id stringlengths 36 36 | source stringclasses 15
values | formatted_source stringclasses 13
values | text stringlengths 2 7.55M |
|---|---|---|---|
4d4d8e60-5251-40f1-b1a9-b5711f9f001e | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | What's the best machine learning newsletter? How do you keep up to date?
I'm looking for a way to make sure I reliably learn about the biggest developments such as [Transformers](https://arxiv.org/abs/1706.03762), AlphaFold, or the grokking [paper](https://arxiv.org/pdf/2201.02177.pdf). I don't currently want to spend... |
4faf6420-068a-4720-bf1e-e188471803bc | trentmkelly/LessWrong-43k | LessWrong | Book review: The Quincunx
The Quincunx is a 1989 novel by Charles Palliser, set in early 1800s England. I want to recommend it to everyone because it's really good, and it might be relevant to the AI transition. Let me try to explain.
The surface level of the book is a kind of mishmash of Dickensian themes. The main ... |
e67b4a7a-63cd-4f7a-bf3d-fe63253cd727 | trentmkelly/LessWrong-43k | LessWrong | Contra Yudkowsky on Epistemic Conduct for Author Criticism
In a comment on the Effective Altruism Forum responding to Omnizoid's "Eliezer Yudkowsky Is Frequently, Confidently, Egregiously Wrong", Eliezer Yudkowsky writes:
> You will mark that in this comment I first respond to a substantive point and show it to be mi... |
8ef32aff-d24d-4586-9160-e695e2c70d95 | trentmkelly/LessWrong-43k | LessWrong | Sleeping Beauty, Cloning Beauty, and Frequentist Arguments In Anthropic Paradoxes
In a recent post, it has been said the Sleeping Beauty Problem is resoundingly considered resolved in academia. As someone who has been following the literature extensively for the past few years, I can positively say that is not true. T... |
f5ffeb75-b7a4-49ac-b2ea-5f804c5ae5c8 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | New paper: AGI Agent Safety by Iteratively Improving the Utility Function
This post is to announce my new paper [AGI Agent Safety by Iteratively Improving the Utility Function](https://arxiv.org/abs/2007.05411). I am also using this post to add some extra background information that is not on the paper. Questions and ... |
4ba35665-2cc2-47a3-8087-5169a211d600 | trentmkelly/LessWrong-43k | LessWrong | The Substitution Principle
Partial re-interpretation of: The Curse of Identity
Also related to: Humans Are Not Automatically Strategic, The Affect Heuristic, The Planning Fallacy, The Availability Heuristic, The Conjunction Fallacy, Urges vs. Goals, Your Inner Google, signaling, etc...
What are the best careers for m... |
7cc30c75-d34c-4ff7-871b-725d2058e7d3 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Take 11: "Aligning language models" should be weirder.
*As a writing exercise, I'm writing an AI Alignment Hot Take Advent Calendar - one new hot take, written ~~every day~~ some days for 25 days. I have now procrastinated enough that I probably have enough hot takes.*
People often talk about aligning language models... |
2bfb160e-a655-4d25-9c31-eeb36f317c07 | trentmkelly/LessWrong-43k | LessWrong | Is there any plausible mechanisms for why taking an mRNA vaccine might be undesirable for a young healthy adult?
We haven't seen any problem in the short term, but is the probability of long term negative side effects (given it's a new technology) sufficiently unlikely for a young healthy person to take it?
Am partic... |
7558959b-b6f1-4e92-b954-2290f894fcfd | trentmkelly/LessWrong-43k | LessWrong | AIDungeon 3.1
We live in a world of magic and sorcery. With a click of my fingers I can do anything I want. Summon a horde of dragons to my bidding. Turn a man into a million shards of glass. Split a mountain in two, and then raise a road to march through it.
I am the mage lord, the most powerful man in all Alaka. He... |
d6e3becf-4396-4f86-8f74-66bc002a99b2 | trentmkelly/LessWrong-43k | LessWrong | Bayesianism for humans: prosaic priors
(followup to What Bayesianism has taught me? and Bayesianism for Humans, relevant to Boring Advice Repository )
There are two insights from Bayesianism which occurred to me and which I hadn't seen anywhere else before.
I like lists in the two posts linked above, so for the sak... |
6d46ad55-fc0a-42b0-ba2c-f5b5d5ae6ac1 | trentmkelly/LessWrong-43k | LessWrong | Lost purposes - Doing what's easy or what's important
Original post: http://bearlamp.com.au/lost-purposes-doing-whats-easy-or-whats-important/
----------------------------------------
A friend came to me with a startup as her main goal. I am keen on my coaching skills so I would from time to time ask her what the t... |
804ea70f-9fce-45c7-bb12-ed1e69a293e3 | trentmkelly/LessWrong-43k | LessWrong | Roadmap: Plan of Action to Prevent Human Extinction Risks
Let’s do an experiment in "reverse crowdfunding”. I will pay 50 USD to anyone who can suggest a new way of X-risk prevention that is not already mentioned in this roadmap. Post your ideas as a comment to this post.
Should more than one person have the same ide... |
e0e8c151-4f08-4c70-9979-bbe927fdb19b | trentmkelly/LessWrong-43k | LessWrong | Progress = Fewer Bad Moments
Most people I help start with a tiny misconception that can wreck their progress.
Whether they’re working on anxiety, insecurity, confidence, etc., they picture progress like this:
They expect progress to feel like getting over a cold: each day the discomfort eases a little more. Sure, t... |
b5b4be00-3b2b-4169-a68d-f2bbf92ae859 | StampyAI/alignment-research-dataset/arxiv | Arxiv | How Should an Agent Practice?.
Introduction
------------
There are many applications of reinforcement learning (RL) in which the natural formulation of the reward function gives rise to difficult computational challenges, or in which the reward itself is unavailable for extended periods of time or is difficult to s... |
34536ae9-8d75-4376-a64e-30c9c39bee28 | trentmkelly/LessWrong-43k | LessWrong | Nonperson Predicates
Followup to: Righting a Wrong Question, Zombies! Zombies?, A Premature Word on AI, On Doing the Impossible
There is a subproblem of Friendly AI which is so scary that I usually don't talk about it, because very few would-be AI designers would react to it appropriately—that is, by saying, "Wow, t... |
df374227-bdaa-4dcb-8b02-f20204b53420 | trentmkelly/LessWrong-43k | LessWrong | What is science?
It seems like a very 101-level question that someone like me shouldn't have, but I am actually confused.
Is there a strict, black-and-white, approximately universally agreed upon definition? I don't get the impression that there is. People seem to use the term to mean different things.
Furthermore, ... |
c490bf42-ee99-40c2-a3c5-5e5a5880855f | trentmkelly/LessWrong-43k | LessWrong | Let’s go meta: Grammatical knowledge and self-referential sentences [ChatGPT]
Language is a medium that humans use for sharing thoughts, attitudes, and feelings. In order to do so it has to be embedded in some physical medium, whether sound, visual signs, of physical gestures. But the thoughts, attitudes, and feelings... |
9749c3b8-007f-4fa3-9e4f-d4f202cabdec | trentmkelly/LessWrong-43k | LessWrong | Simultaneous Footbass and Footdrums II
Getting ready for this Friday's Spark in the Dark I've modified my rhythm stage setup to support a new way of doing more things at once. I have four velocity-sensitive pedals, which I initially used for drums. Then I added bass as an option, and then year ago I got them set up fo... |
90851a2c-c8a2-4543-8712-2042673e1af0 | trentmkelly/LessWrong-43k | LessWrong | Meetup : LessWrong Scotland
Discussion article for the meetup : LessWrong Scotland
WHEN: 15 March 2015 03:00:00PM (+0000)
WHERE: Curler's Rest, 256-260 Byres Rd, Glasgow G12 8SH
After a successful trial meetup (five people!) in November, this will be the first meetup of 2015.
We don't have a definite plan yet, but... |
2f8a40da-604d-40c8-94e7-9874290148ab | trentmkelly/LessWrong-43k | LessWrong | Do you visualize Omega?
I find Omega so profoundly annoying that I don't bother with problems that include it. I know, it's just a philosophical tool, but between not believing that such a thing is possible, and believing that if such a thing were possible, I couldn't recognize it, Omega just has a tremendous ugh fiel... |
7896a264-9760-495e-9422-d6a2a9f37ef8 | trentmkelly/LessWrong-43k | LessWrong | Thoughts on the Feasibility of Prosaic AGI Alignment?
I’d like to preface by saying that I am not an expert on AI by any means, nor am I remotely involved with any kind of research or studies relevant to ML. I have no insight regarding any of the technical or mathematical aspects of discussions about this technology, ... |
3978c7d0-68eb-4bb4-95f2-e813233962fe | trentmkelly/LessWrong-43k | LessWrong | Free Educational and Research Resources
This is a list of free educational resources that are frequently overlooked or somewhat obscure, but still contain a large trove of information. The selection covers software, lessons, research papers, and reference material. I will skip covering online things that I think most ... |
03fa5839-b4f2-4ab5-b6c1-1515ab97b476 | trentmkelly/LessWrong-43k | LessWrong | Result Of The Bounty/Contest To Explain Infra-Bayes In The Language Of Game Theory
A month and a half ago I announced a $500 bounty/prize to explain infrabayes work in the language of game theory. Others added another $980 to the prize pool.
There were no entries which did exactly the thing I asked for, but we did g... |
c9735365-481d-4616-82cb-2e23871b37d6 | StampyAI/alignment-research-dataset/special_docs | Other | Reinforcement Learning in Newcomblike Environments
Reinforcement Learning in Newcomblike
Environments
James Bell
The Alan Turing Institute
London, UK
jbell@posteo.comLinda Linsefors
Independent Researcher
linda.linsefors@gmail.com
Caspar Oesterheld
Department of Computer Science
Duke University
Durham, NC, USA
caspar.... |
3270f6a3-2724-468e-a4f0-f427ccac0afe | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Open Problems Related to Solomonoff Induction
Solomonoff Induction seems clearly "on the right track", but there are a number of problems with it that I've been puzzling over for several years and have not made much progress on. I think I've talked about all of them in various comments in the past, but never collected... |
24697fee-50c4-4b64-89a0-0750db339b69 | trentmkelly/LessWrong-43k | LessWrong | Self-Control of LLM Behaviors by Compressing Suffix Gradient into Prefix Controller
In this paper, we are trying to control model behaviors. For example, by asking saying "You hear someone making fun of a topic you're passionate about", we can control an LLM to behave in an angrier manner. We can also control "any" be... |
fff31d67-5052-4065-9842-2f88929c243b | trentmkelly/LessWrong-43k | LessWrong | Basin broadness depends on the size and number of orthogonal features
TL;DR
For neural networks trained to perfect loss, the broadness of optima in parameter space is given by the number and norm of independent orthogonal features the neural network has. The inner product that defines this "feature norm" and independe... |
7de42041-19df-413d-9641-ef6a38b41b47 | trentmkelly/LessWrong-43k | LessWrong | Trying to approximate Statistical Models as Scoring Tables
For my PhD research, I have been tinkering with the idea of approximating complex, non-linear statistical models (Bayesian networks) as simple, quasi-linear models.
I am making some headway into the problem, but I feel I need yet to clarify what exactly I wa... |
ddce863a-2356-4ea0-a606-7d4b77289c5a | StampyAI/alignment-research-dataset/special_docs | Other | Fast and Easy Infinitely Wide Networks with Neural Tangents
Posted by Samuel S. Schoenholz, Senior Research Scientist and Roman Novak, Research Engineer, Google Research
The widespread success of deep learning across a range of domains such as [natural language processing](https://ai.googleblog.com/2020/02/explor... |
e8ca1c17-59c5-424f-9399-71d0c8467965 | trentmkelly/LessWrong-43k | LessWrong | Some Things That Increase Blood Flow to the Brain
Epistemic status: very shallow google scholar dive. Intended mostly as trailheads for people to follow up on on their own.
previously: https://www.lesswrong.com/posts/h6kChrecznGD4ikqv/increasing-iq-is-trivial
I don't know to what degree this will wind up being a con... |
2e3a2452-6c0e-4160-954a-8e9f888eb3c7 | trentmkelly/LessWrong-43k | LessWrong | Meetup : West LA Meetup 09-06-2011
Discussion article for the meetup : West LA Meetup 09-06-2011
WHEN: 06 September 2011 07:00:00PM (-0700)
WHERE: 10800 West Pico Blvd, Suite 312, Los Angeles, CA 90064
When: 7pm - 9pm September 6th
Where: The Westside Pavillion - on the bridge, which connects Nordstrom 3rd floor w... |
c6601904-ddc9-48af-b882-0a725507c085 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Short-Term AI Alignment as a Priority Cause
In this post, I will argue that *short-term* AI alignment should be viewed as today's greatest priority cause, whether you are concerned by long-term AGI risks or not.
To do so, I will first insist on the fact that AIs are automating *information collection, storage, analys... |
696127f3-c157-45ff-a225-3c9bd4f3fd08 | trentmkelly/LessWrong-43k | LessWrong | ‘Crucial Considerations and Wise Philanthropy’, by Nick Bostrom
|
d800623f-d9a4-49e0-b365-eb0250d756ef | trentmkelly/LessWrong-43k | LessWrong | Anatomy of a Gear
Thankyou to Sisi Cheng (of the Working as Intended comic) for the excellent drawings.
When modelling a physical gearbox, why do we chunk the system into gears? We’re essentially treating each gear as a black box, a discrete component without internal structure of its own, interesting mainly for its ... |
6bcff347-be5b-4906-a4b3-9bda0d3fcb83 | trentmkelly/LessWrong-43k | LessWrong | A Solution to the Desire-Action Conundrum
Note: This post will contain a good amount of philosophical ideas. It has been sectioned for ease.
1. What is the desire-action conundrum? Let's begin by looking at Buddhism's Four Noble Truths:
> * The truth of suffering. Life will almost always contain suffering
> * Th... |
fe5541de-9fe8-42db-ba8c-33a709ea4bc5 | trentmkelly/LessWrong-43k | LessWrong | One-person Universe
One day, professor Wheeler from Princeton called Richard Feynman and exclaimed: "Feynman, I know why all electrons have the same charge and the same mass!"
"Why?" asked Feynman.
"Because they are all the same electron, travelling back and forth in time!"
Upon hearing this story, Zarathustra said... |
1d3ef504-8966-49d8-a6d6-254cc8cb0437 | trentmkelly/LessWrong-43k | LessWrong | [Link] A Darwinian Response to Sam Harris’s Moral Landscape Challenge
I noticed that there has been some earlier discussion about Sam Harris’s Moral Landscape Challenge here at LW. As a writer on the Swedish politico-philosophical blog The Inverted Fable of Reality, I would like to share a response to the challenge, w... |
f8a02c9c-4378-44b2-812d-a09a13f29ac1 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | AISafety.world is a map of the AIS ecosystem
WHERE: National Portrait Gallery, Washington, DC 20001, USA (courtyard)
We'll be meeting to talk about books we've found interesting, exchange books (so bring some!) and decide i... |
a31bf108-bae8-4f73-9503-b7b7e6e397a3 | trentmkelly/LessWrong-43k | LessWrong | "Nahh, that wouldn't work"
After having it recommended to me for the fifth time, I finally read through Harry Potter and the Methods of Rationality. It didn't seem like it'd be interesting to me, but I was really mistaken. It's fantastic.
One thing I noticed is that Harry threatens people a lot. My initial reaction w... |
0ebeb32d-1dc7-43ee-90f9-afcc522c5301 | trentmkelly/LessWrong-43k | LessWrong | Guardian coverage of the Summit [link]
http://www.guardian.co.uk/commentisfree/belief/2011/nov/02/ai-gods-singularity-artificial-intelligence?newsfeed=true |
06bcdceb-9d31-4924-b395-ead47da330b1 | trentmkelly/LessWrong-43k | LessWrong | Dealing with a Major Personal Crisis
This is the earlier promised post about Dealing with a Major Personal Crisis. Please continue reading there but comment here.
The reasons for posting it this way are explained at the end of the link. I hope this approach does what I want it to.
|
a8541650-8778-4d8a-9176-5d0e90853651 | trentmkelly/LessWrong-43k | LessWrong | Open Thread, September 30 - October 6, 2013
If it's worth saying, but not worth its own post (even in Discussion), then it goes here. |
2d184298-186f-4994-8e94-e9447737d709 | trentmkelly/LessWrong-43k | LessWrong | Forecasting Newsletter: Looking back at 2021
Table of contents
* The American Empire Has Alzheimer's.
* Prediction Markets: VC money, searching for DraftKings, predatory pricing, and the race to be last.
* For skilled forecasters, crypto prediction markets are much more profitable than forecasting platforms.
* Bes... |
be813904-95b0-4eec-a915-1945ffbf08a3 | trentmkelly/LessWrong-43k | LessWrong | Ability to solve long-horizon tasks correlates with wanting things in the behaviorist sense
Status: Vague, sorry. The point seems almost tautological to me, and yet also seems like the correct answer to the people going around saying “LLMs turned out to be not very want-y, when are the people who expected 'agents' goi... |
ab8497f1-74af-4097-ac68-d0a294182190 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Navigating public AI x-risk hype while pursuing technical solutions
Public attention on AI x-risk has skyrocketed. I don’t expect this to wane anytime soon. This has a few potentially negative implications for those of us pursuing technical solutions to the problem:
Research Direction Incentives
---------------------... |
1f706bfd-5d0f-42a3-be41-2579d87400c2 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | [AN #107]: The convergent instrumental subgoals of goal-directed agents
**Newsletter #107**
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... |
414f9cde-6f07-4661-9afd-23c59466981d | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Clarifying “What failure looks like”
*Thanks to Jess Whittlestone, Daniel Eth, Shahar Avin, Rose Hadshar, Eliana Lorch, Alexis Carlier, Flo Dorner, Kwan Yee Ng, Lewis Hammond, Phil Trammell and Jenny Xiao for valuable conversations, feedback and other support. I am especially grateful to Jess Whittlestone for long con... |
a06d194a-890e-4c63-911f-471881a02a7f | trentmkelly/LessWrong-43k | LessWrong | What are you working on? June 2012
This is the bimonthly 'What are you working On?' thread. Previous threads are here. So here's the question:
What are you working on?
Here are some guidelines:
* Focus on projects that you have recently made progress on, not projects that you're thinking about doing but haven't s... |
34dd5a0d-6a2d-46fe-a6ab-da614cf9ee81 | trentmkelly/LessWrong-43k | LessWrong | Why Save The Drowning Child: Ethics Vs Theory
(cross-posted on the EA Forum)
A consequentialist, a deontologist, and a virtue ethicist are walking through a forest discussing the fundamental nature of morality. So heated is their disagreement that they startle a nearby small child, who falls into a nearby small pond.... |
fda10ba8-3c5f-42ff-a06d-a52a6edf91b4 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Saint-Petersburg: Game Event
Discussion article for the meetup : Saint-Petersburg: Game Event
WHEN: 24 November 2013 04:00:00PM (+0400)
WHERE: Санкт-Петербург, м. Технологический Институт, ул.1-я Красноармейская, дом 15
Join us for evening of games from lesswrong! For our third meetup we will try out new f... |
14aa3441-9546-48ae-8762-d58832087826 | trentmkelly/LessWrong-43k | LessWrong | Debunking komponisto on Amanda Knox (long)
Rebuttal to: The Amanda Knox Test
If you don't care about Amanda Knox's guilt, or whether you have received unreliable information on the subject from komponisto's post, stop reading now.
[Edit: Let me note that, generally, I agree that discussion of current events should b... |
ddc43471-a61e-410f-90e1-534e3884e97f | trentmkelly/LessWrong-43k | LessWrong | When to join a respectability cascade
This is a response to Scott Alexanders Give Up Seventy Percent Of The Way Through The Hyperstitious Slur Cascade. Maybe read that post first.
I think it becomes more important to struggle against a cascade the more annoying the new word is. There are a variety of factors that co... |
9267f229-0c3d-40d5-be01-58f726564bd0 | StampyAI/alignment-research-dataset/arxiv | Arxiv | Playing Atari with Deep Reinforcement Learning
1 Introduction
---------------
Learning to control agents directly from high-dimensional sensory inputs like vision and speech is one of the long-standing challenges of reinforcement learning (RL).
Most successful RL applications that operate on these domains have reli... |
86ae9fa7-8835-483e-a741-248ff308e8e6 | trentmkelly/LessWrong-43k | LessWrong | How useful for alignment-relevant work are AIs with short-term goals? (Section 2.2.4.3 of "Scheming AIs")
This is Section 2.2.4.3 of my report “Scheming AIs: Will AIs fake alignment during training in order to get power?”. There’s also a summary of the full report here (audio here). The summary covers most of the main... |
0d5cbbf3-b7f4-4da3-9b58-c289295c1018 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Warsaw: Newcomb's Problem & Decision Theories
Discussion article for the meetup : Warsaw: Newcomb's Problem & Decision Theories
WHEN: 30 May 2015 06:30:00PM (+0200)
WHERE: Warsaw
Time: 30 May (Saturday) at 18:30 Place: Cafe Kulturalna, Plac Defilad 1, PKiN, 00-901 Warsaw, Poland
Topic: Newcomb's Problem &... |
94993e86-3d24-49bc-bb36-2148a8b93578 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Was Homer a stochastic parrot? Meaning in literary texts and LLMs
*Cross posted from* [*New Savanna*](https://new-savanna.blogspot.com/2023/04/was-homer-stochastic-parrot-meaning-in.html)*.*
The phrase “stochastic parrot” was coined, I believe, by Emily Bender, a computational linguist and one of the coauthors of th... |
747509b6-900c-4404-933d-c2e094832776 | trentmkelly/LessWrong-43k | LessWrong | [Meetup] NYC Megameetup - Location and Potluck Info
This Saturday, June 23 is the Less Wrong Summer Festival Megameetup. Anyone who feels like coming to New York City is invited to attend. The official meetup description is here.
So far the forecast for Saturday is sunny, so the event will be held outdoors in Central... |
52df6812-c99e-42e8-bea6-49758f5b327e | trentmkelly/LessWrong-43k | LessWrong | Amazon KDP AI content guidelines
Amazon updated the content guidelines for Kindle Direct Publishing to address AI and in the process attempted to define terms AI-generated and AI-assisted. Guidelines like this are how our society tries to engage with AI in the near-term.
> Artificial intelligence (AI) content (text,... |
6655a5c1-4707-4e91-bd95-a43d2b9df02c | trentmkelly/LessWrong-43k | LessWrong | Think in Terms of Actions, not Attributes
How do we model the actions of those around us? This problem has been crucial for human survival since before the Holocene, so the brain has developed clever tricks to solve it. Personal attributes emerged as one of these tricks, as reflected in language: Bob is reliable, whil... |
b5a76f5d-42b0-4ed2-bc12-ae77bc137c8d | LDJnr/LessWrong-Amplify-Instruct | LessWrong | "I remember the exact moment when I began my journey as a rationalist. It was not while reading Surely You’re Joking, Mr. Feynman or any existing work upon rationality; for these I simply accepted as obvious. The journey begins when you see a great flaw in your existing art, and discover a drive to improve, to create n... |
2ab41f86-5d55-4188-a6d2-a72d3e531484 | trentmkelly/LessWrong-43k | LessWrong | Suggestions for chess puzzles
(You can sign up here if you haven't already.)
As suggested by Richard Willis in the comments, I want to present the "player As" with a series of individual isolated chess puzzles, for which the advisors will anonymously present their suggestions. However, there are constraints on what s... |
8eb4a8e2-7aaf-403c-b9af-a87dffc41f8c | trentmkelly/LessWrong-43k | LessWrong | An information-theoretic study of lying in LLMs
We recently published a workshop paper at ICML 2024 which we want to summarize here briefly. Code for the paper can be found on GitHub. This work was done as part of the PIBBSS affiliateship. Thx to @Jayjay for comments on the draft.
TLDR: We analyze the probability dis... |
e4951538-c8d6-4687-9aa8-41d6596a580a | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Why libertarians are advocating for regulation on AI
Motivation: some people on the internet seem confused about why libertarians (and others generally suspicious of government intervention) are advocating for regulation to help prevent x-risk from AI. The arguments here aren't novel and generally seem pretty obvious... |
0dbd5518-953a-4c11-88e0-a57c67b55398 | trentmkelly/LessWrong-43k | LessWrong | The Darwin Pregame
Epistemic Status: True story
This is intended as post two of the sequence Zbybpu’f Nezl.
Previously (required): The Darwin Game
I
This is my reconstruction of my thoughts at the time.
The Darwin Game requires surviving the early, middle and late games.
In the opening, you need to maximize scor... |
5cd939af-1caa-4032-960b-b6411ee319e6 | trentmkelly/LessWrong-43k | LessWrong | Basic Conversational Coordination: Micro-coordination of Intention
Follow up to: Something simple to try in conversations
In this essay, I want to describe a simple, but important, abstraction for how to think about what’s happening in a conversation, and how to make use of that abstraction for helping a conversatio... |
f6b04b53-6ba2-46fe-ad36-f765aba25099 | trentmkelly/LessWrong-43k | LessWrong | Negentropy Overrated?
This post may be interesting to some LWers.
In summary: it looks like our universe can support reversible computers which don't create entropy. Reversible computers can simulate irreversible computers, with pretty mild time and space blowup. So if moral value comes from computation, negentropy p... |
7a0c5404-0abf-454f-aa13-553539af67b2 | trentmkelly/LessWrong-43k | LessWrong | Bringing Little Kids to Contra Dances
Julia and I brought Lily to a contra dance for the first time when she was just a couple weeks old, and have continued bringing her (and now her younger siblings) occasionally as they've gotten older. A friend asked if I could write up something about the logistics of this, so her... |
17890dfc-7d62-417e-aacb-d65690395b77 | StampyAI/alignment-research-dataset/arxiv | Arxiv | MEMO: Test Time Robustness via Adaptation and Augmentation
1 Introduction
---------------
Deep neural network models have achieved excellent performance on many machine learning problems, such as image classification, but are often brittle and susceptible to issues stemming from *distribution shift*. For example, d... |
43b61d62-c15b-4a19-aab8-0422fc1830a8 | trentmkelly/LessWrong-43k | LessWrong | Don't want Goodhart? — Specify the damn variables
> In everyday life, if something looks good to a human, then it is probably actually good (i.e. that human would still think it’s good if they had more complete information and understanding). Obviously there are plenty of exceptions to this, but it works most of the t... |
4ba847f9-4653-4261-9d01-b71d26d7051a | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Strategies for keeping AIs narrow in the short term
*Disclaimer: I don’t have any particular expertise in AI safety, I just had some thoughts and this seemed like the place to put them.*
The [bleak outlook](lesswrong.com/posts/j9Q8bRmwCgXRYAgcJ/miri-announces-new-death-with-dignity-strategy) that Eliezer presents he... |
1be0fa83-0c2a-47fc-806a-05e8755b01c7 | trentmkelly/LessWrong-43k | LessWrong | Suggestions for 31C3 (Chaos Communication Congress)
Hi, I will be attending 31C3, the 31th chaos communication congress in Hamburg today and the next days. For those who don't know about this hacker conference, here is a summary:
> The 31st Chaos Communication Congress (31C3) is an annual four-day conference on techn... |
e1348c37-dacf-45cf-8af9-405cf37207d4 | StampyAI/alignment-research-dataset/blogs | Blogs | making the UD and UDASSA less broken: identifying time steps
making the UD and UDASSA less broken: identifying time steps
------------------------------------------------------------
the [universal distribution](https://www.lesswrong.com/posts/XiWKmFkpGbDTcsSu4/on-the-universal-distribution) ("UD") and [its applicab... |
71965307-fb4b-4053-b201-e197dbcc18a7 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | "Ray Kurzweil and Uploading: Just Say No!", Nick Agar
A new paper has gone up in the November 2011 JET: ["](http://jetpress.org/v22/agar.htm)[Ray Kurzweil and Uploading: Just Say No!"](http://jetpress.org/v22/agar.htm) ([videos](https://whatsortsofpeople.wordpress.com/tag/nick-agar/)) by [Nick Agar](http://www.victori... |
1b209ef2-3418-4539-9c42-775f4375622f | trentmkelly/LessWrong-43k | LessWrong | Can science come to understand consciousness? A problem of philosophical zombies (Yes, I know, P-zombies again.)
In response to the classic Mysterious Answers to Mysterious Questions, I express some skepticism that consciousness is can be understood by science. I postulate (with low confidence) that consciousness is “... |
a2b5bcfb-4299-4aa9-b372-78771ee49bee | StampyAI/alignment-research-dataset/arxiv | Arxiv | Exploiting the Rule Structure for Decision Making within the Independent Choice Logic
454
Exploiting the Rule Structure for Decision Making within the Independent
Choice Logic
David Poole*
Department of Computer Science
University of British Columbia
Vancouver, B.C., Canada V6T 1Z4
poole@cs.ubc.ca
http:// www.... |
8e34da13-dee8-4ade-a517-a6945096703c | trentmkelly/LessWrong-43k | LessWrong | Reforming rot13
rot13.com is a service frequently used here to hide spoilers. I really hate it, though. If I had the time, I would build a simple, but much better alternative. Maybe somebody has more time to do that, so I'll share my rough specification:
* Dynamic. The encoding happens on the client automatically, w... |
615ba297-2e9d-454c-b172-71891e6d83c6 | trentmkelly/LessWrong-43k | LessWrong | Relating Modal Polymorphism to PA with soundness
|
412ea277-3655-4f74-bd17-9b55ca9b0467 | StampyAI/alignment-research-dataset/blogs | Blogs | Paul Christiano on the safety of future AI systems
*By Asya Bergal, 11 September 2019*
[](http://aiimpacts.org/wp-content/uploads/2019/09/paulchristiano.jpeg)Paul Christiano
As part of our AI optimism project, we talked to Paul Ch... |
171e8ceb-9219-460b-9e72-300b173e20b3 | trentmkelly/LessWrong-43k | LessWrong | Estrogen: A trip report
The following post discusses my personal experience of the phenomenology of feminising hormone therapy. It will also touch upon my own experience of gender dysphoria.
I wish to be clear that I do not believe that someone should have to demonstrate that they experience gender dysphoria – howeve... |
37f2d407-366f-4b7c-816b-69d5ab3c3913 | trentmkelly/LessWrong-43k | LessWrong | Harry Potter and the Methods of Rationality predictions
The recent spate of updates has reminded me that while each chapter is enjoyable, the approaching end of MoR, as awesome as it no doubt will be, also means the end of our ability to learn from predicting the truth of the MoR-verse and its future.
With that in mi... |
c4175f68-879c-44a6-a64e-77c04efe2396 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | The Alignment Newsletter #8: 05/28/18
**Highlights**
--------------
**[Solving the Rubik's Cube Without Human Knowledge](https://arxiv.org/abs/1805.07470)** *(Stephen McAleer, Forest Agostinelli, Alexander Shmakov et al)*: This paper proposes *Autodidactic Iteration* (ADI), which is a technique that can be combined w... |
d40c2bdb-2503-4d09-a661-499ca17ca336 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | A Gym Gridworld Environment for the Treacherous Turn
**EDIT**: posted here for feedback and discussion. I plan to continue working on different models/environments, so feel free to suggest improvements.
(**tl;dr**: In an attempt to better understand the treacherous turn, I created a gridworld environment where an ag... |
682602be-670f-46f1-9add-681c9dd390c6 | trentmkelly/LessWrong-43k | LessWrong | Wiki Quote Attributions
After an interesting discussion the last time this topic came up, I wanted people's input on the best way to attribute a quote to the LessWrong Wiki. The passage is obviously based on Eliezer's work, but he himself didn't write the summary. From the quote logs, it was Z. M. Davis who originally... |
d63bbc82-1f3e-4a85-bac3-38a187b04d50 | trentmkelly/LessWrong-43k | LessWrong | The State of the Art of Scientific Research on Polyamoury
The idea of polyamoury is one that interests me. However, while such books as The Ethical Slut have done a good job of providing me with tools to understand and possibly handle the challenges and rewards involved, I found them unsatisfying in that they were lar... |
0dd5a119-4b12-4ed0-be68-9c86fa77ea08 | trentmkelly/LessWrong-43k | LessWrong | Alignment 201 curriculum
This curriculum, a follow-up to the Alignment Fundamentals curriculum (the ‘101’ to this 201 curriculum), aims to give participants enough knowledge about alignment to understand the frontier of current research discussions. It assumes that participants have read through the Alignment Fundamen... |
2b016ca2-e044-4744-a945-32cc2154eb8c | trentmkelly/LessWrong-43k | LessWrong | September Less Wrong Meetup aka Eliezer's Bayesian Birthday Bash
In honor of Eliezer's Birthday, there will be a Less Wrong meetup at 5:00PM this Saturday, the 11th of September, at the SIAI Visiting Fellows house (3755 Benton St, Santa Clara CA). Come meet Eliezer and your fellow Less Wrong / Overcoming Bias members... |
6f2416bd-830e-4ee7-8c0c-b0b8e1794ce9 | trentmkelly/LessWrong-43k | LessWrong | The origins of the steam engine: An essay with interactive animated diagrams
This is a guest post written by Anton Howes and animated by Matt Brown of Extraordinary Facility. This project was sponsored by The Roots of Progress, with funding generously provided by The Institute:
> Steam power did not begin with the st... |
aadf0f28-e03f-48cd-b73b-f69789b02da5 | trentmkelly/LessWrong-43k | LessWrong | A brainteaser for language models
I came up with the following puzzle the other day:
Q: Solve the puzzle: 63 = x = 65536
A: x =
The intended answer is in the form of a number.
text-davinci-003 guesses my intended answer at 11.8% probability, which is the second-highest probability for any answer.
(This is somew... |
fa14a397-d080-4285-9674-62cdcb268a51 | trentmkelly/LessWrong-43k | LessWrong | Avoiding Moral Fads?
(content warning: murder, medical abuse)
A few years ago, my old middle school and another school in its district were renamed; they had originally been named after once-esteemed educational figures who were now in disrepute thanks in large part to their links to eugenics. I believe at the time I... |
d6fd6ef2-ee62-4ba9-aec2-871c58259ffc | trentmkelly/LessWrong-43k | LessWrong | Meetup Tip: The Next Meetup Will Be. . .
Summary
One of the most useful things to tell people at a meetup is when the next meetup will be. This requires you to know when the next meetup will be. If you don't know when the next meetup will be then the next best equivalent is telling people how you will announce the nex... |
fe37f888-8157-4bca-96b1-1d27f8db5f09 | StampyAI/alignment-research-dataset/arxiv | Arxiv | UbuntuWorld 1.0 LTS - A Platform for Automated Problem Solving & Troubleshooting in the Ubuntu OS
UbuntuWorld 1.0 LTS
-------------------
The main components of the proposed system (Figure [3](#S0.F3 "Figure 3 ‣ Motivation ‣ UbuntuWorld 1.0 LTS - A Platform for Automated Problem Solving & Troubleshooting in the Ubu... |
52ce947d-c1a5-4f42-b17a-01cdad5fd1eb | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Four mindset disagreements behind existential risk disagreements in ML
I've previously written that good ideas and conversations about AGI seem to have propagated through ML [weirdly slowly](https://www.lesswrong.com/posts/Rkxj7TFxhbm59AKJh/the-inordinately-slow-spread-of-good-agi-conversations-in-ml).
A different we... |
b243b604-ed1b-4f5c-a308-21888ac4e577 | trentmkelly/LessWrong-43k | LessWrong | Let's split the cake, lengthwise, upwise and slantwise
This post looks at some of the current models for how two agents can split the gain in non-zero sum interactions. For instance, if you and a buddy have to split £10 between the two of you, where the money is discarded if you can't reach a deal. Or there is an oppo... |
f85e8ba0-b593-4f8e-8558-e203a4bc04a8 | trentmkelly/LessWrong-43k | LessWrong | Improving Cryonics - Regulations and Ethical Considerations
Here is my understanding - correct me if I'm wrong:
Cryonics is only allowed once a person is determined legally dead: when the heart stops beating.
One of the reasons why they have to be dead seems to be that the majority of the population consider cryonic... |
a408b1bd-75fb-49aa-b635-a6e77d113631 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | NTIA - AI Accountability Announcement
The Department of Commerce is announcing a new policy on AI Accountability today, with an event at 12pm EST, April 11th.
The linked page also includes a link to the new policy.
There is an opportunity to leave comments on the policy until June 10th. |
2e44a6da-558f-4e29-8c6f-e0ecddcb320d | trentmkelly/LessWrong-43k | LessWrong | Why I Am Changing My Mind About AI Risk
|
77e45c8d-729c-42ba-8c77-40f15b939c21 | StampyAI/alignment-research-dataset/blogs | Blogs | Will AI see sudden progress?
*By Katja Grace, 24 February 2018*
Will advanced AI let some small group of people or AI systems take over the world?
AI X-risk folks and others have accrued lots of arguments about this over the years, but I think this debate has been disappointing in terms of anyone changing anyone e... |
b3023175-319a-4e01-baf7-f6a6ea1b8753 | trentmkelly/LessWrong-43k | LessWrong | Coordinate-Free Interpretability Theory
Some interpretability work assigns meaning to activations of individual neurons or small groups of neurons. Some interpretability work assigns meaning to directions in activation-space. These are two different ontologies through which to view a net’s internals. Probably neither ... |
5d94f8c3-fb2a-4a70-bc71-4de07993999e | StampyAI/alignment-research-dataset/special_docs | Other | AI, Governance Displacement, and the (De)Fragmentation of International Law
of AI's effects on the legal and institutional scaffolding of the international legal order. How can we analyse the direct and indirect effects of AI technologies on the viability, form, or functioning of the international legal order? This pa... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.