id stringlengths 36 36 | source stringclasses 15
values | formatted_source stringclasses 13
values | text stringlengths 2 7.55M |
|---|---|---|---|
8dbe68a0-3f07-45d9-b691-2f584ba9283a | StampyAI/alignment-research-dataset/special_docs | Other | Key Concepts in AI Safety: An Overview
MARCH 2021
Key Concepts in AI Safety: An Overview
CSET Issue Brief
AUTHORS Tim G. J. Rudner Helen Toner
Center for Security and Emerging Technology | 1 This paper is the first installment in a series on “AI safety,” an area of machine learning research that aims to identify... |
0cbcb270-4034-4641-929a-f6f879fa690f | StampyAI/alignment-research-dataset/youtube | Youtube Transcripts | 239. Soares, Tallinn, and Yudkowsky discuss AGI cognition
hello and welcome to session 239
in the ai safety.com reading group
tonight we'll be discussing part of the
post suarez talin and yukowski discuss
agi cognition
elias rykowski is
the senior researcher at the machine
intelligence research foundation jan
talon is... |
1fe4ba0c-f281-4ba6-943a-95c5e4cfc5d8 | trentmkelly/LessWrong-43k | LessWrong | Beliefs and Disagreements about Automating Alignment Research
Epistemic status: Mostly organizing and summarizing the views of others.
Thanks to those whose views I summarized in this post, and to Tamera Lanham, Nicholas Kees Dupuis, Daniel Kokotajlo, Peter Barnett, Eli Lifland, and Logan Smith for reviewing a draft.... |
8ce694be-0385-47cf-a3d9-769b812f60fd | trentmkelly/LessWrong-43k | LessWrong | Pittsburgh meetup Nov. 20
We'll be meeting at 7:00 this Saturday, at the Starbucks at 417 South Craig St (not on the CMU campus, as previously stated).
Hope to see you there! |
b1838fd0-22f4-47d4-9f76-27ff775a1b1c | StampyAI/alignment-research-dataset/youtube | Youtube Transcripts | Aligning ML objectives with human values
welcome to the afternoon their session
first talk key by Paul Christiana
about AI line hey so this talk will
cover some slightly unusual topics so
you should feel unusually free to stop
for questions or clarifications or
heckling my voice is not doing super
well so if I start w... |
0a812c40-110e-4eac-9c01-68cc378804ef | trentmkelly/LessWrong-43k | LessWrong | Call for Contributions: Implied Marginal Value of Time from Good/Services
If you buy a good or service with the hope that doing so will save you time, then doing so creates a lower bound on the marginal value of your time. For example, if you expect that filing your taxes by hand will take 5 hours, and filing your tax... |
d0416c7a-ae26-44d5-a13a-484068564f5e | trentmkelly/LessWrong-43k | LessWrong | The American Information Revolution in Global Perspective
In “What if they gave an Industrial Revolution and nobody came?” I reviewed The British Industrial Revolution in Global Perspective, by Robert Allen. In brief, Allen’s explanation for the Industrial Revolution is that Britain had high wages and cheap energy, wh... |
40365c6a-6cd0-4637-b2c1-d5f0bfdde5cd | trentmkelly/LessWrong-43k | LessWrong | What is the most significant way you have changed your mind in the last year?
Part of what it means to be a rationalist is to be able to change our minds when exposed to evidence.
What was the most significant way you have changed your mind in the last year? |
dcd2f4ba-90a4-4a4f-a3c3-ca04483f3312 | trentmkelly/LessWrong-43k | LessWrong | The Domain of Your Utility Function
Unofficial Followup to: Fake Selfishness, Post Your Utility Function
A perception-determined utility function is one which is determined only by the perceptual signals your mind receives from the world; for instance, pleasure minus pain. A noninstance would be number of living huma... |
dcdec78d-eab5-4b19-b0e7-fd9fa409356c | trentmkelly/LessWrong-43k | LessWrong | Working Through the Controlled Demolition Conspiracy
I hope I'm not breaking any taboos here. It's been a while since I've come onto the discussion section and I admit I'm not too up to date on the topics.
I'm having difficulty responding to someone who is convinced that 7WTC was brought down by controlled demolition... |
e7b0dde5-2283-44d6-8c69-d2a037d73881 | trentmkelly/LessWrong-43k | LessWrong | Against intelligence
Two years ago I wrote some pragmatic arguments that “human-like AI” is hard to develop and would be fairly useless. My focus was on the difficulty of defining a metric for evaluation and the cost-effectiveness of human brains.
But I think I failed to stress another fundamental point, which is tha... |
777adc07-0b69-4fef-90d3-80457d5c6d29 | trentmkelly/LessWrong-43k | LessWrong | Ideas for Next Generation Prediction Technologies
2021 Note: This was written and posted in December, 2016. The date it shows on LessWrong is 2019; I believe this refers to a time the post was (very minorly) updated, as part of moving it to the Prediction-Driven Collaborative Reasoning Systems sequence.
Prediction ma... |
c407cf1f-a411-4593-b220-f82d8b90d2f8 | trentmkelly/LessWrong-43k | LessWrong | Distillation Of DeepSeek-Prover V1.5
https://arxiv.org/abs/2408.08152 - "DeepSeek-Prover-V1.5: Harnessing Proof Assistant Feedback for Reinforcement Learning and Monte-Carlo Tree Search"
https://github.com/deepseek-ai/DeepSeek-Prover-V1.5
TL;DR
DeepSeek-Prover-V1.5 is an improved open-source language model for theo... |
90bc37d2-18ba-40a3-8ea1-b8021f6f7624 | trentmkelly/LessWrong-43k | LessWrong | Aieee! The stupid! it burns!
Last Wednesday (2010 Dec 01), BBC Radio 4 broadcast a studio discussion on the question: "should we actively try to extend life itself?" The programme can be listened to from the BBC here for one week from broadcast, and is also being repeated tomorrow (Saturday Dec 04) at 22:15 BST. (ETA:... |
4486e9d7-4128-4bfd-a2ab-f650a40258d4 | trentmkelly/LessWrong-43k | LessWrong | Don't Jump or I'll...
[Note: I am not a psychologist. I know next to nothing about mental illness and have never suffered from it myself. This story deals with suicide.]
The man stepped out onto the roof. The gray clouds above were illuminated not by the moon or stars, but by the light pollution of the city below. Ev... |
c9fa851f-15c4-403f-8884-43920385ce36 | trentmkelly/LessWrong-43k | LessWrong | Lonely Dissent
Asch’s conformity experiment showed that the presence of a single dissenter tremendously reduced the incidence of “conforming” wrong answers. Individualism is easy, experiment shows, when you have company in your defiance. Every other subject in the room, except one, says that black is white. You become... |
2ab4bacc-c769-48dc-a378-3187c2b6de6d | trentmkelly/LessWrong-43k | LessWrong | Unit Test Everything
Inspiration
* Specializing in Problems We Don't Understand
* You Are Not Measuring What You Think You Are Measuring
* Humans are very reliable agents
* Epistemic Legibility
The SELEX Experiment
For my MS research, I'm generating an aptamer for IL-10, a crucial anti-inflammatory immune signa... |
91321ecb-ace3-4ce6-8dc2-9527b0dc03c8 | trentmkelly/LessWrong-43k | LessWrong | Social Insight: Status Exchange: When an Insult Is a Compliment, When a Compliment Is an Insult
Some rough synonyms for status include respect, prestige, and "coolness."
Conceptually, the idea I sometimes think of when I try to describe "status" in its constituent parts is that to have status is to have people fee... |
c26823d5-f22f-4e29-9c3b-82a75e6db5e4 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | The Self in Artificial Consciousness: A Buddhist Investigation into Advanced AI
Hi all,
I'm sharing this undergrad thesis paper at the request of a fellow Effective Altruism member. It dives into some of the questions that arise around the topic of advanced (and possibly conscious) AI, what that even means, and what... |
2afad330-9a00-44db-98f8-436cab6bb228 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Trends in the dollar training cost of machine learning systems
Summary
=======
1. Using a [dataset](https://docs.google.com/spreadsheets/d/1beiholXkLpiYoeyDL64bOHLI-mXZ-_8ZnY1Ed6Sfjh0/edit?usp%3Dsharing) of 124 machine learning (ML) systems published between 2009 and 2022,[[1]](#fnobr3yxci06) I estimate that the cost... |
83da37f4-3cba-4063-b161-7c08bb17c04c | trentmkelly/LessWrong-43k | LessWrong | Covid 2/9/23: Interferon λ
There are two big pieces of pandemic news this week.
One is that there is growing alarm about the dangers of avian flu, or H5N1. Likely mammal-to-mammal transmission has been identified, a lot of wild animals are infected, and there is a real danger that this will become a pandemic. I will ... |
62397645-16e2-4543-9d93-ac82ba1fc298 | trentmkelly/LessWrong-43k | LessWrong | On Walmart, And Who Bears Responsibility For the Poor
Note: Originally posted in Discussion, edited to take comments there into account.
----------------------------------------
Yes, politics, boo hiss. In my defense, the topic of this post cuts across usual tribal affiliations (I write it as a liberal criticizing o... |
cd76e948-5d94-4156-ba16-9b2b297492d6 | trentmkelly/LessWrong-43k | LessWrong | Weekly LW Meetups
This summary was posted to LW Main on December 12th. The following week's summary is here.
Irregularly scheduled Less Wrong meetups are taking place in:
* Atlanta December Meetup - Game Night!: 27 December 2014 07:00AM
* Australia onling hangout: 14 December 2014 07:00PM
* Copenhagen: December M... |
4c0a8e65-2539-4520-8dd5-345e748477a4 | trentmkelly/LessWrong-43k | LessWrong | Steering subsystems: capabilities, agency, and alignment
Human brains have steering subsystems. LLMs and most RL agents do not. Steering systems, as defined by Max H, are goal-directed AI systems, or optimizers. Here I focus on steering subsystems: the parts of human and AI cognitive systems most directly relevant to ... |
08f45f3f-407f-4051-95fe-18a000f11d6a | trentmkelly/LessWrong-43k | LessWrong | Ten Thousand Years of Solitude
This is a linkpost for the article "Ten Thousand Years of Solitude", written by Jared Diamond for Discover Magazine in 1993, four years before he published Guns, Germs and Steel. That book focused on Diamond's theory that the geography of Eurasia, particularly its large size and common c... |
f0ec1480-0f78-4918-b4c2-d95c1f024102 | trentmkelly/LessWrong-43k | LessWrong | Keeping it (less than) real: Against
ℶ₂ possible people or worlds
Epistemic status and trigger warnings: Not rigorous in either math or physics. Not proof-read by any third party yet. May contain original research. Lengthy. Some sections are tongue-in-cheek. Anthrophics. Cosmology. Theism.
TL;DR: There are good reas... |
089f9aeb-260f-4b96-865a-494cb7e1336d | trentmkelly/LessWrong-43k | LessWrong | Optimal Employment
Related to: Best career models for doing research?, (Virtual) Employment Open Thread
In the spirit of offering some practical real world advice, let's talk about employment rationality. Let’s talk about optimal employment.1
You're young, smart, and hoping to have a positive impact on the world. Ma... |
6650e69d-e3fe-43c2-81ab-45c8f5807983 | StampyAI/alignment-research-dataset/arxiv | Arxiv | Time-Efficient Reward Learning via Visually Assisted Cluster Ranking.
1 Introduction
---------------
Learning reward functions from human feedback has been a successful paradigm in both sequential decision-making problems Abbeel and Ng ([2004](#bib.bib1)); Ziebart et al. ([2008](#bib.bib46)); Finn et al. ([2016](#b... |
537f2ebd-634d-4329-a1ac-429b8c94e322 | StampyAI/alignment-research-dataset/blogs | Blogs | ∀V: A Utopia For Ever
∀V: A Utopia For Ever
---------------------
∀V (read "universal voluntaryism" or "univol") is my utopia proposal. for people who are familiar with me or my material, this may serve as more of a clarification than an introduction; nevertheless, for me, this will be the pos... |
c80fe53e-5d89-45f5-8e72-c5c212d3c21e | trentmkelly/LessWrong-43k | LessWrong | Just a casual question regarding MIRI
Currently I am planning to start a mathematics degree when I enter university, however my interest has shifted largely to computational neuroscience and related fields, so I'm now planning to switch to an AI degree when I go to study. Having said that, MIRI has always posed intere... |
3871f273-78f2-4643-a158-8d25cbd0e036 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Are we in an AI overhang?
Over on [Developmental Stages of GPTs](https://www.lesswrong.com/posts/3nDR23ksSQJ98WNDm/developmental-stages-of-gpts), orthonormal mentions
> it at least reduces the chance of a [hardware overhang.](https://aiimpacts.org/hardware-overhang/)
An *overhang* is when you have had the ability t... |
c1a150f5-1f9d-49ff-b43f-a06faa1a043c | trentmkelly/LessWrong-43k | LessWrong | Evening drawing
This evening I became tempted by a YouTube video of an artist painting a portrait, which led me to be tempted by another such video, and then more of them, and then by one of these artists’ websites, and then by my own pencils and paper. (I did not become tempted by YouTube videos advertising breaking ... |
f4f13f52-df27-4cb4-9caa-059b4735ddbd | trentmkelly/LessWrong-43k | LessWrong | Robot Programmed To Love Goes Too Far (link)
http://www.muckflash.com/?p=200
Might be a nice story to point out to people who think "friendly" is easy.
|
4ac65264-2271-46ec-883a-c4a2477e9b73 | trentmkelly/LessWrong-43k | LessWrong | Using Flashcards for Deliberate Practice
Note: As in many of my recent posts, these ideas are new to me, and I'm just starting to test them out. This is part of a larger project of overhauling and refining my approach to scholarship. I'm excited to hear feedback and criticism. I anticipate this project will take years... |
d49ca0a9-d670-46b2-96a2-dd543032e25e | StampyAI/alignment-research-dataset/arbital | Arbital | Vingean reflection
[Vinge's Principle](https://arbital.com/p/1c0) implies that when an agent is designing another agent (or modifying its own code), it needs to approve the other agent's design without knowing the other agent's exact future actions.
Deep Blue's programmers decided to run Deep Blue, *without* knowing ... |
bd4f015a-6d9c-49ca-adc9-691cddce2ef8 | trentmkelly/LessWrong-43k | LessWrong | The cup-holder paradox
I'm shopping for a car, and I've spent many hours this past month reading user reviews of cars. There are seven things American car buyers have cared and complained about consistently for at least the past ten years. In roughly decreasing importance:
* Performance
* Gas mileage
* Frequency ... |
6af6d77f-3805-4e36-be8e-cb5dea8994df | trentmkelly/LessWrong-43k | LessWrong | [Link] First almost fully-formed human [foetus] brain grown in lab, researchers claim
This seems significant:
> An almost fully-formed human brain has been grown in a lab for the first time, claim scientists from Ohio State University. The team behind the feat hope the brain could transform our understanding of neuro... |
c881b753-73f0-47d8-814f-94d3c0bec48d | trentmkelly/LessWrong-43k | LessWrong | How truthful is GPT-3? A benchmark for language models
This is an edited excerpt of a new ML paper (pdf, code) by Stephanie Lin (FHI Oxford), Jacob Hilton (OpenAI) and Owain Evans (FHI Oxford). The paper is under review at NeurIPS.
Title: TruthfulQA: Measuring how models mimic human falsehoods
Abstra... |
4ff3a905-9397-4b03-ae02-dc72562c6bd7 | trentmkelly/LessWrong-43k | LessWrong | Shuttling between science and invention
“The Link Between Science and Invention: The Case of the Transistor,” a 1962 paper by Richard Nelson at the RAND Corporation, provides a fascinating case study in invention at the frontiers of science, and the relation between the two. In it, we can see that research—and researc... |
027c7e23-75c6-4927-a4c4-1e7b194506f7 | trentmkelly/LessWrong-43k | LessWrong | The Craft And The Codex
None |
dd65a055-2968-4ad9-87c6-352cc038e991 | trentmkelly/LessWrong-43k | LessWrong | Building and Entertaining Couples
Cross-posted, as always, from Putanumonit.
----------------------------------------
A friend told me recently that she’s looking for a husband to settle have kids, but she’s facing a problem: most of the men she goes on dates with fail to excite her intellectually in the first hour ... |
f27a6630-f5d0-490e-a831-ff02d3b55a20 | trentmkelly/LessWrong-43k | LessWrong | On AI Scaling
I’ve avoided talking about AI, mostly because everyone is talking about it. I think the implications of AI scaling were clear a while ago, and I mostly said my piece back then.
It’s time to take another crack at it.
Scaling laws
Scaling laws are a wonderful innovation[1]. With large enough training ru... |
8711c277-2fae-4f1d-8d21-ece8946b2e65 | StampyAI/alignment-research-dataset/blogs | Blogs | Help MIRI in a Massive 24-Hour Fundraiser on May 6th
**Update:** We’re now liveblogging the fundraiser [here](http://intelligence.org/2014/05/06/liveblogging-the-svgives-fundraiser/).
[](http://svgives.razoo.com/story/Machine-Intelligence-Research-I... |
5813ec01-dc29-4318-8119-a53f430fcfbe | trentmkelly/LessWrong-43k | LessWrong | Rationalization Maximizes Expected Value
TLDR: After making a choice, we often change how desirable we think the options were to better align with that choice. This is perfectly rational. How much you’ll enjoy an outcome depends on both (1) what you get, and (2) how you feel about it. Once what you get is settled, if ... |
b3c7842e-6ccf-44e6-a6c5-4417ea6fd70e | trentmkelly/LessWrong-43k | LessWrong | Procedural knowledge gap: public key encryption
In the spirit of uncovering procedural knowledge gaps, I'd like to know how to use public key encryption.
Is there some website which generates public and private keys, and lets you encode and decode according to those keys?
I'd love if there was some way I could send ... |
05cbf531-57a4-4379-9751-368d3ccd6197 | trentmkelly/LessWrong-43k | LessWrong | Reverse (intent) alignment may allow for safer Oracles
This is a core concept in my AI-in-a-box success model.
What is reverse alignment?
Consider the world-state optimized for the AI's utility, on one hand, and the world-state optimized for human utility, on the other. Let us call the first state "AI victory," and... |
8987eb3f-45f8-470c-b699-8702136b039d | trentmkelly/LessWrong-43k | LessWrong | Easy Intelligence Augmentation or Internet Wackaloonery?
On January 4, PJ Eby sent around an email linking an... interesting... website. The claim on the particular webpage he linked was as follows:
* the normal span of your breath is critical to how well your mental faculties can function
* the best activity for i... |
7020dbf7-29f6-4b67-8655-a597f19d0c54 | trentmkelly/LessWrong-43k | LessWrong | Realizations in Regards to Masculinity
This isn't a bait of some sort, but rather just a self-realization on how we view masculinity. I, for context, am a man, rather a boy who was not treated fairly in regards of being a man.
Growing up, I wasn't treated well by my father and I would often get the belt whenever I s... |
7e4358a0-5c75-45cb-8397-015e66107b02 | trentmkelly/LessWrong-43k | LessWrong | Race Along Rashomon Ridge
Produced As Part Of The SERI ML Alignment Theory Scholars Program 2022 Research Sprint Under John Wentworth
Two Deep Neural Networks with wildly different parameters can produce equally good results. Not only can a tweak to parameters leave performance unchanged, but in many cases, two neu... |
0fc81bd6-df1c-4b45-a090-a97a2eefc999 | trentmkelly/LessWrong-43k | LessWrong | Thread for making 2019 Review accountability commitments
The LessWrong 2019 Review is upon us. We're out of the nomination phase, and starting the review phase. 121 posts have passed into this second stage.
Now we join forces to answer some crucial questions:
* Are the claims in these posts true?
* Are they impo... |
2a2e53b3-543a-48c4-8853-f8f8890acbcc | trentmkelly/LessWrong-43k | LessWrong | Proposal: Twitter dislike button
The popular story of Twitter’s role in the ruin of civilization is that it is a runaway trash fire of reciprocal anger and offense, where otherwise nice people are possessed by overwhelming outrages, and drawn into throwing their own energy behind creating the vilest and most vindictiv... |
3c18278d-ebe4-45a9-9df9-6c2c5c17300d | trentmkelly/LessWrong-43k | LessWrong | Meetup : Fort Lauderdale
Discussion article for the meetup : Fort Lauderdale
WHEN: 21 April 2012 06:00:00PM (-0400)
WHERE: Starbucks (2026 North Flamingo Road, Pembroke Pines, FL 33027)
As usual, during this meeting we will be bringing articles of interest to discuss. We are very informal and all newcomers are enco... |
c53f529b-ccff-45e3-b262-3dd5a3e513d5 | trentmkelly/LessWrong-43k | LessWrong | Why do ‘respectable’ women want dead husbands?
I find this hostile wife phenomenon pretty confusing for several reasons:
1. Wanting other people dead is generally considered an evil character trait. Most people either don’t have such wishes or don’t admit to them. This is especially the case when the person you pref... |
86645ede-ebff-44e1-bee3-942ff5dbed70 | trentmkelly/LessWrong-43k | LessWrong | Open Source Automated Interpretability for Sparse Autoencoder Features
Generated by Dalle
Background
Sparse autoencoders recover a diversity of interpretable, monosemantic features, but present an intractable problem of scale to human labelers. We investigate different techniques for generating and scoring text expl... |
457a8e15-e5ee-482a-a356-b8dd189a538a | trentmkelly/LessWrong-43k | LessWrong | The Jolly Green Giant Chronicles [ChatGPT]
Cross-posted from New Savanna.
At the end of December, on the 30th, I gave ChatGPT the following prompt:
> What happened when the Jolly Green Giant had a colorless green idea after eating a stew laced with ergot?
To which it replied in that deadpan pedantic way it has:
>... |
d25f1cba-0b09-40ff-acb9-70ec84fa3e5b | trentmkelly/LessWrong-43k | LessWrong | Superintelligent Introspection: A Counter-argument to the Orthogonality Thesis
John Wentworth serendipitously posted How To Write Quickly While Maintaining Epistemic Rigor when I was consigning this post to gather dust in drafts. I decided to write it up and post it anyway. Nick Bostrom's orthogonality thesis has neve... |
1d98d632-9b82-4ec3-b0ca-5c09945f1db7 | LDJnr/LessWrong-Amplify-Instruct | LessWrong | "Introduction
Due to long inferential distances it's often very difficult to use knowledge or understanding given by rationality in a discussion with someone who isn't versed in the Art (like, a poor folk who didn't read the Sequences, or maybe even not the Goedel, Escher, Bach !). So I find myself often forced to use ... |
58400c7a-2d2c-49d4-842a-a0023e543033 | trentmkelly/LessWrong-43k | LessWrong | Link: Why and how to debate charitably
Even though this was written by a current Less Wrong poster (hi, pdf23ds!), I don't think it has been posted here: Why and how to debate charitably (pg. 2, comments). (Edit: The original pdf23ds.net site has sadly been lost to entropy – Less Wrong poster MichaelBishop found a rep... |
f8efef1d-5c98-49e2-88cb-f7fd5ebf2a87 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Is behavioral safety "solved" in non-adversarial conditions?
I’m trying to crystalize something I was said to a friend recently: I think that techniques like [RLHF](https://openai.com/research/learning-from-human-preferences) and [Constitutional AI](https://arxiv.org/abs/2212.08073) seem to be sufficient for making la... |
f0e29f25-717f-4c60-8a1b-f3c0fe170f22 | trentmkelly/LessWrong-43k | LessWrong | Plant Seeds of Rationality
After his wife died, Elzéard Bouffier decided to cultivate a forest in a desolate, treeless valley. He built small dams along the side of the nearby mountain, thus creating new streams that ran down into the valley. Then, he planted one seed at a time.
After four decades of steady work, the... |
60e4d0b9-4a6a-4796-a16b-b89cb7fd6c64 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | An Open Philanthropy grant proposal: Causal representation learning of human preferences
*This is a proposal I wrote for the recent [Open Philanthropy call on AI Alignment projects](https://www.openphilanthropy.org/focus/global-catastrophic-risks/potential-risks-advanced-artificial-intelligence/request-for-proposals-f... |
698aee0d-24fd-4593-811e-803eae3f7cab | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Compositional language for hypotheses about computations
**TLDR:** I sketch a language for describing (either Bayesian or infra-Bayesian) hypotheses about *computations* (i.e. computable[[1]](#fn-hxDKRSC8mTKCyak9e-1) logical/analytic facts). This is achieved via a novel (AFAIK) branch of automata theory, which general... |
e579157b-3064-4cd9-b0be-a15a50be7c2b | StampyAI/alignment-research-dataset/special_docs | Other | It's not too soon to be wary of AI: We need to act now to protect humanity from future superintelligent machines
JAMIE CHUNG/TRUNK ARCHIVE; HAND: JUSTIN METZ
editor’s note: This article is based on a chapter of the author’s newly released book , Human Compatible: Artificial Intelligence
and the Problem of Control, pu... |
5ef39979-dc9d-49a1-980a-a633103e8741 | trentmkelly/LessWrong-43k | LessWrong | AISN #51: AI Frontiers
Welcome to the AI Safety Newsletter by the Center for AI Safety. We discuss developments in AI and AI safety. No technical background required.
In this newsletter, we cover the launch of AI Frontiers, a new forum for expert commentary on the future of AI. We also discuss AI 2027, a detailed sce... |
374ec53a-68f4-472a-b80f-4677d4cedb59 | trentmkelly/LessWrong-43k | LessWrong | Kelly betting vs expectation maximization
People talk about Kelly betting and expectation maximization as though they're alternate strategies for the same problem. Actually, they're each the best option to pick for different classes of problems. Understanding when to use Kelly betting and when to use expectation maxi... |
8b803372-42aa-4cf1-9c10-db3aa66f6957 | trentmkelly/LessWrong-43k | LessWrong | Breaking Down Goal-Directed Behaviour
When we speak about entities 'wanting' things, or having 'goal-directed behaviour', what do we mean?
Because most of the actors that we (my human readers and I) attentively interact frequently with are (presumably) computationally similar (to each other and to ourselves), it is e... |
9ae9edb2-6df9-4a9d-8309-0e68c976e92a | trentmkelly/LessWrong-43k | LessWrong | May-June 2025 Progress in Guaranteed Safe AI
There will be a AIxFM conference in the Bay Area Q4, according to a little birdie.
Morph ships big autoformalization result in 3599 lines of Lean
They have human decomposition in the latex/lean blueprint, into 67 lemmas with human spotchecking. Still, I’m impressed with t... |
3fe4f679-ba43-4610-8fdd-89dbae13b5cd | trentmkelly/LessWrong-43k | LessWrong | Argue Politics* With Your Best Friends
Epistemic Status: I endorse this strongly but don’t think I’m being original or clever at all.
Until recently — yesterday, in fact — I was seriously wrong about something.
I thought that it was silly when I saw people spending lots of energy arguing with their closest friends w... |
5b8a0c29-f499-4d8b-8346-cd7c73758d30 | StampyAI/alignment-research-dataset/arxiv | Arxiv | Deep Video Portraits
|
ca82dc7e-7513-4830-bcbf-70022d9a63ee | trentmkelly/LessWrong-43k | LessWrong | Anti-Corruption Market
General Secretary Lu called his most trusted advisor Zhou into his office.
"Close the door," said the General Secretary.
Zhou closed the door with quivering arms.
"Don't worry. You're doing great. In fact, that's why I called you in today. I'm starting a new anti-corruption campaign," said th... |
dbeae2f6-8b3f-498a-a64e-82ae92a15d70 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | More to explore on 'Risks from Artificial Intelligence'
### The development of artificial intelligence
* [*AlphaGo - The Movie - DeepMind*](https://tinyurl.com/vsj22235) *- A documentary exploring what artificial intelligence can reveal about the 3000-year-old game of Go, and what that can teach us about the future p... |
de470f0b-248f-44dd-8e91-6a58d599013f | trentmkelly/LessWrong-43k | LessWrong | Reverse RSS Stats
In 2013 I built a weird feature where clicking on " ◂◂ RSS" at the top of a post gets you an RSS feed that runs in reverse from that post, with a 'new' old post with the publication of each new new post. It's a bit hard to think about, and reminds me of trying to keep tenses straight when time travel... |
c1d4d0f6-cdfc-4b1b-a3aa-6a2b68193da9 | trentmkelly/LessWrong-43k | LessWrong | If You Were Brilliant When You Were Ten...
(If I do anything wrong here, please tell me. I don't know what I'm doing and would benefit from being told what I've got wrong, if anything. I've never made a top-level post here before.)
So, it seems like most people here are really smart. And a lot of us, I'm betting, wil... |
60e50218-cbf6-4a9f-8189-c5171f71bf8a | StampyAI/alignment-research-dataset/arbital | Arbital | Real number (as Dedekind cut)
%%comment: Mnemonics for defined macros: \Ql = Q left, \Qr = Q right, \Qls = Q left strict, \Qrs = Q right strict.%%
The rational numbers have a problem that makes them unsuitable for use in calculus — they have "gaps" in them. This may not be obvious or even make sense at first, because... |
f28de976-af8e-4f6c-b29f-e984dac4ffbf | StampyAI/alignment-research-dataset/youtube | Youtube Transcripts | R Dobbe: Towards a Systematic and Realistic Practice for Developing Safe and Democratically Sound AI
go ahead
thank you evgeny and hi everyone
um it's a pleasure to be here and be
back uh to present
uh in the ai tech community i remember
uh
visiting delft from new york a couple
years ago when when it
when it all got s... |
3b5c5bb1-369f-43b8-825e-fc8d8f435034 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | How I Learned to Stop Worrying and Love MUM
*Epistemic status: Speculative*
The title is a tongue-in-cheek reference to Google AI's latest showcase: [Multitask Unified Model](https://blog.google/products/search/introducing-mum/), or MUM for short. Further details can be found in their arXiv paper; [Rethinking Search... |
457f39d3-d350-4b99-870d-ae9fa574fda7 | StampyAI/alignment-research-dataset/arxiv | Arxiv | Toward Campus Mail Delivery Using BDI
1 Introduction
---------------
Autonomous systems are designed in such a manner that they can intelligently react to ever-changing environments and operational conditions. Given such flexibility, they can accept goals and set a path to achieve these goals in a self-responsible ... |
c9b8458a-e08a-4873-b9ca-75a7f0096599 | trentmkelly/LessWrong-43k | LessWrong | Some questions for the people at 80,000 Hours
I originally wrote this for the EAForum and later realised they could be linked? Anyway, thought this might be of interest to the people at LW (I don't expect the folks at 80k to answer here as well as the EAForum)
---
1. By recommending people work at the big AI labs (w... |
e0c5216d-7d11-4deb-ab12-deb97509dba5 | trentmkelly/LessWrong-43k | LessWrong | Reference class of the unclassreferenceable
One of the most useful techniques of rationality is taking the outside view, also known as reference class forecasting. Instead of thinking too hard about particulars of a given situation and taking a guess which will invariably turned out to be highly biased, one looks at o... |
50a09091-6713-4e0f-9084-034d3ae1a07b | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Are there substantial research efforts towards aligning narrow AIs?
I am certainly not an expert, but what I have read in the field of AI alignment is almost solely targeted at handling the question of how to make sure that a general AI does not start optimizing for outcomes that humans do not want to happen.
This is... |
9c730193-4003-4881-8ff1-61cbc645cd28 | trentmkelly/LessWrong-43k | LessWrong | Petrov Day Retrospective, 2023 (re: the most important virtue of Petrov Day & unilaterally promoting it)
When Stanislav Petrov's missile alert system pinged, the world was not watching. Russia was not watching. Perhaps a number of superiors in the military were staying in the loop about Stanislav's outpost, waiting fo... |
e6b56c85-9493-48c7-aa27-0b3c787859dc | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | What Makes A Good Measurement Device?
In some sense, any new measurement device automates a part of research. A thermometer automates the task of sticking one's finger in something to check how hot it is, a scale automates the task of holding something to check how heavy it is, etc. The automated version is not only m... |
4afaf59f-bfcb-4f82-8860-fb2bcd5a1791 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Contest: 250€ for translation of "longtermism" to German
I'm starting a German translation project and I think there will be some strong path dependencies. So **I'm offering 50€ for the best translation of the word "longtermism" into German.**
**Simon Grimm offers another 200€ if there are any especially promising id... |
26783735-c9f1-4d51-861f-805ca6e0a943 | trentmkelly/LessWrong-43k | LessWrong | Eating Boogers
Should you eat boogers practice automucophagy?
The science on this is surprisingly lacking. Maybe you’re one of the normal people that find it gross and refuses to engage in the behavior for fear of breaking social norms.
But what about us, bold thinkers, outspoken intelligentsia willing to thrust thr... |
e8b485e3-6b34-43b2-a2e9-afd6a351e842 | StampyAI/alignment-research-dataset/special_docs | Other | Are we living at the hinge of history
Are we living at the hinge
of history?
William MacAskill
Global Priorities Institute | September 2020
GPI Working Paper No. 12-2020
Are We Living At The Hinge Of History? 0. Introduction In the final pages of On What Matters, Volume II (2011), Derek Pa... |
cfc17657-c684-492c-b441-56ddf5e9fc08 | trentmkelly/LessWrong-43k | LessWrong | Weekly LW Meetups: Atlanta, Austin, Berlin, Brussels, Cambridge MA, Cincinnati, Durham, Vancouver, Washington DC
This summary was posted to LW Main on April 19th. The following week's summary is here.
There are upcoming irregularly scheduled Less Wrong meetups in:
* April Atlanta Meetup! Friday, April 19th: 19 Apri... |
74557e9c-992a-4301-be4c-d89e2506eb97 | trentmkelly/LessWrong-43k | LessWrong | Microstartup Stories: Initial Thoughts
A while back I read this post saying that you should write a business plan. Previously I thought business plans were dumb. I recall that being the conventional wisdom in the tech world and hearing people like Paul Graham and David Heinemeier Hansson say so. But that blog post cha... |
c9531413-2e1d-46e3-92e7-339db6ac2c18 | trentmkelly/LessWrong-43k | LessWrong | Can This Idea Dramatically Improve Effective Vegan Activism?
In "At what cost, carnivory?" by Gregory Lewis, ACE's estimation was that the cost of making one vegan for one year was 4.38$ (2015) on average. The answer to improving vegan activism could be to target students and have a charity that copied and held Gary Y... |
a1917f66-4050-4c90-bc32-1181e12e75bc | StampyAI/alignment-research-dataset/arxiv | Arxiv | The Singularity May Never Be Near
1 Introduction
---------------
There is both much optimism and pessimism around artificial
intelligence (AI) today.
The optimists are investing millions of dollars,
and even in some cases billions of dollars into AI.
The pessimists, on the other hand, predict
that AI will end many ... |
e992d05c-09b4-4ada-a593-cae10c0d5bb7 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Alignment Implications of LLM Successes: a Debate in One Act
**Doomimir**: Humanity has made no progress on the alignment problem. Not only do we have no clue how to align a powerful optimizer to our "true" values, we don't even know how to make AI "corrigible"—willing to let us correct it. Meanwhile, capabilities con... |
60fd5849-31fe-43df-83bb-dff800a991fa | trentmkelly/LessWrong-43k | LessWrong | Alignment Newsletter #34
Find all Alignment Newsletter resources here. In particular, you can sign up, or look through the database of all summaries.
Highlights
Scalable agent alignment via reward modeling (Jan Leike): This blog post and the associated paper outline a research direction that DeepMind's AGI safety te... |
a102acb8-e670-44bd-8a0c-917dcf803792 | trentmkelly/LessWrong-43k | LessWrong | Helsinki meetup Saturday 9.4 (cancelled)
Time: Saturday 9th of April from 15:00 onwards
Place: Cafe Picnic at Yliopistonkatu 5
This is the second Helsinki meetup. The last one was successful, going on for three hours with around ten people there. Like before, lurkers and newbies are extremely welcome, as well as peo... |
ddd4dc4d-c0b2-4d0f-bc16-ee462ca4cb9d | StampyAI/alignment-research-dataset/blogs | Blogs | Randall Larsen and Lynne Kidder on USA bio-response
[Colonel Randall Larsen](http://www.randalllarsen.com/), USAF (Ret), is the National Security Advisor at the [UPMC Center for Health Security](http://www.upmchealthsecurity.... |
d7452ea5-dd51-4eca-80d0-bd89becbad5b | trentmkelly/LessWrong-43k | LessWrong | Gizmo Watch Review
Summary: the watches do the basic things they need to and have let us give our kids more freedom, but there are also a bunch of annoying limitations and they're 36% more expensive than they look. I'm overall happy we got them.
Our older two kids, Lily and Anna, are ten and eight, and are mature eno... |
505e6a8e-e27f-4a3c-9472-daeaa6627d4b | trentmkelly/LessWrong-43k | LessWrong | [AN #135]: Five properties of goal-directed systems
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 summaries that have ever been in the newsletter.
Aud... |
c7a4b062-f455-46c0-a33c-65350aa910c9 | trentmkelly/LessWrong-43k | LessWrong | If a nuke is coming towards SF Bay can people bunker in BART tunnels?
|
819534a3-f474-4795-bdf6-2cc4e49f38a9 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Kyiv: 29 December 2016 18:15 (+0200)
Discussion article for the meetup : Kyiv: 29 December 2016 18:15 (+0200)
WHEN: 29 December 2016 06:15:00PM (+0200)
WHERE: Lva Tolstoho St, 3, Kyiv, Ukraine
On the agenda are short self-introductions and topical discussion. You can find more information about the upcomin... |
33a74ab4-21b7-411d-9e91-e0301b6f35a9 | trentmkelly/LessWrong-43k | LessWrong | Show LW: Fallacyfiles
http://www.fallacyfiles.org
> What are the fallacy files?
> I began collecting and studying logical fallacies about thirty-eight years ago, when I first became interested in logic. This collection took two forms:
> A collection of named fallacies—such as "ad hominem"—that is, types of bad reas... |
4be18b27-0746-4ea4-9885-1c279cc0296c | trentmkelly/LessWrong-43k | LessWrong | Shikamaru vs. the Logical Fallacies
Presenting: Shikamaru vs. the Logical Fallacies, a rationalist fanfic set in the Naruto universe and starring a realistically rational Shikamaru. The writer is a friend of mine going under the alias FanFicBias, and I am an editor along with the author.
We welcome reviews on FanFict... |
16c716b9-8316-4cf8-bc9a-881f12008870 | trentmkelly/LessWrong-43k | LessWrong | Thought experiments on simplicity in logical probability
A common feature of many proposed logical priors is a preference for simple sentences over complex ones. This is sort of like an extension of Occam's razor into math. Simple things are more likely to be true. So, as it is said, "why not?"
Well, the analogy h... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.