id stringlengths 36 36 | source stringclasses 15
values | formatted_source stringclasses 13
values | text stringlengths 2 7.55M |
|---|---|---|---|
0d251aa2-8e1f-4e67-839c-468c41b84e8a | trentmkelly/LessWrong-43k | LessWrong | Conditioning Predictive Models: Outer alignment via careful conditioning
This is the second of seven posts in the Conditioning Predictive Models Sequence based on the paper “Conditioning Predictive Models: Risks and Strategies” by Evan Hubinger, Adam Jermyn, Johannes Treutlein, Rubi Hudson, and Kate Woolverton. Each p... |
76a2cb41-791b-4c70-b5b8-d6accd77ced4 | trentmkelly/LessWrong-43k | LessWrong | Clothing is Hard (A Brief Adventure into my Inefficient Brain)
This is an active solicitation for suggestions on how to train it differently.
Apparently, this morning I put on my underwear wrong.
Upon noticing that they were on incorrectly, I took them off by turning them inside out on the Z axis (top of head t... |
a004c3ea-8632-4d72-9b1b-da7d564ef8d3 | trentmkelly/LessWrong-43k | LessWrong | AI-kills-everyone scenarios require robotic infrastructure, but not necessarily nanotech
Tl;DR: Nanotech seems the only option for AI to kill everybody; but we can imagine a prosaic-AI-killing-everyone using three steps: taking over a country, building robotic infrastructure, releasing biological weapons.
Arguments ... |
8671c984-8ad4-4481-ad47-105972c69963 | trentmkelly/LessWrong-43k | LessWrong | XAI releases Grok base model
> We are releasing the base model weights and network architecture of Grok-1, our large language model. Grok-1 is a 314 billion parameter Mixture-of-Experts model trained from scratch by xAI.
>
> This is the raw base model checkpoint from the Grok-1 pre-training phase, which concluded in ... |
16a3ecc4-cc6c-4bcf-b5c5-33b67301711b | trentmkelly/LessWrong-43k | LessWrong | Money threshold Trigger Action Patterns
In American society, talking about money is a taboo. It is ok to talk about how much money someone else made when they sold their company, or how much money you would like to earn yearly if you got a raise, but in many different ways, talking about money is likely to trigger som... |
4a256856-3c8c-460f-82df-54bcf03a7933 | trentmkelly/LessWrong-43k | LessWrong | Look Under the Light Post
There's a well-worn story you've probably heard before about looking for keys under a light post. If you've not heard it before:
> A man leaves the bar after a long night of drinking. He stumbles all the way home only to find when he gets there that he doesn't have his keys. He knows he had ... |
a23bee5b-3dc5-4fcb-b5f5-0a7f734fa6a3 | trentmkelly/LessWrong-43k | LessWrong | Finance - accessibility (asking for textbook recommendations)
> Finance is obfuscated, not made accessible, for it is why professionals are paid. Barriers of entry are intentional
Anonymous finance professional
Context - While debating the legitimacy and usefulness of industry-knowledge barriers of entry is arguably... |
c2ccbd13-9718-4588-8f7c-d386f986bbbe | trentmkelly/LessWrong-43k | LessWrong | pplapi is a virtual database of the entire human population.
|
75213417-a601-44c3-93d2-60a14c05bf22 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Potential Alignment mental tool: Keeping track of the types
*Epistemic status: Came up with what I was thinking of as a notation. Realised that there was a potentially useful conceptual tool behind it.*
Consider this diagram.
 try to familiarize yourself with the sizes of various things using e.g. spaced repetition and (2) c... |
9df57d9a-4793-4fa7-9984-6d4d9fb2b45b | trentmkelly/LessWrong-43k | LessWrong | Low Hanging Fruit- Basic bedroom decorating
This is a weird topic for a discussion post. Why are you writing it?
In a recent post-mini-camp skills exchange thread, I jokingly mentioned that I could "help make your room look Awesome!" To my surprise, a couple people were actually interested. Note: I'm not an interior ... |
772285d7-1f31-4876-a3e6-2e6e251041e3 | trentmkelly/LessWrong-43k | LessWrong | What question would you like to collaborate on?
In Why don't long running conversations happen on LessWrong? adamzerner writes:
> Here is how things currently work:
>
> * Someone writes a post.
> * It lingers around the front page for a few days. During this time, conversations emerge in the comments section.
> ... |
0937ba07-2b1a-4439-aad6-d0da710eb098 | StampyAI/alignment-research-dataset/arxiv | Arxiv | A Brandom-ian view of Reinforcement Learning towards strong-AI
1 Introduction
---------------
The analytic philosophy of Robert Brandom situates itself with the demarcation question: what does it mean to be us? In the process, Brandom presents a layer-cake picture of sapience, thus categorizing agents into one of t... |
0e1a8328-69b7-4c34-beb0-4f1874d994b8 | trentmkelly/LessWrong-43k | LessWrong | Countering Self-Deception: When Decoupling, When Decontextualizing?
What's the difference, conceptually, between these three noticings, if any?
> "I'm only enjoying that wine because of what it signals"
>
> "I'm only enjoying that food because I know it's organic"
>
> "I'm only enjoying that movie scene because I k... |
e8ddb1b3-dbcd-413d-a402-cf9e4f88a3a4 | trentmkelly/LessWrong-43k | LessWrong | Morality in Fantasy Worlds
This comment by Carinthium in my earlier post has got me to think, and my apologies if this subject was raised before - couldn't find it.
> 1- In a hypothetical world where God actually existed, it would be an OBJECTIVE belief and thus it would be worth saving people. (Just as with people k... |
9dd421a0-8855-42e6-a52e-b14342897725 | trentmkelly/LessWrong-43k | LessWrong | Missed Distinctions
When we lump unlike things together, it confuses us and opens holes in our theories. I'm not normally one to read about diets, dieting advice, or anything of that sort, but in today's article about the Shangri-La Diet, I saw an important distinction that no one's talked about. Something Eliezer sai... |
1213cfca-27bc-4b5b-8c0e-f0ec207972d3 | trentmkelly/LessWrong-43k | LessWrong | The Deep Lore of LightHaven, with Oliver Habryka (TBC episode 228)
This is a link to the latest Bayesian Conspiracy episode. Oliver tells us how Less Wrong instantiated itself into physical reality via LightHaven, along with a bit of deep lore of foundational Rationalist/EA orgs. He also gives a surprisingly nuanced (... |
4afe25cf-d0c1-4b4e-a6ea-5aef812c7e30 | trentmkelly/LessWrong-43k | LessWrong | Brief Query- An Idea
I had an idea- has it been done before, and if not shouldn't somebody try to do it? I live in Melbourne where LWers aren't organised, but perhaps the New York branch or some other group could try this? (IF it hasn't been done, that is- I haven't seen in mentioned, which is why I'm checking)
Idea:... |
7ac414ae-9602-4fab-ac47-6de24e448bda | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Evaluating the feasibility of SI's plan
(With Kaj Sotala)
SI's current R&D plan seems to go as follows:
1. Develop the perfect theory.
2. Implement this as a safe, working, Artificial General Intelligence -- and do so before anyone else builds an AGI.
The Singularity Institute is almost the only group w... |
4d2d6584-993e-40e5-84b4-71e3ff84fc81 | StampyAI/alignment-research-dataset/arbital | Arbital | Provability predicate
A provability predicate is a formula $P$ of a theory satisfying the Hilbert-Bernais derivability conditions. If the diagonal theorem is applicable in the theory as well, then [https://arbital.com/p/55w](https://arbital.com/p/55w) and [Gödel's second incompleteness theorem](https://arbital.com/p/)... |
d677f2b6-89e1-4990-b813-1abfa66a39e2 | trentmkelly/LessWrong-43k | LessWrong | AI for AI safety
(Audio version here (read by the author), or search for "Joe Carlsmith Audio" on your podcast app.
This is the fourth essay in a series that I’m calling “How do we solve the alignment problem?”. I’m hoping that the individual essays can be read fairly well on their own, but see this introduction for... |
9060356d-fc72-4fba-a944-3c5bc7f4940b | trentmkelly/LessWrong-43k | LessWrong | introduction to thermal conductivity and noise management
basics
Most people understand, to some extent, the principle of Fourier's law - that heat transfer is proportional to temperature difference. Most people reading this probably also understand triple-glazed windows, fiberglass insulation, and vacuum flasks, but:... |
035192f2-4138-45df-b9ec-df627cc182d8 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Interpretability’s Alignment-Solving Potential: Analysis of 7 Scenarios
*This is the second post in the sequence “Interpretability Research for the Most Important Century”. The first post, which introduces the sequence, defines several terms, and provides a comparison to existing works, can be found here:* [*Introduct... |
e24de49f-9ec8-4795-98fb-1d8373b90c45 | StampyAI/alignment-research-dataset/arxiv | Arxiv | WebGPT: Browser-assisted question-answering with human feedback
1 Introduction
---------------
A rising challenge in NLP is long-form question-answering (LFQA), in which a paragraph-length answer is generated in response to an open-ended question. LFQA systems have the potential to become one of the main ways peopl... |
3e26f556-12ad-4cfb-9d63-d54470913eb1 | trentmkelly/LessWrong-43k | LessWrong | What could small scale disasters from AI look like?
If an AI system or systems goes wrong in the near term and causes harm to humans in a way which is consistent with or supportive of alignment being a big deal, what might it look like?
I'm asking because I'm curious about potential fire-alarm scenarios (including th... |
76c5e67f-210d-4365-9ce7-66cb3506ec98 | trentmkelly/LessWrong-43k | LessWrong | Radical Probabilism [Transcript]
(Talk given on Sunday 21st June, over a zoom call with 40 attendees. Abram Demski is responsible for the talk, Ben Pace is responsible for the transcription)
Talk
Abram Demski: I want to talk about this idea that, for me, is an update from the logical induction result that came out o... |
562ffab3-f0d1-458b-a6cf-866751f2f5b2 | trentmkelly/LessWrong-43k | LessWrong | Notes on Wisdom
This post examines the virtue of wisdom. It mostly explores what other people have learned about this virtue, rather than highlighting my own opinions about it, though I’ve been selective about what I found interesting or credible, according to my own inclinations. I wrote this not as an expert, but as... |
f2948061-4ed8-4755-bd05-38ffc45caf69 | trentmkelly/LessWrong-43k | LessWrong | Less Anti-Dakka
It is written in More Dakka:
> If something is a good idea, you need a reason to not try doing more of it.
Taken at face value, it implies the following:
> If something is a bad idea, you need a reason to not try doing less of it.
Labels/concepts, such as More Dakka, Inadequate Equilibria, etc poin... |
7ac1f331-8565-4b8f-9eed-cab53492d0cf | trentmkelly/LessWrong-43k | LessWrong | Brain Preservation
Most people, given the option to halt aging and continue in good heath for centuries, would. Anti-aging research is popular, but medicine is only minimally increasing lifespan for healthy adults. You, I, and everyone we know have bodies that are incredibly unlikely to make it past 120. They're just ... |
7f877cca-b6c9-41f9-93e1-e675caebeb8c | trentmkelly/LessWrong-43k | LessWrong | High school activities and medical school admissions
Many people consider medicine to be a desirable profession because of the high status of the job of being a doctor, the opportunity to help people, and the high income. Whether conventional views about what it’s like to be in medicine are well grounded is unclear, b... |
18b01ae7-226d-44cb-9f71-88ced2fd52f7 | trentmkelly/LessWrong-43k | LessWrong | University of Edinburgh, Master's Artificial Intelligence
This article is part of a series of articles on different European master's programs related to artificial intelligence and machine learning.
Basic data on the degree
* Duration: 1 year (or 2 years part-time)
* Cost: ~15.000€ p.a. for EU students (before Br... |
9bfebb6e-4738-49d4-a8e5-8e17698b3d55 | trentmkelly/LessWrong-43k | LessWrong | Near-mode cryonics: A thought experiment
(x-post: LessDead)
Set-up: In a world that’s otherwise the same, an anthropomorphic God comes to you with a gun and a check, and offers you the following deal, should you accept it. (God can perfectly see the future and is honest.)
Dilemma: God will give you a check of X USD ... |
9bd21f92-b77c-42a3-ace7-804dd806dbe6 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Evaluating hidden directions on the utility dataset: classification, steering and removal
*Produced as part of the SERI ML Alignment Theory Scholars Program - Summer 2023 Cohort, under the mentorship of Dan Hendrycks*
We demonstrate different techniques for finding concept directions in hidden layers of an LLM. We pr... |
75235501-c3aa-4f45-8d9b-88d505ef80c0 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | AI Governance Course - Curriculum and Application
*[Edit: The course has received major updates since this post was made. **See [here](https://course.aisafetyfundamentals.com/governance) for the most recent course information and curriculum**. I've removed the largely outdated information from this post.]*
While man... |
866f20cb-01f0-4173-a9df-abd3969c92c9 | trentmkelly/LessWrong-43k | LessWrong | MIRI course list book reviews, part 1: Gödel, Escher, Bach
I'm Nate. I was recently introduced to the ideas of existential risk and unfriendly AI. I've decided to read the books suggested in the MIRI course list. I'll review the books as I read them. Repeating the knowledge is expected to help solidify it. Public acco... |
c883ea12-f721-423e-9696-6ea1648b4cbd | trentmkelly/LessWrong-43k | LessWrong | Musicians and Mouths
Most musicians massively underuse their mouths, and the ones that use their mouths underuse their fingers. This is very weird and I don't understand it.
The most common instruments I see are probably piano, guitar, drums, and violin. Add in mandolin, bass, accordion, cello, viola, and banjo, and ... |
0b20e82e-b449-4767-930d-8cbd23f247af | StampyAI/alignment-research-dataset/arbital | Arbital | Logarithm base 1
There is no [https://arbital.com/p/-3nd](https://arbital.com/p/-3nd) base 1, because no matter how many times you multiply 1 by 1, you get 1. If there _were_ a log base 1, it would send 1 to 0 (because $\log_b(1)=0$ for every $b$), and it would also send 1 to 1 (because $\log_b(b)=1$ for every $b$), w... |
36da573e-053f-4bbb-b75e-665eb3ddf115 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Towards a Formalisation of Logical Counterfactuals
**Introduction**
This is an attempt to build a basis for formalising logical counterfactuals. It's not workable as is, but I think it can potentially solve the problem of ["outthinking the counterfactual"](https://www.lesswrong.com/posts/q9DbfYfFzkotno9hG/example-dec... |
ee5c108f-0308-44e4-a15d-437be71d1e71 | trentmkelly/LessWrong-43k | LessWrong | Dominic Cummings: how the Brexit referendum was won
|
eb56ed45-ea58-4c7a-92e8-ea4ab057e0f6 | trentmkelly/LessWrong-43k | LessWrong | Wired Headsets for Video Calls
I've recently become a fan of wired headsets for video calls for two reasons: headsets are good at keeping out background noise, and wired is good for minimizing latency.
When I started working from home I would need to constantly mute and unmute myself on video calls. I'm working from ... |
87f5005b-975c-4433-8035-d8dc7304fa8c | awestover/filtering-for-misalignment | Redwood Research: Alek's Filtering Results | id: post3861
A putative new idea for AI control; index here . This post will use the formalism of this post to illustrate some well known philosophical thought experiments and show why learning algorithms are not sufficient to solve them. Examples Death and life extension A human consists of two agents, A 100 and A 1 ... |
5cbec9f4-517a-4bcb-afc2-d71d45ec1d28 | trentmkelly/LessWrong-43k | LessWrong | A Guide to AI 2027
This post aims to help readers navigate the recent AI 2027 website, which contains a plethora of links and references that encourage rabbit-holing on many different aspects of the scenario and associated forecasts.
If this is helpful in any way to readers of AI 2027, let me know. If you see someth... |
c5c50407-c948-4732-9768-92465b6ad854 | trentmkelly/LessWrong-43k | LessWrong | Seeking models of LW’s aversion to religion
Hi there. I’m a longtime LW-er posting under a pseudonym, in the hopes of being able to make my ignorance visible without unnecessary drama. Please don’t out me; it seems to me that the ability to expose one’s ignorance pseudonymously is a pro-epistemic norm that we’ll be be... |
602dd65a-c22f-4999-899d-5e6204731c20 | trentmkelly/LessWrong-43k | LessWrong | DARPA accepting proposals for explainable AI
"The XAI program will focus the development of multiple systems on addressing challenges problems in two areas: (1) machine learning problems to classify events of interest in heterogeneous, multimedia data; and (2) machine learning problems to construct decision policies f... |
c44c072e-d71f-4aef-8862-0b384dc45ede | trentmkelly/LessWrong-43k | LessWrong | Built-In Bundling For Faster Loading
I used to work on mod_pagespeed, which would rewrite web pages so they would load faster. A key component was bundling: combining multiple small resources into a single larger one to reduce overhead. For example, it had tools to combine multiple CSS, JS, and images.
There were ser... |
da043a68-5359-45bf-a522-7874cb7a54aa | trentmkelly/LessWrong-43k | LessWrong | Inference cost limits the impact of ever larger models
I sometimes notice that people in my community (myself included) assume that the first "generally human-level" model will lead to a transformative takeoff scenario almost immediately. The assumption seems to be that training is expensive but inference is cheap so ... |
9604c17f-2d6a-4ba8-aedd-5e08c524b4fd | trentmkelly/LessWrong-43k | LessWrong | Arrow grid game
There’s something I like about having different systems all the time. Apparently. |
eb575c99-1186-41bc-98b5-a8219d74dbcd | trentmkelly/LessWrong-43k | LessWrong | Dennett's "Consciousness Explained": Chpt 2
Chapter 2: Explaining Consciousness
This chapter is all about dualism and why it's bad. I find this chapter incoherent, because Dennett never defines dualism carefully, and confuses it with the theistic views historically associated with dualism.
[Long stretches of my own ... |
f0a15c02-b5d7-41e9-a03e-fca6cf0ff50b | trentmkelly/LessWrong-43k | LessWrong | Why *I* fail to act rationally
There is a lot of talk here about sophisticated rationality failures - priming, overconfidence, etc. etc. There is much less talk about what I think is the more common reason for people failing to act rationally in the real world - something that I think most people outside this communit... |
ca6d11a9-1d8a-4e7a-906a-d3013040be06 | trentmkelly/LessWrong-43k | LessWrong | What can we learn from Microsoft's Tay, its inflammatory tweets, and its shutdown?
http://www.wired.com/2016/03/fault-microsofts-teen-ai-turned-jerk/
Could this be a lesson for future AIs? The AI control problem?
[future AIs may be shutdown, and matyred..] |
acb349c0-b301-44ee-82a3-e5655e68b193 | trentmkelly/LessWrong-43k | LessWrong | Self-help, hard and soft
Summary: Four types of trying to be better.
Confidence: 90% that this beats the 0-D or 1-D idea. 50% that this is the best phrasing or 2D projection.
I get annoyed when I click a post and find that it's not a list of things to try, or a cool tool for reasoning or deciding. But I don't endors... |
ba43e570-326b-446a-a8f6-f2e235cfa58e | trentmkelly/LessWrong-43k | LessWrong | Pandemic Prediction Checklist: H5N1 (6/14)
H5N1 likely has spread among minks, the first major spread from mammal to mammal
Changelog available at bottom of post.
Here's what we can do about it:
Five ways to stop a bird->mink->human H5N1 pandemic
Overall score: 5/14
* Transmissibility: H5N1 is bird flu. It can j... |
98804ce7-a490-4e5e-aac1-3a4b09b93b9d | StampyAI/alignment-research-dataset/arxiv | Arxiv | Imitating Latent Policies from Observation
1 Introduction
---------------
Humans often learn from and develop experiences through mimicry. It is noteworthy that we can mirror behavior through only the observation of state trajectories without knowing the underlying actions (e.g., the exact kinematic forces) and int... |
e5deacf3-f1bb-4f65-96bf-3f911ca6d3e0 | trentmkelly/LessWrong-43k | LessWrong | Getting diagnosed for ADHD if I don't plan on taking meds?
Several people have told me that it is likely I have ADHD & I should go to a doctor and get diagnosed. At the same time, I have no desire to take medication. Given that this is the case, is there any point in going to a doctor & getting diagnosed. After all, ... |
291b2110-eb14-4977-8015-058c1fb2ca7d | trentmkelly/LessWrong-43k | LessWrong | Prediction challenge: Zika and birth rates
I've been wondering about good new topics for LW, and prediction might be one of them.
The effect of the Zika virus-- and human reactions to it-- on birth rates has the combination of being hard enough to be interesting, not being heavily plowed over by partisans, and having... |
3440dd90-dcfe-411b-bb04-71a24990b048 | StampyAI/alignment-research-dataset/distill | Distill Scientific Journal | The Paths Perspective on Value Learning
Introduction
------------
In the last few years, reinforcement learning (RL) has made remarkable progress, including [beating world-champion Go players](https://deepmind.com/research/alphago/), [controlling robotic hands](https://blog.openai.com/learning-dexterity/), and eve... |
df2584bb-f7a3-468d-869f-a0268e7f6586 | trentmkelly/LessWrong-43k | LessWrong | Raymond Smullyan has died
|
dff0c25b-b184-466d-b74d-3761f38fb116 | trentmkelly/LessWrong-43k | LessWrong | Yet another failed utopia
A DeepMind-like AI is trained and executed on a decentralized supercomputer like the Golem network, perhaps one that also permits microtransactions like the IOTA Tangle.
The algorithm collects texts posted by citizens on public fora and processes them by topic using sentiment analysis algori... |
8f730656-5da8-4791-9d9d-9525698a762f | trentmkelly/LessWrong-43k | LessWrong | 2018 Prediction Contest - Propositions Needed
Summary: I think it is worth running a contest to measure ability to make accurate predictions, and am prepared to put up a prize of $200 for the winner. I need to identify a base of 20-50 propositions, all likely to settle within 2-10 months from now, and would like to as... |
4843fcfe-55a8-4a65-bce7-ef00fa393351 | trentmkelly/LessWrong-43k | LessWrong | The Packaging and the Payload
I.
As I've run and studied meetups, there's a useful metaphor that's become more important to how I think about them. For most meetups, there's the packaging, and a payload, and these are related but useful to approach separately. Allow me to expand.
The payload is the thing you actually... |
de0a3a54-9259-4529-a16c-352e48eb4c9d | trentmkelly/LessWrong-43k | LessWrong | An AI risk argument that resonates with NYTimes readers
Ezra Klein of NYT put out a surprisingly sympathetic post on AI risk in the Sunday edition. It even quotes Paul Christiano and links back to LessWrong!
But what I'm actually here to talk about is the top reader-recommended comment on the article as of Sunday 11p... |
683aed9a-1c2b-4576-b15f-8c31ff54f3bd | trentmkelly/LessWrong-43k | LessWrong | Intellectual Hipsters and Meta-Contrarianism
Related to: Why Real Men Wear Pink, That Other Kind of Status, Pretending to be Wise, The "Outside The Box" Box
> WARNING: Beware of things that are fun to argue -- Eliezer Yudkowsky
Science has inexplicably failed to come up with a precise definition of "hipster", but fr... |
7473f2ef-ec31-42f6-921e-52f826e59fcc | StampyAI/alignment-research-dataset/special_docs | Other | Sino-Western cooperation in AI safety _ Brian Tse _ EA Global - San Francisco 2019-by Centre for Effective Altruism-video_id 3qYmLRqemg4-date 20190829
# Brian Tse Sino-Western cooperation in AI safety - EA Forum
\_International cooperation is essential if we want to capture the benefits of advanced AI while minimizin... |
3a7c0753-26bd-4890-97ca-256f484e3b9b | trentmkelly/LessWrong-43k | LessWrong | How can I delete my less wrong account?
Where can I get my less wrong account deleted?
No, the 'delete account' function doesn't work, and I am not posting from the account I want to have deleted as I don't want the name to stay in the google cache for another couple of months
I already asked that question here; I g... |
ea4949fa-b5b6-44d9-a3c2-23b5629b2acf | awestover/filtering-for-misalignment | Redwood Research: Alek's Filtering Results | id: post3003
This [EDIT: final version , presentation ] is a design for a conservative agent that I worked on with Marcus Hutter. Conservative agents are reluctant to make unprecedented things happen. The agent also approaches at least human-level reward acquisition. The agent is made conservative by being pessimistic... |
66ad20a8-ba78-4997-ad6f-b8be9185bda2 | trentmkelly/LessWrong-43k | LessWrong | In partially observable environments, stochastic policies can be optimal
I always had the informal impression that the optimal policies were deterministic (choosing the best option, rather than some mix of options). Of course, this is not the case when facing other agents, but I had the impression this would hold when... |
1525876d-8069-43dd-bf56-879d6c03088d | trentmkelly/LessWrong-43k | LessWrong | Supporting Effective Altruism through spreading rationality
So does spreading rationality contribute to Effective Altruism? I certainly think so, as a rationality popularizer and an Effective Altruist myself. My own donations of money and time is focused on my project, Intentional Insights, of trying to spread rationa... |
b455c0a5-f897-4a11-a732-0784074ef731 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Foresight for AGI Safety Strategy: Mitigating Risks and Identifying Golden Opportunities
This post is about I think why we should be using strategic foresight for AGI governance work. I am writing this without having read all the AGI governance / strategy reports and reading materials. I was planning on putting more e... |
4d6ba4f7-5b34-4108-9f3c-7537bdb51432 | trentmkelly/LessWrong-43k | LessWrong | How hard do we really want to sell cryonics?
For a lot of serious charitable causes, people expend tremendous resources on 'raising awareness' which probably, on net, accomplishes little or nothing for the cause it nominally supports. For cryonics, though, the technology already exists, the target audience can pay for... |
9eee6a78-aeaa-474c-a77d-f92632bcb8a0 | StampyAI/alignment-research-dataset/arxiv | Arxiv | Analyzing Inverse Problems with Invertible Neural Networks
1 Introduction
---------------
When analyzing complex physical systems, a common problem is that hidden system parameters, which scientists would like to know, cannot be measured directly.
For many of these problems, scientists have developed sophisticated ... |
66ba8525-b519-4532-9ebb-c160808d84d5 | trentmkelly/LessWrong-43k | LessWrong | [SEQ RERUN] Science Isn't Strict Enough
Today's post, Science Isn't Strict Enough was originally published on 16 May 2008. A summary (taken from the LW wiki):
> Science lets you believe any damn stupid idea that hasn't been refuted by experiment. Bayesianism says there is always an exactly rational degree of belie... |
b27ede1d-215b-4c0f-a224-f8b3a82b27e4 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Japan AI Alignment Conference Postmortem
The goal
========
[Conjecture](https://conjecture.dev/) collaborated with [Araya](https://www.araya.org/en/)to host a two day AI Safety conference in Japan, the first [Japan AI Alignment Conference](https://www.lesswrong.com/posts/tAQRxccEDYZY5vxvy/japan-ai-alignment-conferenc... |
d37a418e-1d57-4a46-b777-9e17a3b7d771 | trentmkelly/LessWrong-43k | LessWrong | Weekly LW Meetups
This summary was posted to LW Main on January 13th. The following week's summary is here.
Irregularly scheduled Less Wrong meetups are taking place in:
* Kyiv: 14 January 2017 18:15 (+0200): 14 January 2017 06:15PM
The following meetups take place in cities with regular scheduling, but involve a ... |
142d1ebb-bac8-4dfd-92e4-a7b335ea4b58 | trentmkelly/LessWrong-43k | LessWrong | Reconsider the anti-cavity bacteria if you are Asian
Update: further research has led me to believe that people of all races should test themselves for ALDH deficiency before using Lumina. Even if you don't exhibit AFR symptoms when drinking alcohol, your ALDH activity may still be decreased.
Many people in the ratio... |
53c61b21-0f06-4425-9eca-f9624354d6ab | trentmkelly/LessWrong-43k | LessWrong | [AN #100]: What might go wrong if you learn a reward function while acting
Newsletter #100 (!!)
Alignment Newsletter is a weekly publication with recent content relevant to AI alignment around the world. Find all Alignment Newsletter resources here. In particular, you can look through this spreadsheet of all summarie... |
018bc1f8-7529-4013-a119-0adfa8d45c88 | trentmkelly/LessWrong-43k | LessWrong | Sabine "Bee" Hossenfelder (and Robin Hanson) on How to fix Academia with Prediction Markets
"Don’t ask what science can do for you." is the clever title.
As I understand it, the prediction markets in science would not be easily influenced by the scientists and they would take into account a group of people currently ... |
59e79f52-1595-4b65-af94-c14ef1cc7fc6 | trentmkelly/LessWrong-43k | LessWrong | Rationality Quotes December 2014
Another month, another rationality quotes thread. The rules are:
* Please post all quotes separately, so that they can be upvoted or downvoted separately. (If they are strongly related, reply to your own comments. If strongly ordered, then go ahead and post them together.)
* Do not ... |
f48a42a3-934a-48e7-a9e2-8786132ab097 | StampyAI/alignment-research-dataset/arxiv | Arxiv | Dream to Control: Learning Behaviors by Latent Imagination
1 Introduction
---------------
Figure 1: Dreamer learns a world model from past experience and efficiently learns farsighted behaviors in its latent space by backpropagating value estimates bac... |
a0c6aaed-7f01-435d-abdb-c0166d9e8595 | StampyAI/alignment-research-dataset/special_docs | Other | The average utilitarian’s solipsism wager
The following prudential argument is relatively common in my circles: [We probably live in a simulation](http://www.simulation-argument.com/), but if we don’t, our actions matter much more. Thus, expected value calculations are dominated by the utility under the assumption tha... |
511cac2a-4a73-4c16-b6a8-80bebddcfe0b | trentmkelly/LessWrong-43k | LessWrong | What's up with ChatGPT and the Turing Test?
ChatGPT looks like it would pass the Turing Test, the gold standard of benchmarks measuring whether an AI has reached human-level intelligence. Yet Googling around, it doesn't seem that anyone has put on a full Turing Test.
Please comment with your thoughts on whether or h... |
b2c19650-d05e-441b-8b58-1933f5e2ef46 | StampyAI/alignment-research-dataset/special_docs | Other | Formalizing preference utilitarianism in physical world models
Synthese
DOI 10.1007/s11229-015-0883-1
Formalizing preference utilitarianism in physical world
models
Caspar Oesterheld1
Received: 23 April 2015 / Accepted: 27 August 2015
© The Author(s) 2015. This article is published with open access at Springerlink.com... |
5ea3daa8-e939-491a-b084-ed6854627035 | trentmkelly/LessWrong-43k | LessWrong | Chapter 91: Roles, Pt 2
A/N: This chapter does not contain a spoiler for any particular Orson Scott Card novel. It's a metaphor.
----------------------------------------
Shortly after, there was another knock upon the storeroom door.
"If you actually care about my mental health," the boy said without looking up, "y... |
7c3575e4-2174-4fe8-9f14-92f49e9fc1e0 | trentmkelly/LessWrong-43k | LessWrong | Guidelines for cold messaging people
I just received a random email from a stranger in which they ask me 5-10 hard-to-answer questions on effective altruism, emotional health, and moral cluelessness that bother them.
Reflecting on this, I came up with a list of eight rules of thumb you might want to consider when yo... |
d656d916-7e41-43bd-b604-d161e582331d | trentmkelly/LessWrong-43k | LessWrong | Where are the quick, easy wins?
The Efficient Markets Hypothesis suggest they shouldn't exist but in its strong form it's wrong. Otherwise there would be far fewer businesspeople who are successful in multiple largely unrelated industries.
Alternatively, those people are special and those opportunities are not availa... |
99f8360a-8975-4084-93ee-a53ba0a2563b | trentmkelly/LessWrong-43k | LessWrong | What's up with psychonetics?
http://deconcentration-of-attention.com/psychonetics.html
Psychonetics sounds a lot like something that rationalists would be into or at least know about but searching it on LessWrong gives a grand total of one mention in a nine-year old comment about it being a thing that rationalists mi... |
534c233d-8b2c-4b82-9e12-de23c195b281 | trentmkelly/LessWrong-43k | LessWrong | Hire (or Become) a Thinking Assistant
Of the posts I've delayed writing for years, I maybe regret this one the most.
I think more people (x-risk focused people in particular) should consider becoming (and hiring) metacognitive assistants. This is the single largest performance boost I know of – hire someone to sit wi... |
5cab7506-00b6-48dd-89d4-f6a1dfa3a3a0 | trentmkelly/LessWrong-43k | LessWrong | [SEQ RERUN] Two More Things to Unlearn from School
Today's post, Two More Things to Unlearn from School, was originally published on 12 July 2007. A summary (taken from the LW wiki):
> School encourages two bad habits of thought: (1) equating "knowledge" with the ability to parrot back answers that the teacher expect... |
7af1c896-1946-48fe-96e1-a45979a85672 | trentmkelly/LessWrong-43k | LessWrong | Less to Greater, Bookmarklet Edition
Lesser2Greater is a Chrome extension that redirects from lesswrong.com to greaterwrong.com. I’d like to use it, but I’m not a Chrome user. I do, however, use bookmarklets, so I made one to replicate Lesser2Greater’s functionality. |
0f1f7c03-0616-41d5-a03c-ae8c9d9b8cac | trentmkelly/LessWrong-43k | LessWrong | Spaniard Lesswrongers, Unite!
Por favor contesten en la sección de comentarios si son Españoles o residentes en España. La idea es entrar en comunicación, y poder discutir problemas de racionalidad propios de España. Que ésta página es muy EEUU-céntrica y aquí tenemos problemas endémicos.
Nota: Los hispanoamerican... |
d3125dfc-b57d-4fc0-ad7a-49d4f910e12e | trentmkelly/LessWrong-43k | LessWrong | A philosopher's critique of RLHF
In the spring, I went to a talk with Brian Christian at Yale. He talked about his book, The Alignment Problem, and then there was an audience Q&A. There was a really remarkable question in that Q&A, which I have transcribed here. It came from the Yale philosophy professor L.A. Paul. I ... |
c51db264-8d13-49a1-9c5a-1c68f9e0260a | trentmkelly/LessWrong-43k | LessWrong | Inefficient doesn't mean indifferent, but it might mean wimpy.
My composition teacher in college told me that in some pottery schools, the teacher holds up your pot, examines it, comments on it, and then smashes it on the floor. They do this for your first 100 pots.
In that spirit, this post's epistemic status is SMA... |
ef58145b-265c-44c7-a2d3-6c2ea791214d | trentmkelly/LessWrong-43k | LessWrong | SIAI awareness opportunity
I'm using today's "Skynet becomes self-aware" Facebook event to link to http://singinst.org/. Today (intentionally) is also the release date for Portal 2, which you should know features an unfriendly, if entertaining, AGI. |
4c49c661-45d2-4d08-8216-3fa4152026fe | trentmkelly/LessWrong-43k | LessWrong | Can We Place Trust in Post-AGI Forecasting Evaluations?
TLDR
Think "A prediction market, where most questions are evaluated shortly after an AGI is developed." We could probably answer hard questions more easily post-AGI, so delaying them would have significant benefits.
Motivation
Imagine that select pre-AGI legal ... |
da03bb0f-90a2-4f50-9d11-c1f01771da8c | trentmkelly/LessWrong-43k | LessWrong | A Detailed Critique of One Section of Steven Pinker’s Chapter “Existential Threats” in Enlightenment Now (Part 1)
This is the first of three posts; the critique has been split into three parts to enhance readability, given the document's length. For the official publication, go here.
Key findings:
—> The first quart... |
c2f2a719-998c-4f94-a1c2-36e5a0db1b21 | trentmkelly/LessWrong-43k | LessWrong | [SEQ RERUN] Use the Try Harder, Luke
Today's post, Use the Try Harder, Luke was originally published on 02 October 2008. A summary (taken from the LW wiki):
> A fictional exchange between Mark Hamil and George Lucas over the scene in Empire Strikes Back where Luke Skywalker attempts to lift his X-wing with the for... |
b3f03324-983a-4159-a42e-cd2a338b760d | trentmkelly/LessWrong-43k | LessWrong | Project ideas: Sentience and rights of digital minds
This is part of a series of lists of projects. The unifying theme is that the projects are not targeted at solving alignment or engineered pandemics but still targeted at worlds where transformative AI is coming in the next 10 years or so. See here for the introduct... |
de8f3c6c-e589-4c53-954d-261ca8dbd3b4 | trentmkelly/LessWrong-43k | LessWrong | [LINK] SMBC on cryonics
http://www.smbc-comics.com/index.php?db=comics&id=2405
|
094ff290-f334-4901-acc9-20c584a63ebc | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Reviews of “Is power-seeking AI an existential risk?”
(Edited 10/14/22 to add Lifland review. Edited 7/10/23 to add Levinstein review.)
Open Philanthropy solicited reviews of my draft report [“Is power-seeking AI an existential risk?”](https://docs.google.com/document/d/1smaI1lagHHcrhoi6ohdq3TYIZv0eNWWZMPEy8C8byYg/ed... |
c3d041be-c991-4e93-b540-55e5e4836148 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | All AGI Safety questions welcome (especially basic ones) [July 2023]
**tl;dr: Ask questions about AGI Safety as comments on this post, including ones you might otherwise worry seem dumb!**
Asking beginner-level questions can be intimidating, but everyone starts out not knowing anything. If we want more people in the ... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.