82f4-59eb4c9eae16 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Giant (In)scrutable Matrices: (Maybe) the Best of All Possible Worlds
It has become common on LW to refer to "giant inscrutable matrices" as a problem with modern deep-learning systems.
To clarify: deep learning models are trained by creating giant blocks of random numbers -- blocks with dimensions like 4096 x 512 x... |
4191d26d-d2b6-413d-b5f3-864ab10dea6e | trentmkelly/LessWrong-43k | LessWrong | LW Team Updates - September 2019
To better communicate site updates, we’re going to experiment with a once-monthly updates post. We’ll pin this post pin on the homepage for roughly a week and then continue to post updates here throughout the month.
Please also feel free to use the comments section on this post as a S... |
485821ed-b888-4e92-95a4-4fb74aace81d | trentmkelly/LessWrong-43k | LessWrong | The Weighted Majority Algorithm
Followup to: Worse Than Random, Trust In Bayes
In the wider field of Artificial Intelligence, it is not universally agreed and acknowledged that noise hath no power. Indeed, the conventional view in machine learning is that randomized algorithms sometimes perform better than unrandom... |
22af7161-78a8-4c2f-95a4-4ba2d7c6c323 | trentmkelly/LessWrong-43k | LessWrong | Understanding Machine Learning (II)
,,
This is part 2 of a three part series on the fundamentals of Machine Learning as presented by this book. It builds heavily onto part I.
Uniform Convergence
Uniform convergence is a property which a hypothesis class may satisfy that will end up being equivalent to PAC learning, ... |
2e0e0209-619e-4de6-ac02-abcec55fcc31 | StampyAI/alignment-research-dataset/aisafety.info | AI Safety Info | Could AI alignment research be bad? How?
While the hope is that AI alignment research will have a good outcome, there are a few noteworthy ways it could be bad.
## 1. Accelerating capabilities
Many aspects of alignment research are also relevant for increasing AI capabilities. If a particular approach to alignment r... |
18bad9be-4af5-46f9-8da7-c493ccab6c05 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Am I secretly excited for AI getting weird?
*This post is arguably darker than my* [*other one*](https://www.lesswrong.com/posts/K4urTDkBbtNuLivJx/why-i-think-strong-general-ai-is-coming-soon)*. I don't make any persuasive arguments about AI forecasting here; if you don't feel like looking at doominess, feel free to s... |
2e149e24-7ae1-4498-95b4-78db30f69847 | trentmkelly/LessWrong-43k | LessWrong | Book review: WEIRDest People
Book review: The WEIRDest People in the World, by Joseph Henrich.
Wow!
Henrich previously wrote one of the best books of the last decade. Normally, I expect such an author's future books to, at best, exhibit regression toward the mean. But Henrich's grand overview of humanity's first few... |
35ce9af1-82e9-4647-b250-b819adf916a5 | trentmkelly/LessWrong-43k | LessWrong | Explicit content
The song The Sound of Silence has been on my mind, and ... what if they actually meant the lyrics? What if in lots of popular songs, people were actually trying to tell people a thing?
I know that "popular music contains messages for me" is a typical crazy-person thing to say. But, to be fair, they a... |
efb71083-055c-4681-a38d-2fac6d581749 | trentmkelly/LessWrong-43k | LessWrong | HCH is not just Mechanical Turk
HCH, introduced in Humans consulting HCH, is a computational model in which a human answers questions using questions answered by another human, which can call other humans, which can call other humans, and so on. Each step in the process consists of a human taking in a question, option... |
c4d51732-1b15-4a41-825e-7869db6871b6 | trentmkelly/LessWrong-43k | LessWrong | Paying people to vote for good policies.
Confidence level/s: very speculative, low
The problem
It seems that one of the key optimization problems in democratic decision making is that we often have little incentive to vote for policies that we truly think will increase the public utility.
As per Jess Whittlestone a... |
cf0eb89a-bb14-447d-8890-d2b57993e348 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | [Linkpost] My attempt at trying to summarize 'Intro to ML Safety'
After going through [Intro to ML Safety](https://course.mlsafety.org/), I decided to condense the course into a ~30-page document. While I do not think this comes close to a substitute for the course, I figured this might be useful as a litmus test to s... |
61417769-1b92-4600-9b79-061e4f7508f0 | trentmkelly/LessWrong-43k | LessWrong | Q&A with experts on risks from AI #3
[Click here to see a list of all interviews]
I am emailing experts in order to raise and estimate the academic awareness and perception of risks from AI.
Dr. Pei Wang is trying to build general-purpose AI systems, compare them with human intelligence, analyze their theoretical as... |
f93661b2-0fc4-43ae-9d7a-caf1e424ff9a | trentmkelly/LessWrong-43k | LessWrong | DeepMind: Generally capable agents emerge from open-ended play
EDIT: Also see paper and results compilation video!
> Today, we published "Open-Ended Learning Leads to Generally Capable Agents," a preprint detailing our first steps to train an agent capable of playing many different games without needing human interac... |
27e095bc-a075-4dc9-803a-de57aa992b39 | trentmkelly/LessWrong-43k | LessWrong | Cryonics prices were more complicated when I double-clicked
I know there are a few folks signed up for cryonics on LessWrong. At time of writing, I have decided not to do so. This covers part of the research I did while deciding: how expensive is cryonics?
Life Insurance
I am in Canada where only one insurance compa... |
954f65a1-7236-4e1b-a82f-16d7e375d928 | trentmkelly/LessWrong-43k | LessWrong | How safe "safe" AI development?
There are now at least two AI development companies purporting to research safe AI and both have existed for a couple years now, so I think it's worth taking another look at how safe "safe" AI development is.
One of those companies is GoodAI. I don't know a lot about them beyond what's... |
92cee731-c82d-4708-98a3-da98a91d341e | trentmkelly/LessWrong-43k | LessWrong | Social class amongst the intellectually gifted
Something that I've come to realize is that as a practical matter, intellectually gifted people who haven't developed very strong ability in a quantitative subject tend to be at a major disadvantage relative to those who have. The quantitative subjects that I have in mind... |
5dbf68d4-a970-489c-b568-f999134282ba | trentmkelly/LessWrong-43k | LessWrong | C19 Prediction Survey Thread
I'm curious about people's current predictions of future:
* US confirmed case count
* US death count
As a schelling point, let's use May 1 (5/01/2020) as the prediction date, and worldometer as the data source.
Predictions should be in the form of a simple distribution such as a norma... |
e7ae2f17-3968-4709-8c42-3b5195a79dd6 | trentmkelly/LessWrong-43k | LessWrong | Problem relaxation as a tactic
It's easier to make your way to the supermarket than it is to compute the fastest route, which is yet easier than computing the fastest route for someone running backwards and doing two and a half jumping jacks every five seconds and who only follows the route p percent of the time. Some... |
94911430-45a6-477f-9d74-7005cf45d3e1 | trentmkelly/LessWrong-43k | LessWrong | Revising Stages-Oversight Reveals Greater Situational Awareness in LLMs
Summary
The Stages-Oversight benchmark from the Situational Awareness Dataset tests whether large language models (LLMs) can distinguish between evaluation prompts (such as benchmark questions) and deployment prompts (real-world user inputs). This... |
c8fb46e9-201a-4e86-9b19-af245f0f3d64 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Meetup: Vancouver
Discussion article for the meetup : Meetup: Vancouver
WHEN: 27 July 2013 01:30:00PM (-0700)
WHERE: 2505 W Broadway, Vancouver, BC
Tomorrow we will be discussing the Landmark Education forum, a self-improvement workshop attended by some of our members, who reflect positively on it and woul... |
232d219a-404f-45ae-be0c-52886767d8c5 | trentmkelly/LessWrong-43k | LessWrong | Racial Dating Preferences and Sexual Racism
Note: This is a linkpost from my personal substack. This is on a culture war topic, which is not normally the focus of my blogging. Rationalist friends suggested that this post might be interesting and surprising to LW readers.
Summary
* People widely exclude romantic and... |
5e819f89-db2d-498e-9ccc-a2157a67056c | trentmkelly/LessWrong-43k | LessWrong | Live web-forum Q&A on Friendly AI, Thu. May 24 (Hebrew)
On May 24 from 7 to 9pm Israel time, I will be answering questions and leading a discussion (in Hebrew) sponsored by the Galileo popular science magazine.
The topic of discussion will be my article "Superhuman Intelligence, Unhuman Intelligence," from the May ed... |
19ae93aa-a722-48ac-ae6e-392efe1881e5 | trentmkelly/LessWrong-43k | LessWrong | Is This Thing Sentient, Y/N?
Builds up on: Consciousness and the Brain by Stanislas Dehaene. Good summaries may be found here or here, though reading them is not strictly necessary.
Synopsis: I claim to describe the exact mental structure that allows qualia.
----------------------------------------
Background: Wha... |
50c5c987-1fbd-4fca-9bf3-c0e5bc93599d | trentmkelly/LessWrong-43k | LessWrong | The Existential Dread of Being a Powerful AI System
I was reading about honeypots today. Honeypots (in AI evaluations, at least) are described as testing scenarios that could trick an AI into revealing deceptive or misaligned intent. My first thought was that, especially for current generation large language models, a... |
91de9e69-a8a4-40f0-aa23-e3e4bab7a6f5 | trentmkelly/LessWrong-43k | LessWrong | EA & LW Forum Weekly Summary (23rd - 29th Jan '23)
Supported by Rethink Priorities
This is part of a weekly series summarizing the top posts on the EA and LW forums - you can see the full collection here. The first post includes some details on purpose and methodology. Feedback, thoughts, and corrections are welcomed... |
d19396d2-13e5-46c4-9f55-b1430dca6319 | trentmkelly/LessWrong-43k | LessWrong | Evaluating the historical value misspecification argument
ETA: I'm not saying that MIRI thought AIs wouldn't understand human values. If there's only one thing you take away from this post, please don't take away that. Here is Linch's attempted summary of this post, which I largely agree with.
Recently, many people h... |
e376cedd-1cb1-4332-9c9c-608a97fac5ea | trentmkelly/LessWrong-43k | LessWrong | Dangers of deference
[Written September 02, 2022. Note: I'm likely to not respond to comments promptly.]
Sometimes people defer to other people, e.g. by believing what they say, by following orders, or by adopting intents or stances. In many cases it makes sense to defer, since other people know more than you about m... |
4dd64072-377e-4000-8638-99c8dbdf5571 | trentmkelly/LessWrong-43k | LessWrong | mindlevelup: 2 Years of Blogging
An overview of how blogging for me went in 2017. Goes over some highlights of the year and outlines what I think might be worth writing more essays about in 2018. |
8eee1bed-f477-46aa-8959-623c53a53c2a | trentmkelly/LessWrong-43k | LessWrong | Noticing Frame Differences
Previously: Keeping Beliefs Cruxy
----------------------------------------
When disagreements persist despite lengthy good-faith communication, it may not just be about factual disagreements – it could be due to people operating in entirely different frames — different ways of seeing, thin... |
2f7c5e30-d947-4437-87b7-e5295744ff4d | trentmkelly/LessWrong-43k | LessWrong | Am I Really an X?
As I understand it, there is a phenomenon among transgender people where no matter what they do they can't help but ask themselves the question, "Am I really an [insert self-reported gender category here]?" In the past, a few people have called for a LessWrong-style dissolution of this question. This... |
a68537e8-a3cb-4ab3-b949-52d73e50f879 | trentmkelly/LessWrong-43k | LessWrong | Yudkowsky vs Trump: the nuclear showdown.
Sorry for the slightly clickbait-y title.
Some commenters have expressed, in the last open thread, their disappointment that figureheads from or near the rationality sphere seemed to have lost their cool when it came to this US election: when they were supposed to be calm and... |
2ad40887-7d4b-41bd-8a2e-ef27c13fb91f | trentmkelly/LessWrong-43k | LessWrong | Priming with Hypothetical questions
I came across this article this morning via a blog post from
http://solutionfocusedchange.blogspot.com/.
http://sd1.myipcn.org/science/article/pii/S0749597811001099
"Wolves in sheep’s clothing: How and when hypothetical questions influence behavior" by Sarah G. Moore and others. ... |
3baccb20-2ec6-4a90-a622-f88633c720da | trentmkelly/LessWrong-43k | LessWrong | Chapter 55: The Stanford Prison Experiment, Pt 5
In a scarred and ruined corridor, lit by dim gas lights, a boy slowly crept forward, one hand stretched out, toward the unmoving snake that was the body of his teacher.
Harry was only a meter away from the snake's body when he first felt it, tickling at the edge of his... |
c81174b0-cfb5-49e2-b6ba-164ecc3cfab6 | awestover/filtering-for-misalignment | Redwood Research: Alek's Filtering Results | id: post3918
A putative new idea for AI control; index here . In a preceding post , I showed how to get a wide variety of probability and utility estimates out of a boxed AI, without fearing the AI would mislead us. Specifically, given some bounded utility u and some stochastic message Y of length l , we could get est... |
cb812f43-b616-4823-af5c-d335674f651f | trentmkelly/LessWrong-43k | LessWrong | Multiple Factor Explanations Should Not Appear One-Sided
In Policy Debates Should Not Appear One-Sided, Eliezer Yudkowsky argues that arguments on questions of fact should be one-sided, whereas arguments on policy questions should not:
> On questions of simple fact (for example, whether Earthly life arose by natural ... |
9ca01a38-1f60-4bd8-9b8a-579df7c4f315 | trentmkelly/LessWrong-43k | LessWrong | An Interactive Shapley Value Explainer
This Shapley Value Calculator breaks down the formula in simple steps, using cats. |
0157b599-b08b-4ce1-9400-053d4df92d71 | trentmkelly/LessWrong-43k | LessWrong | One-dimensional vs multi-dimensional features in interpretability
> Chris Olah's “What is a Linear Representation? What is a Multidimensional Feature?” (July Circuits Update) prompted a moment of pause for me regarding the term "one-dimensional feature." I initially conflated that phrase with the number of dimensions ... |
57d605d3-a47c-4159-a26c-672422fcf9e2 | trentmkelly/LessWrong-43k | LessWrong | Maximal lotteries for value learning
Suppose Ann asks robot Bob to buy her favorite fruit. Bob doesn't know what Ann's favorite fruit is, but he knows that Ann likes either apples, bananas or oranges, each with probability 1/3, and doesn't like the other two fruits at all. At the groceries, for his money Bob can buy 1... |
387e1635-a265-4b72-9d0b-e9b844fde4f6 | StampyAI/alignment-research-dataset/arbital | Arbital | A googolplex
The number $10^{10^{100}}$ is named "a googolplex." That is, a googolplex is the number represented (in [https://arbital.com/p/-decimal_notation](https://arbital.com/p/-decimal_notation)) as a 1 followed by a [https://arbital.com/p/-42m](https://arbital.com/p/-42m) zeroes, i.e., $10^{googol}$, which is $ ... |
e5349196-d4bb-4537-88b9-de6551799ce3 | trentmkelly/LessWrong-43k | LessWrong | Should we remove markdown parsing from the comment editor?
As was just brought up in the discussion about the vote, the current markdown-parsing in the comments is causing quite a few bugs and the associated convenience might not be worth the costs of the bugs. We now have an experimental option in your user-profile ... |
eab17037-2a6a-47a8-b77c-06c19e6763cd | trentmkelly/LessWrong-43k | LessWrong | Stargate AI-1
There was a comedy routine a few years ago. I believe it was by Hannah Gadsby. She brought up a painting, and looked at some details. The details weren’t important in and of themselves. If an AI had randomly put them there, we wouldn’t care.
Except an AI didn’t put them there. And they weren’t there at ... |
0624cc21-3ee0-4004-9508-61c77fae5637 | trentmkelly/LessWrong-43k | LessWrong | [SEQ RERUN] When Life is Cheap, Death is Cheap
Today's post, When Life Is Cheap, Death Is Cheap was originally published on November 24, 2008. A summary:
> Human cultures have historically adapted to scenarios in which they were the victims of genocide, etc. Ems would likely respond this way as well.
Discuss the... |
3827d432-67b2-471b-a1a5-76b705988a38 | trentmkelly/LessWrong-43k | LessWrong | What's the difference between GAI and a government?
I have zero technical AI alignment knowledge. But this question has kept recurring to me for like a year now so I thought I'd ask.
A lot of the arguments for the danger of GAI revolve around the notion that an agent that is smarter than a human is un-boxable, self-c... |
d19d9fc9-03ea-4db5-9984-9ae30e3acc35 | trentmkelly/LessWrong-43k | LessWrong | Domain-specific SAEs
TLDR: Current SAE training doesn't specifically target features we care about, e.g. safety-relevant ones. In this post, we compare three ways use to SAEs to efficiently extract features relevant to a domain of interest.
Introduction
If Sparse Autoencoders (SAEs) are to be useful for alignment, t... |
67d6b270-a479-4813-8b89-0c9098b23c7f | trentmkelly/LessWrong-43k | LessWrong | They are made of repeating patterns
Epistemis status: an obvious parody.
— You won't believe me. I've found them.
— Whom?
— Remember that famous discovery by Professor Prgh'zhyne about pockets of baryonic matter in open systems that minimize the production of entropy within them? They went further and claimed that... |
73198d49-4e98-4702-ae05-892a066ef9b2 | StampyAI/alignment-research-dataset/arxiv | Arxiv | Aligning Superhuman AI with Human Behavior: Chess as a Model System
1. Introduction
----------------
Artificial intelligence is becoming increasingly intelligent, equalling and surpassing peak human performance in an increasing range of domains (Esteva et al., [2017](#bib.bib7); Silver
et al., [2018](#bib.bib15)). ... |
50fbd2eb-093f-4214-a98f-71769571dc2c | trentmkelly/LessWrong-43k | LessWrong | Will compute bottlenecks prevent a software intelligence explosion?
Epistemic status – thrown together quickly. This is my best-guess, but could easily imagine changing my mind.
Intro
I recently copublished a report arguing that there might be a software intelligence explosion (SIE) – once AI R&D is automated (i.e. ... |
7a8b632c-73c1-4acc-a4fb-f5e228fe2532 | trentmkelly/LessWrong-43k | LessWrong | The "show, don't tell" nature of argument
Consider a statement of the form "based on knowledge and common sense and estimating the probabilities of alternative hypotheses, I believe that X". Or, perhaps, "based on Bayesian reasoning I have come to the conclusion X". How much, when an interlocutor of yours uses such a ... |
2c5d1d7a-f953-4ddd-b08c-db8e0a4be6f6 | trentmkelly/LessWrong-43k | LessWrong | Why Would Belief-States Have A Fractal Structure, And Why Would That Matter For Interpretability? An Explainer
Yesterday Adam Shai put up a cool post which… well, take a look at the visual:
Yup, it sure looks like that fractal is very noisily embedded in the residual activations of a neural net trained on a toy pro... |
25b84362-4ffe-4368-b823-e3ff8a13f18d | trentmkelly/LessWrong-43k | LessWrong | "It isn't magic"
People keep saying "AI isn't magic, it's just maths" like this is some kind of gotcha.
Turning lead into gold isn't the magic of alchemy, it's just nucleosynthesis.
Taking a living human's heart out without killing them, and replacing it with one you got out a corpse, that isn't the magic of necroma... |
9c6b2527-6469-49a7-92f3-577d452292c6 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Conceptual Problems with UDT and Policy Selection
Abstract
========
UDT doesn't give us conceptual tools for dealing with multiagent coordination problems. There may have initially been some hope, because a UDT player can select a policy which incentivises others to cooperate, or because UDT can reason (EDT-style) th... |
d1492465-05fe-40de-b6b2-f65135b2a567 | trentmkelly/LessWrong-43k | LessWrong | Truthseeking when your disagreements lie in moral philosophy
[Status: latest entry in a longrunning series]
My last post on truthseeking in EA vegan advocacy got a lot of comments, but there’s one in particular I want to highlight as highly epistemically cooperative. I have two motivations for this:
* having just s... |
82b0251f-2a9c-4a95-9f2f-4819a20358d7 | trentmkelly/LessWrong-43k | LessWrong | One Sided Policy Debate - The Science of Literature
On HackerNews, this article was linked. The general idea is that companies are studying what people like to read, to help authors produce books that people like to read.
Now, for me, when I look at this idea, I see some down sides, but I certainly see some benefits ... |
04623621-c5c1-41e2-8720-0b39332f7c97 | trentmkelly/LessWrong-43k | LessWrong | Thoughts on AGI consciousness / sentience
Summary
I’ll walk through my take on how consciousness works (basically endorsing “illusionism”) and why I think computer programs can be conscious, and how that relates to moral and ethical obligations towards conscious entities in general and AIs in particular.
Up to here, ... |
6723bb7d-16b1-4847-82da-32fa6339ac78 | trentmkelly/LessWrong-43k | LessWrong | Standing on a pile of corpses
[In the darkest day of 2018, it is proper to think about the darkness that surrounds us]
When we think about the history of humanity, we focus on its highlights.
Galileo discovering the moons of Jupiter. Edward Jenner developing the first pox vaccine. Emmy Noether setting the mathematic... |
84efa0b9-72e7-4e53-8643-5ee20e272475 | trentmkelly/LessWrong-43k | LessWrong | Meetup : London Social Meetups - 23/03 and 30/03
Discussion article for the meetup : London Social Meetups - 23/03 and 30/03
WHEN: 23 March 2014 02:00:00PM (+0000)
WHERE: Shakespeare's Head, Africa House, 64 Shakespeare's Head, Africa House, 64-68 Kingsway, London WC2B 6BG, UK-68 Kingsway, London WC2B 6BG, UK
LessW... |
f8c1e172-c0f0-4d50-834f-92fb78efcea2 | trentmkelly/LessWrong-43k | LessWrong | Weekly LW Meetups
This summary was posted to LW Main on November 18th. The following week's summary is here.
The following meetups take place in cities with regular scheduling, but involve a change in time or location, special meeting content, or simply a helpful reminder about the meetup:
* Bay Area Winter Solstic... |
ec5aa0ce-483a-4c40-8acd-09f96823b6eb | trentmkelly/LessWrong-43k | LessWrong | Biases are engines of cognition
This work was done while at Conjecture.
Humans are not perfect reasoners. Cognitive biases are not biases; they are engines of cognition, each exploiting regularities of the universe. Combining enough of these “cognitive biases” was enough to create a general intelligence but doesn’t ... |
fa0e6e2c-78f5-4cee-95e5-46073a93ba8a | trentmkelly/LessWrong-43k | LessWrong | Open Thread, April 15-30, 2013
If it's worth saying, but not worth its own post (even in Discussion), then it goes here. |
d0d09af4-0f9a-4c93-bc74-cffb55714db2 | trentmkelly/LessWrong-43k | LessWrong | How good is a human's gut judgement at guessing someone's IQ?
In her latest post, Sarah Constantin writes:
> Whatever ability IQ tests and math tests measure, I believe that lacking that ability doesn’t have any effect on one’s ability to make a good social impression or even to “seem smart” in conversation.
I remem... |
f248e699-a6cc-40ef-bd75-886aa07a44a7 | trentmkelly/LessWrong-43k | LessWrong | D&D.SCP: Anomalous Acquisitions
STORY (somewhat longer than usual, still skippable)
When you graduated top of your class from Data Science School, you didn't care where you ended up working, you just wanted to find the highest-paid job possible. (Your student loans may have had an impact on this decision).
(Yes, I k... |
9cf7a7c3-1de7-4f7c-8dfd-94dd32c69e3a | StampyAI/alignment-research-dataset/youtube | Youtube Transcripts | 197. Ben Garfinkel on Scrutinizing Classic AI Risk Arguments 2
Hello and welcome to session 197 in the AI
Safety Reading Group.
Tonight we’ll continue with the podcast
of Ben Garfinkel on Scrutinizing Classic AI
Risk Arguments.
Ben Garfinkel is still a research fellow and
Howie Lempel, Robert Wiblin, and Keiran Harris... |
90c2bf92-77d3-4ee5-be73-7a4b1e301c5a | trentmkelly/LessWrong-43k | LessWrong | Double Crux — A Strategy for Mutual Understanding
Preamble
Double crux is one of CFAR's newer concepts, and one that's forced a re-examination and refactoring of a lot of our curriculum (in the same way that the introduction of TAPs and Inner Simulator did previously). It rapidly became a part of our organizational s... |
b723af6b-c11f-431d-b55c-25e8bb0f9258 | trentmkelly/LessWrong-43k | LessWrong | Hypothetical Paradoxes
When we form hypotheticals, they must use entirely consistent and clear language, and avoid hiding complicated operations behind simple assumptions. In particular, with respect to decision theory, hypotheticals must employ a clear and consistent concept of free will, and they must make all infor... |
eacbd3a0-8bb9-4b4e-97f2-005f40dd9756 | trentmkelly/LessWrong-43k | LessWrong | Rationalist Community Hub in Moscow: 3 Years Retrospective
Short summary: Moscow rationalist community is organized around a local venue called Kocherga. Kocherga hosts 40–50 rationality-adjacent events per month, teaches a few dozen people per year in CFAR-style applied rationality workshops, and has ambitious plans ... |
7969da75-f28b-49dd-a74b-6c49fec92985 | trentmkelly/LessWrong-43k | LessWrong | Multispecies Metagenomic Calibration
Cross-posted from my NAO Notebook.
This is something I wrote internally in late-2022. Sharing it now with light edits, additional context, and updated links after the idea came up at the Microbiology of the Built Environment conference I'm attending this week.
Metagenomic sequenc... |
5b9b1a4f-67d4-4a21-b991-f01efcd2cc74 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | AI Safety Fundamentals: An Informal Cohort Starting Soon!
Hi Everyone,
We'll be starting an informal cohort on [AI Safety Fundamentals](https://aisafetyfundamentals.com/ai-alignment-curriculum).
Come join-us on [Discord](https://discord.gg/gW9Kq6VHgM). We'll do a poll next week to find a slot that works for most ... |
80c028d6-ae8c-44bd-a64a-bc36b5700fb4 | StampyAI/alignment-research-dataset/arbital | Arbital | Intension vs. extension
To give an "intensional definition" is to define a word or phrase in terms of other words, as a dictionary does. To give an "extensional definition" is to point to examples, as adults do when teaching children. The preceding sentence gives an intensional definition of "extensional definition"... |
6f0da215-6d64-49d2-aa07-b4d477275f30 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Weekly Chicago Meetups Resume 5/26
Discussion article for the meetup : Weekly Chicago Meetups Resume 5/26
WHEN: 26 May 2012 01:00:00PM (-0500)
WHERE: 360 N Michigan Ave, Chicago IL
Because of all of the craziness happening in downtown Chicago, with the NATO meeting and associated protests, the 5/19 Chicago... |
e444aebe-89b4-45eb-b2ea-da2144699173 | trentmkelly/LessWrong-43k | LessWrong | My thesis (Algorithmic Bayesian Epistemology) explained in more depth
In March I posted a very short description of my PhD thesis, Algorithmic Bayesian Epistemology, on LessWrong. I've now written a more in-depth summary for my blog, Unexpected Values. Here's the full post:
***
In January, I defended my PhD thesi... |
44eb1e64-0653-4da8-a1f0-ae5dfe20f232 | trentmkelly/LessWrong-43k | LessWrong | Thousands of malicious actors on the future of AI misuse
Announcing the results of a 2024 survey by Convergence Analysis. We’ve just posted the executive summary below, but you can read the full report here.
In the largest survey of its kind, Convergence Analysis surveyed 2,779 malicious actors on how they would mis... |
bfe0806d-7973-420f-9694-2b9f541a315e | StampyAI/alignment-research-dataset/arxiv | Arxiv | Norms for Beneficial A.I.: A Computational Analysis of the Societal Value Alignment Problem
1 Introduction
---------------
Several applications already have an Artificial Intelligent system (A.I.) taking decisions in place of their owners. It is expected that in the future, such delegation of decisions will become ... |
0a1d0f41-420a-45a8-80cc-7d190e5585ce | trentmkelly/LessWrong-43k | LessWrong | Minimal intuitionistic logic as a setting for logical counterfactuals
|
228333d5-07d9-403f-a838-02b0e91d49bb | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Xanadu, GPT, and Beyond: An adventure of the mind
Cross posted [from New Savanna](https://new-savanna.blogspot.com/2023/08/xanadu-gpt-and-beyond-adventure-of-mind.html).
I've posted a new w... |
4efa0a17-4ee2-4e01-87eb-ab19a7a3e28c | StampyAI/alignment-research-dataset/special_docs | Other | Maximum Margin Planning
Maximum Margin Planning
Nathan D. Ratliff ndr@ri.cmu.edu
J. Andrew Bagnell dbagnell@ri.cmu.edu
Robotics Institute, Carnegie Mellon University, Pittsburgh, PA. 15213 USA
Martin A. Zinkevich maz@cs.ualberta.ca
Department of Computing Science, University of Alberta, Edmonton, AB T6G 2E1, Canada
Abs... |
d2e01118-e82d-41a5-bace-0f7fae987a7c | trentmkelly/LessWrong-43k | LessWrong | What should you do regarding footwear to protect against COVID-19?
From [Effectiveness of personal protective measures in prevention of nosocomial transmission of severe acute respiratory syndrome]:
"Univariate analysis showed that mask, gown, gloves, goggles, footwear, "hand-washing and disinfecting", gargle, "membr... |
1195c162-7d7e-4e54-9ba5-e05e9e74cf56 | trentmkelly/LessWrong-43k | LessWrong | Beta-Beta Testing: Frontpage Rework [Update - further tweak]
We have a rework of the front page on Lesserwrong, that we'd like to sanity-check before either deploying live or putting more work into.
We've deployed it to our beta-beta site, Lessestwrong.com. (The main shift here is in how the list of posts work, and h... |
6387bc03-ba9c-4681-bbe0-7500a785c5f6 | trentmkelly/LessWrong-43k | LessWrong | Covid Home Test Experiment
I was exposed to COVID 4 days ago. I received the Pfizer vaccine second dose 4 months ago. I have done 2 rapid tests today. One came back positive, one negative. I would like to do 18 more and report back my data. However, I cannot afford that many home tests.
Would anyone like to sponsor t... |
c901be1f-a126-42ac-ba93-5b7caec2c57b | trentmkelly/LessWrong-43k | LessWrong | Algorithm detects multiple authors in of the Old Testament [link]
http://www.haaretz.com/jewish-world/israeli-software-aims-to-shed-light-on-the-bible-1.370343
The more interesting aspect of the story is that the algorithm was originally developed by an (rather impressive IMO) orthodox jewish CS researcher - http://u... |
61d82dea-7b63-41f5-af37-8cdbec567521 | trentmkelly/LessWrong-43k | LessWrong | On characterizing heavy-tailedness
[CONTEXT: For a while I have been meaning to engage with a literature review on heavy tailed distributions. Instead of just indefinitely postponing the project I resolved to write some preliminary thoughts on the topic, so I can get started on understanding the concept better with a ... |
27437b8d-1e34-4afe-994c-9d219f1630da | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | [AN #119]: AI safety when agents are shaped by environments, not rewards
Alignment Newsletter is a weekly publication with recent content relevant to AI alignment around the world. Find all Alignment Newsletter **[resources here](http://rohinshah.com/alignment-newsletter/)**. In particular, you can look through **[thi... |
90f8baab-f40f-456c-9961-a3b860d54507 | trentmkelly/LessWrong-43k | LessWrong | Confidence levels inside and outside an argument
Related to: Infinite Certainty
Suppose the people at FiveThirtyEight have created a model to predict the results of an important election. After crunching poll data, area demographics, and all the usual things one crunches in such a situation, their model returns a gre... |
eee8d929-f8ad-4722-aa85-d6b67795e8a7 | trentmkelly/LessWrong-43k | LessWrong | Trick-or-treating in Covid Times
Somerville has asked its residents not to trick-or-treat:
> Mayor Joseph A. Curtatone and the Somerville Board of Health announced Halloween guidance and related updates today. They strongly urge all community members to forgo trick-or-treating in favor of lower-risk activities as def... |
59cda092-cdeb-4c14-9a6a-2332a39535bf | trentmkelly/LessWrong-43k | LessWrong | Plans are Recursive & Why This is Important
Epistemic status: Reference. Highly confident. I rely on what is presented here extensively in my own thinking.
Plans are recursive. Any plan can be decomposed into parts and those parts can be in turn be decomposed into further parts and so on until it is senseless to deco... |
3b127536-40cf-4160-8387-fa757cf7d394 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Frankfurt: New LW apartment and guests
Discussion article for the meetup : Frankfurt: New LW apartment and guests
WHEN: 08 November 2014 02:00:00PM (+0100)
WHERE: Frankfurt, Heerstrasse
We have just moved in with each other! And we have guests. One Earning to give person (and EA) from London and one Effect... |
ca205455-5667-4b2c-be42-cc69055d6678 | trentmkelly/LessWrong-43k | LessWrong | Transcription and Summary of Nick Bostrom's Q&A
INTRO: From the original posting by Stuart_Armstrong:
> Underground Q&A session with Nick Bostrom (http://www.nickbostrom.com) on existential risks and artificial intelligence with the Oxford Transhumanists (recorded 10 October 2011).
>
> http://www.youtube.com/watch?v... |
f0ab7e7e-dd73-4c10-8127-3ac42298f1b3 | trentmkelly/LessWrong-43k | LessWrong | Second order taste
Last year I got my friend a gift card to Beast and Cleaver for his wedding, a local butcher shop and restaurant.
I was pretty proud of the gift. He and his wife said the meal was amazing and I wouldn't expect them to say that if it weren't true. Getting something on the wedding registry or perhaps ... |
e62a33e7-f80c-4a2c-89e1-875124d49614 | trentmkelly/LessWrong-43k | LessWrong | Osaka
The more I learn about urban planning, the more I realize that the American city I live in is dystopic. I'm referring specifically to urban planning, and I'm not being hyperbolic. Have you ever watched the teen dystopia movie Divergent? The whole city is perfectly walkable (or parkourable, if you're Dauntless). ... |
d9094d7e-3739-47a7-a36a-cd53280ef907 | trentmkelly/LessWrong-43k | LessWrong | What's the evidence on falling testosteron and sperm counts in men?
One occasionally hears some concerns about falling testosteron/sperm counts, usually in some narrative about the Good Olden days when Men were still real Men, etc. It sounds a little like ' they are turning the frogs gay' type of stuff, but perhaps th... |
e2668833-061e-4bdb-943b-0e6c2a954577 | trentmkelly/LessWrong-43k | LessWrong | No Universally Compelling Arguments
What is so terrifying about the idea that not every possible mind might agree with us, even in principle?
For some folks, nothing—it doesn't bother them in the slightest. And for some of those folks, the reason it doesn't bother them is that they don't have strong intuitions about ... |
fbf040a5-40fa-46b4-a9ae-75954fbcbf1c | trentmkelly/LessWrong-43k | LessWrong | Progress links and tweets, 2022-10-12
Announcements
* Our World in Data is hiring a Senior Data Visualization Engineer
* World’s Fair Co looking for space in San Francisco to host meetings with creatives; can anyone help? (@camwiese)
Links
* “A well-known Dussehra tradition is to worship all weapons, tools, instr... |
c21bbbb9-119f-48a8-b7e6-a164fbfa5a8d | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Framings of Deceptive Alignment
In this post I want to lay out some framings and thoughts about deception in misaligned AI systems.
Types of Deception
------------------
There seem to be two different things which people mean by deception which have different causes and likely different effects. Because these are b... |
aea85b6d-d0d9-435f-b203-2e3e81c96dda | trentmkelly/LessWrong-43k | LessWrong | Moonlight Reflected
Previously in sequence: Navigation by Moonlight
Cross-posted from SecondPerson.dating.
----------------------------------------
I’m very proud of Navigation by Moonlight, and not just because it got a warm reception and resonated with many readers. I’m proud because I didn’t understand it well e... |
c2330cf8-1f73-4216-8b2d-2d0b16503a3a | trentmkelly/LessWrong-43k | LessWrong | Progress links and tweets, 2022-07-12
Links
* Patrick Collison on universal covid vaccines: we could have them very soon, if it weren’t for bureaucratic roadblocks
* “Current funding organizations cluster heavily around a few management styles, warping the research ecosystem” (Ben Reinhardt)
* The Tony Blair Instit... |
57ab1653-e2e9-4d54-9317-b3bfc490dd32 | trentmkelly/LessWrong-43k | LessWrong | Edinburgh Lesswrong meetup
According to statistics, Edinburgh has a reasonable number of LW readers. Currently no Edinburgh meetups exist for Less Wrong so we might as well arbitrarily set one in the near future and see if it could work out (if there are enough people who are interested in a meetup here).
Date/time: ... |
dca4c37c-6412-4206-9f0a-6433c0f4f976 | StampyAI/alignment-research-dataset/blogs | Blogs | 2018 Update: Our New Research Directions
For many years, MIRI’s goal has been to resolve enough fundamental confusions around [alignment](https://intelligence.org/2016/12/28/ai-alignment-why-its-hard-and-where-to-start/) and intelligence to enable humanity to think clearly about technical AI safety risks—and to do thi... |
f1d31e11-2f67-4df4-b6a1-4da468027809 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | When is Goodhart catastrophic?
*Thanks to Aryan Bhatt, Eric Neyman, and Vivek Hebbar for feedback.*
*This post gets more math-heavy over time; we convey some intuitions and overall takeaways first, and then get more detailed. Read for as long as you're getting value out of things!*
TLDR
----
How much should you opt... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.