id stringlengths 36 36 | source stringclasses 15
values | formatted_source stringclasses 13
values | text stringlengths 2 7.55M |
|---|---|---|---|
9ca7169c-71b6-45f2-8c54-0ac753b746e9 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Logical Counterfactuals & the Cooperation Game
**Update**: Originally I set the utility for AB or BA to -10, -10; but I've now realised that unnecessarily complicates the problem.
---
Logical counterfactuals (as in Functional Decision Theory) are more about your state of knowledge than the actual physical state of... |
5e6d30aa-04b8-4016-be8f-9b9a50549ff3 | trentmkelly/LessWrong-43k | LessWrong | Spending Update 2020
It's been a bit over two years since our last spending update ( 2018, 2016, 2014). We're no longer tracking our expenses to the dollar in most cases, except where that's required for taxes, but since we rent out part of our house it turns out that this covers all the big categories.
For past upda... |
9dc8626f-2916-4b10-b083-9430ebf49646 | trentmkelly/LessWrong-43k | LessWrong | Unpopularity of efficiency
I feel like ‘efficiency’ is often scowled at. It is associated with factories and killing and commercialization, and people who are no fun. Things are openly criticized for being oriented toward efficiency. Nobody hopes to give their children an efficient childhood or asks for an efficient V... |
42ecc700-75ff-49b8-8940-deb794a050b3 | LDJnr/LessWrong-Amplify-Instruct | LessWrong | "At a family party some years ago, one of my uncles remarked on how little science really knows. For example, we still have no idea how gravity works - why things fall down. "Actually, we do know how gravity works," I said. (My father, a Ph.D. physicist, was also present; but he wasn't even touching this one.) "We do?"... |
8c2110a9-2c09-4111-a7c8-b0a0afd3c226 | StampyAI/alignment-research-dataset/arxiv | Arxiv | Exploiting Model Equivalences for Solving Interactive Dynamic Influence Diagrams
Journalof Artificial Intelligence Research 43(2012)211-2 55 Submitted 8/11;published 02/12
Exploiting Model Equivalences forSolving InteractiveDynamic
Influence Diagrams
YifengZeng YFZENG@CS.AAU.DK
Dept. of Computer Science
Aalborg Universi... |
430a13b9-a4c8-4033-a7be-e8df83e2d9a2 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | The Credit Assignment Problem
*This post is eventually about* [*partial agency*](https://www.lesswrong.com/s/HeYtBkNbEe7wpjc6X)*. However, it's been a somewhat tricky point for me to convey; I take the long route. **Epistemic status:** slightly crazy.*
---
I've occasionally said "Everything boils down to credit as... |
fd964eb8-c157-4617-b109-4d034008a016 | trentmkelly/LessWrong-43k | LessWrong | Benchmarking an old chess engine on new hardware
I previously explored the performance of a modern chess engine on old hardware (1, 2). Paul Christiano asked for the case of an old engine running on modern hardware. This is the topic of the present post.
State of the art
Through an online search, I found the CCRL Bl... |
f631d62c-f636-4ec9-961b-39222f00c264 | trentmkelly/LessWrong-43k | LessWrong | The best you can
In fiction, protagonists narrow their focus until the difference between success and failure on their specific task seems like the difference between victory and defeat. Batman attempts to solve the mystery while ensuring that nobody dies; meanwhile, children in Africa suffer from Malaria. The crew in... |
eec0a80a-6d45-468c-aaec-ff8faa299917 | trentmkelly/LessWrong-43k | LessWrong | The substance of money
Not much epistemic effort here: this is just an intuition that I have to model a vast and possibly charged field. I'm calling upon the powers of crowd-finding to clarify my views.
Tl;dr: is the debate in economics over the nature of money really about the definition or just over politics?
I'm ... |
20fe0d89-5e3f-4866-b95d-bb79c22206be | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Intro to AI risk for AI grad students?
Ideas for topics, points, or resources that would be **especially valuable to get across to incoming graduate students in Artificial Intelligence**?
Context
-------
On October 18, I will be giving a ≤1.5-hour lecture to the incoming class of 50+ graduate students in the AI prog... |
e0fe2b6e-b7b4-47c9-9e23-bdfbf695d550 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | AMA on EA Forum: Ajeya Cotra, researcher at Open Phil
Hi all, I'm [Ajeya](https://www.openphilanthropy.org/about/team/ajeya-cotra), and I'll be doing an AMA on the EA Forum (this is a linkpost for my announcement there). I would love to get questions from LessWrong and Alignment Forum users as well -- please head on o... |
74dbfdc8-24da-4c4a-a41c-de7ddb2ccd3b | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | What is a VNM stable set, really?
Previously: [Book report: Theory of Games and Economic Behavior (von Neumann & Morgenstern)](https://www.lesswrong.com/posts/qRKyZGcoio9JhdmvX/book-report-theory-of-games-and-economic-behavior-von)
*Summary: Von Neumann and Morgenstern defined the notion of stable set in order to ch... |
68c51c1a-cf6d-4317-9751-081d5e523fb0 | trentmkelly/LessWrong-43k | LessWrong | AI #116: If Anyone Builds It, Everyone Dies
If Anyone Builds It, Everyone Dies is the title of the new book coming September 16 from Eliezer Yudkowsky and Nate Sores. The ‘it’ in question is superintelligence built on anything like the current AI paradigm, and they very much mean this literally. I am less confident in... |
ab50fdee-d585-4dbd-aa7e-aeb205165e39 | StampyAI/alignment-research-dataset/arxiv | Arxiv | Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database
1 Introduction
---------------
Governments, corporations, and individuals are increasingly deploying intelligent systems to safety-critical problem areas, including transportation NTSB ([2017](#bib.bib15 "Collision Between ... |
f6a8ce78-05f7-4c74-ae5f-58a9ed27c79e | trentmkelly/LessWrong-43k | LessWrong | Subagents, introspective awareness, and blending
> In this post, I extend the model of mind that I've been building up in previous posts to explain some things about change blindness, not knowing whether you are conscious, forgetting most of your thoughts, and mistaking your thoughts and emotions as objective facts, w... |
69cb4a2f-ba81-4024-9964-41f8324b3475 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Enhancing Corrigibility in AI Systems through Robust Feedback Loops
**Abstract**
============
This article explores the concept of corrigibility in artificial intelligence and proposes a detailed framework for a robust feedback loop to enhance corrigibility. The ability to continuously learn and correct errors is cri... |
e3231a50-3812-494c-af3c-6620adc073cb | LDJnr/LessWrong-Amplify-Instruct | LessWrong | "(content warning: nuclear war, hypothetical guns, profanity, philosophy, one grammatically-incorrect comma for readability’s sake)This is a very special time of year, when my whole social bubble starts murmuring about nuclear war, and sometimes, some of those murmurers, urging their listeners to worry more, will state... |
dc9f1574-3bd4-42b2-a6d6-53347801cc21 | trentmkelly/LessWrong-43k | LessWrong | Kevin T. Kelly's Ockham Efficiency Theorem
There is a game studied in Philosophy of Science and Probably Approximately Correct (machine) learning. It's a cousin to the Looney Labs game "Zendo", but less fun to play with your friends. http://en.wikipedia.org/wiki/Zendo_(game) (By the way, playing this kind of game is e... |
7c674adc-00f2-4839-adcd-1ec5a5bfff0d | StampyAI/alignment-research-dataset/youtube | Youtube Transcripts | #029 GPT-3, Prompt Engineering, Trading, AI Alignment, Intelligence
welcome back to the machine learning
street talk youtube channel
today yannick and i in particular
changed our mind about gbt3
a little bit v-drag posted a very
interesting comment on the kanalihi
video
a couple of episodes ago giving some
examples of... |
744cd204-e455-4aeb-b9e1-985a5f410d0a | trentmkelly/LessWrong-43k | LessWrong | Whence the symptoms of social media?
Crossposted from world spirit sock puppet.
A thing I liked about The Social Dilemma was the evocative image of oneself being in an epic contest for one’s attention with a massive and sophisticated data-nourished machine, tended by teams of manipulation experts. The hopelessness of... |
6efca051-2f08-4273-9a3d-2c3c99c3251f | trentmkelly/LessWrong-43k | LessWrong | The Developmental Role of Play
When I was a kid, maybe around 8 or 9, I had a cabinet full of Legos. To the casual observer it was a few messy shelves crammed with plastic blocks, but to me it was a portal into another world. Opening it up I’d see little buildings and people resting in suspended animation, and as I tu... |
c70474e8-da2e-4c20-89ad-b1926ff5c819 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | A Brief Introduction to ACI, 2: An Event-Centric View
In the [previous chapter](https://www.lesswrong.com/posts/iCmdssymD8SFoM2NP/a-brief-introduction-to-algorithmic-common-intelligence-aci) , we have introduced the basic principle of ACI:
*Intelligent Agents should behave the same way as past behaviors which are doi... |
77ebbd6c-5572-4268-9d84-b736adb1428b | trentmkelly/LessWrong-43k | LessWrong | What are fun little puzzles / games / exercises to learn interesting concepts?
Things in the category of the 2-4-6 puzzle where you need to guess the rule and can try as many examples as you want to figure out the puzzle, and it teaches 'gung jr ner cebar gb pbasvezngvba ovnf' (https://rot13.com/).
Please describe th... |
ef5e46e8-85ad-4659-9fd0-dee8f5639137 | trentmkelly/LessWrong-43k | LessWrong | Small Habits Shape Identity: How I became someone who exercises
This morning, I lingered in my bed until late. Despite that, I went for a run. It started okay, but upgraded to great after a few minutes. Running felt good, natural, what I should be doing right now. I even went for a longer path than initially planned, ... |
cdb1ab38-e298-438f-a359-24560053ec9e | trentmkelly/LessWrong-43k | LessWrong | [Book Review] "The Signal and the Noise: Why So Many Predictions Fail—But Some Don’t.", by Nate Silver
Here's a link to a review, by The Economist, of a book about prediction, some of the common ways in which people make mistakes and some of the methods by which they could improve:
Looking ahead : How to look ahead—a... |
ebdba2d7-e62f-4886-a06f-6983313b2511 | trentmkelly/LessWrong-43k | LessWrong | Who are some people you met that were the most extreme on some axis?
Characteristics like:
* person that had the most imaginary friends
* person that was the most extroverted
* person that lived the most online
* person that recorded their life the most
* person that asked the most questions online
* person tha... |
fb87d13f-9b56-4988-9d5f-7d1420917668 | awestover/filtering-for-misalignment | Redwood Research: Alek's Filtering Results | id: post2637
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. Audio version here (may not be up yet). Pl... |
fd6a80c3-077a-45b3-82b0-1229af1e6a6f | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | A rough and incomplete review of some of John Wentworth's research
This is going to be a half-assed review of John Wentworth's research. I studied his work last year, and was kinda hoping to write up a better review, but am lowering my standards on account of how that wasn't happening.
Short version: I've been unimpr... |
a08c61a8-efbb-40a3-9465-124de997355f | trentmkelly/LessWrong-43k | LessWrong | Some secondary statistics from the results of LW Survey
Global LW (N=643) vs USA LW (N=403) vs. Average US Household (Comparable Income) Income Bracket LW Mean Contributions USA LW Mean Contribution US Mean Contributions** [1] LW Mean Income USA LW Mean Income US Mean*** Income [1] LW Contributions /Income USA ... |
0c7799e8-2cb2-461a-a41e-d6eccad0c74d | StampyAI/alignment-research-dataset/lesswrong | LessWrong | The inordinately slow spread of good AGI conversations in ML
Spencer Greenberg wrote [on Twitter](https://twitter.com/SpencrGreenberg/status/1536760598486196225):
> Recently @KerryLVaughan has been critiquing groups trying to build AGI, saying that by being aware of risks but still trying to make it, they’re reckles... |
33130c27-2abe-4cc6-9201-c76f646e7883 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | What would it mean to understand how a large language model (LLM) works? Some quick notes.
*Cross-posted from* [*New Savanna*](https://new-savanna.blogspot.com/2023/10/what-would-it-mean-to-understand-how.html)*.*
I don’t mean “understand” in any deep philosophical sense. I mean only a rough and ready sense of the w... |
abb34d02-452d-40d6-b864-d91df20dd6a2 | trentmkelly/LessWrong-43k | LessWrong | Life Extension versus Replacement
Has anyone here ever addressed the question of why we should prefer
(1) Life Extension: Extend the life of an existing person 100 years
to
(2) Replacement: Create a new person who will live for 100 years?
I've seen some discussion of how the utility of potential people fits into a ... |
241979eb-3f2c-4daf-8eed-e1b28142e924 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | [Intro to brain-like-AGI safety] 7. From hardcoded drives to foresighted plans: A worked example
*Part of the*[*“Intro to brain-like-AGI safety” post series*](https://www.alignmentforum.org/s/HzcM2dkCq7fwXBej8)*.*
*(This post substantially overlaps with my post from last August,* [*Value loading in the human brain: a... |
9e18ab0f-8fe6-4015-a0f3-a82a3dbe7781 | trentmkelly/LessWrong-43k | LessWrong | Are long-form dating profiles productive?
There's been a noticeable trend over the last few years of in this community of writing long-form dating profiles. They have the benefit of being slightly costly signals, requiring significantly more time and attention and a much greater degree of disclosure than can be devote... |
c5187bd4-9ec3-47f9-b254-b6be7e436d42 | trentmkelly/LessWrong-43k | LessWrong | Many people lack basic scientific knowledge
Intelligence varies more than it may appear. I tend to live and work with people near my own intelligence level, and so―probably―do you. I know there's at least two tiers above me. But there's even more tiers below me.
A Gallup poll of 1,016 Americans asked whether the Eart... |
0584985e-6fe1-40a0-b04f-4df63b970b4a | trentmkelly/LessWrong-43k | LessWrong | Crisis and opportunity during coronavirus
Note: please put on your own oxygen mask first. Don’t engage with this post if you haven’t taken appropriate measures to prepare yourself and your family; and plausibly don’t engage if you haven’t taken measures to ensure you can do so while staying stable and grounded.
We f... |
f8425cac-a2b6-479e-a31e-45ef533f8b58 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Dual use of artificial-intelligence-powered drug discovery
H/T [Aella](https://twitter.com/Aella_Girl/status/1503490760607469576).
A company that made machine learning software for drug discovery, on hearing about the security concerns for these sorts of models, asked: "huh, I wonder how effective it would be?" and w... |
8d69c159-5fe8-4edc-ab4c-d5b93e036cb7 | trentmkelly/LessWrong-43k | LessWrong | Needing Better PR
I've been having a bit of a back-and-forth with a friend about what appears to be a charisma problem with the SIAI, and was hoping you lovely folks had thoughts on the matter. My friend was going through the Eliezer Q&A videos, specifically Question #7, "What's your advice for Less Wrong readers who ... |
e842de74-3031-4a43-88fc-eb30cb1ab4d6 | trentmkelly/LessWrong-43k | LessWrong | Strong-Misalignment: Does Yudkowsky (or Christiano, or TurnTrout, or Wolfram, or…etc.) Have an Elevator Speech I’m Missing?
[Intro Note: (1) this "post" is, technically, a "question", for which there's a separate category on LW. I debated posting it as a "question", but it also involves considerable argumentation for... |
63a555c6-9eb7-4165-9304-a1feaad3530e | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | The Inner Alignment Problem
.mjx-chtml {display: inline-block; line-height: 0; text-indent: 0; text-align: left; text-transform: none; font-style: normal; font-weight: normal; font-size: 100%; font-size-adjust: none; letter-spacing: normal; word-wrap: normal; word-spacing: normal; white-space: nowrap; float: none; dir... |
c23ddc4f-dd1e-4262-928e-351e2c179fb3 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Review of 'But exactly how complex and fragile?'
I've thought about (concepts related to) the fragility of value [quite](https://www.lesswrong.com/s/7CdoznhJaLEKHwvJW/p/w6BtMqKRLxG9bNLMr#Objective_vs_value_specific_catastrophes) [a bit](https://www.lesswrong.com/posts/Xts5wm3akbemk4pDa/non-obstruction-a-simple-concept... |
8fce0b1c-fed3-4bd3-8a91-5422c61e91f0 | trentmkelly/LessWrong-43k | LessWrong | Deep limitations? Examining expert disagreement over deep learning
A recent publication by Carla Zoe Cremer, who's working at the Future of Humanity Institute:
> We conducted 25 expert interviews resulting in the identification of 40 limitations of the deep learning approach and 5 origins of expert disagreement. Thes... |
b6e86e59-b9c3-4a5e-bd3a-d0a125b2ebb9 | trentmkelly/LessWrong-43k | LessWrong | Alignment and Deep Learning
Between the recent AI breakthroughs and Eliezer's open admission of how bleak the chances of alignment are, everyone is speaking up and contributing what they can. It seems to me that there's a route that very few people are talking about that stands a better chance of successful alignment... |
d5dacb18-0175-4a60-b752-d39a9c36b7b6 | trentmkelly/LessWrong-43k | LessWrong | Food4Me - personalised nutrition initiative
I stumbled on this project during my research: food4me.org
The complete mapping of the human genome sequence in 2000 introduced the possibility of individualised medicine, including personalised nutrition. During this time the field of “nutrigenomics” emerged, which exa... |
421200ec-235f-4e67-b804-78837c090326 | trentmkelly/LessWrong-43k | LessWrong | Group Rationality Diary, March 16-31
This is the public group instrumental rationality diary for March 16-31.
> It's a place to record and chat about it if you have done, or are actively doing, things like:
>
> * Established a useful new habit
> * Obtained new evidence that made you change your mind about some be... |
c27c792e-3874-461a-9d43-ca7f91630d6b | trentmkelly/LessWrong-43k | LessWrong | Khan Academy: Introduction to programming and computer science
Khan Academy now also features a Computer Science category. There are not many lessons yet but about 3 new videos are being added each day. They are going to add CS exercises soon too.
If you don't want to wait for the exercises, there is always the incre... |
47016d8b-58ba-4c6a-b53a-6f99041e5774 | trentmkelly/LessWrong-43k | LessWrong | Analysis of COVID-19 superspread events (linkpost)
FYI, an analysis of COVID-19 superspread events is posted here. I just found it online; I was not involved with it.
Short version: The major mode of transmission (in such events, YMMV) seems to be ballistic projection of dropules while coughing, singing, or speaking ... |
a7b02505-b589-46ae-8cae-e8e2acfd6303 | StampyAI/alignment-research-dataset/arxiv | Arxiv | Machine Learning Model Development from a Software Engineering Perspective: A Systematic Literature Review
I Introduction
---------------
In Software Engineering (SE), researchers and practitioners have spent decades on developing tools and methodologies to create, manage and assemble complex software modules. Soft... |
a75bae9c-c302-48bf-ade9-867499235f5f | trentmkelly/LessWrong-43k | LessWrong | Munich Meetup, Saturday September 10th, 2PM
When: Saturday, September 10th, 2PM.
Where: Munich Central Station, Coffee Fellows cafe, Bahnhofsplatz 2, First floor. I will be there with a Lesswrong sign. If you can't find it, you can call me: 0160-93132663 .
According to this doodle-survey 5 people (including me) c... |
7fdb0422-6965-4e84-8054-87bdad1ca93b | trentmkelly/LessWrong-43k | LessWrong | Harry Potter and the Methods of Rationality discussion thread, March 2015, chapter 118
This is a new thread to discuss Eliezer Yudkowsky’s Harry Potter and the Methods of Rationality and anything related to it. This thread is intended for discussing chapter 118.
Plans for next chapter release:
> Ch. 119 (9800 words)... |
f0fe6585-03a4-467f-b408-853676dcbcd1 | StampyAI/alignment-research-dataset/blogs | Blogs | Robust Cooperation: A Case Study in Friendly AI Research
The paper “[Robust Cooperation in the Prisoner’s Dilemma: Program Equilibrium via Provability Logic](http://arxiv.org/abs/1401.5577)” is among... |
7d5b81d9-8fe4-401d-b41e-c6636424c0fe | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | A Critique of Non-Obstruction
Epistemic status: either I’m confused, or non-obstruction isn’t what I want.
This is a response to Alex Turner’s [Non-Obstruction: A simple Concept Motivating Corrigibility](https://www.alignmentforum.org/posts/Xts5wm3akbemk4pDa/non-obstruction-a-simple-concept-motivating-corrigibility).... |
9ef9d536-1d72-4212-818d-07d63ac2f3d1 | trentmkelly/LessWrong-43k | LessWrong | The Allais Paradox and the Dilemma of Utility vs. Certainty
Related to: The Allais Paradox, Zut Allais, Allais Malaise, and Pascal's Mugging
You've probably heard the Allais Paradox before, where you choose one of the two options from each set:
Set One:
1. $24000, with certainty.
2. 97% chance of $27000, 3% ch... |
7c4eea7f-7c10-4915-8306-160b7b60fc67 | trentmkelly/LessWrong-43k | LessWrong | Collapsing the Belief/Knowledge Distinction
There is no distinction between belief and knowledge.
Why?
Because whatever we think is knowledge is only, and can ever be, what we believe to be true. If we do not believe in it, we do not think it is true.
How do we get to truth? By virtue of our previously held beliefs... |
a66ba69e-b05b-42e9-a272-7201bc19ae74 | trentmkelly/LessWrong-43k | LessWrong | Babyeater's dilemma
Imagine it's the future, and everything has gone according to plan. Humanity has worked out its own utility function, f0, and has worked out a strategy S0 to optimize it.
Humanity has also run a large number of simulations of how alien worlds evolve. It has determined that of those civilizations w... |
43abbb6a-d3c8-497e-af6b-e23ea3acba79 | trentmkelly/LessWrong-43k | LessWrong | Mind Control and Me
Reading Eliezer Yudkowsky's works have always inspired an insidious feeling in me, sort of a cross between righteousness, contempt, the fun you get from understanding something new and gravitas. It's a feeling that I have found to be pleasurable, or at least addictive enough to go through all of hi... |
a4e8bc9b-022d-4ef9-b7c5-60e69179e46a | awestover/filtering-for-misalignment | Redwood Research: Alek's Filtering Results | id: 4d19ce85beab1502
Alignment faking in large language models \ Anthropic Skip to main content Skip to footer Claude API Solutions Research Commitments Learn News Try Claude Alignment Alignment faking in large language models Dec 18, 2024 Read the paper Most of us have encountered situations where someone appears to ... |
d1183a7b-f122-439d-8c52-8f6bebbc705b | StampyAI/alignment-research-dataset/blogs | Blogs | Conversation with Rohin Shah
AI Impacts talked to AI safety researcher Rohin Shah about his views on AI risk. With his permission, we have transcribed this interview.
**Participants**
----------------
* [Rohin Shah](https://rohinshah.com/) — PhD student at the Center for Human-Compatible AI, UC Berkeley
* Asya Ber... |
77fcad94-0e06-4874-a024-875125a5678e | trentmkelly/LessWrong-43k | LessWrong | Cleaning a SoundCraft Mixer
In Kingfisher Cecilia uses a condenser on her fiddle but combines it with effects pedals. This is a bit awkward to get working, and we've been using a Soundcraft Notepad mini-mixer. Unfortunately we had some issues recently where adjusting the input gain would make horrible scratchy noises:... |
c8723bbd-a0ae-443b-a886-7eb4915e0bdd | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Looking Deeper at Deconfusion
Introduction
============
My PhD thesis probably wins the prize of weirdest ever defended at my research lab. Not only was it a work of theory of distributed computing in a formal methods lab, but it didn’t even conform to what theory of distributed computing is supposed to look like. Wi... |
bba2dea4-e708-4727-9fb2-ac36eae91821 | trentmkelly/LessWrong-43k | LessWrong | Gradual takeoff, fast failure
Overview
This post aims to speculate on some concrete ways that things could go awry in a world of gradual takeoff, where AI capabilities increase smoothly and predictably on human timescales.
(At this point, "human timescales" are probably best measured in months or even weeks, as oppos... |
c9e27d05-29de-427e-89ee-a5255c0e591a | trentmkelly/LessWrong-43k | LessWrong | [Link and commentary] The Offense-Defense Balance of Scientific Knowledge: Does Publishing AI Research Reduce Misuse?
This is (partly) a linkpost for a paper published earlier this year by Toby Shevlane and Allan Dafoe, both researchers affiliated with the Centre for the Governance of AI. Here’s the abstract:
> There... |
ae25f364-07fa-485d-ae50-e975b231ba3e | trentmkelly/LessWrong-43k | LessWrong | Kindergarten in NYC: Much More than You Wanted to Know
Kindergarten in NYC: Much More than You Wanted to Know
My son is turning five next year, which means one of the most important transitions in his childhood and potentially his life: starting Kindergarten. I always thought New York City moms who obsessed over th... |
665a61e4-e342-4996-9deb-52f8917d6e59 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Search versus design
*This work was supported by OAK, a monastic community in the Berkeley hills. It could not have been written without the daily love of living in this beautiful community. The work involved in writing this cannot be separated from the sitting, chanting, cooking, cleaning, crying, correcting, fundrai... |
04f48a57-5a2c-4be8-9069-96e5c94f25c6 | trentmkelly/LessWrong-43k | LessWrong | Some thoughts on meta-probabilties
I often like to think of my epistemic probability assignments in terms of probabilities-of-probabilities, or meta-probabilities. In other words, what probability would I assign that my probability estimate is accurate? Am I very confident, am I only mildly confident, or do I only hav... |
e13b34af-51ca-4ddd-ba57-5ecf76fc0f67 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Distribution Shifts and The Importance of AI Safety
Preface
=======
The following text is my submission for the [AI Safety Public Materials contest](https://www.lesswrong.com/posts/gWM8cgZgZ9GQAYTqF/usd20k-in-bounties-for-ai-safety-public-materials). In it, I try to lay out the importance of AI Safety Research to peo... |
1ef4cb08-5400-49a9-a7d6-fa462f209a05 | LDJnr/LessWrong-Amplify-Instruct | LessWrong | "When I started studying the art of studying, I wanted to understand the role of book learning. How do we best learn from a textbook, scientific article, or nonfiction book? What can a student of average intelligence do to stay on top of their homework? Is it possible to improve your annual knowledge growth rate by one... |
c4ba14b3-e570-4800-827d-ee54b6778496 | trentmkelly/LessWrong-43k | LessWrong | Adaptive Logics: the best of both worlds?
Consider the following premises from which we would like to reason:
1. Abraham Lincoln was born in Michigan.
2. Abraham Lincoln was born in Kentucky.
3. Eliezer Yudkowsky has a beard or he has blond hair.
4. Eliezer Yudkowsky doesn't have blond hair.
We'd like to be able... |
c32840ff-0393-4992-a8b4-4e414640bf4f | trentmkelly/LessWrong-43k | LessWrong | Are the effects of UFAI likely to be seen at astronomical distances?
My comment to a discussion of great filters/existential risk:
How likely is it that a UFAI disaster would produce effects we can see from here? I think "people can't suffer if they're dead" disasters (failed attempt at FAI) is possibly more likely t... |
8f9f45ff-c234-4e60-a467-95ed9a1a70ea | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | On unfixably unsafe AGI architectures
There's loads of discussion on ways that things can go wrong as we enter the post-AGI world. I think an especially important one for guiding current research is:
*Maybe we'll know how to build **unfixably** unsafe AGI, but can't coordinate not to do so.*
As a special case, I w... |
ba0d79c6-e24c-4289-b284-500a1cafaa79 | trentmkelly/LessWrong-43k | LessWrong | How do you whittle path-dependent (probabilistic) questions?
It's a skill I lack and moreso, a desire I find myself lacking; to structure discussion in such a way that the later probability estimates don't contaminate the earlier ones they really shouldn't.
For example, I am reading Zvi's post on buying/selling/holdi... |
9e6b70b4-0c22-4b19-a81a-dbecb82fe6eb | StampyAI/alignment-research-dataset/lesswrong | LessWrong | There have been 3 planes (billionaire donors) and 2 have crashed
I'd rather not go into the details about which billionaires are which, so if it's actually 4 and 3 or 6 and 4, then that may or may not be debatable. I'm much more worried about whether MIRI survives the decade.
It seems to me like this is a good place ... |
7898c8a6-309b-406f-ad67-7fcb41b1ae59 | StampyAI/alignment-research-dataset/arxiv | Arxiv | Learning in two-player games between transparent opponents
1 Introduction
---------------
Reinforcement learning is concerned with constructing agents that learn to achieve high rewards in a diverse set of environments. Multi-agent reinforcement learning studies learning in environments that contain other agents, p... |
bcb2dfd8-4d94-4c5f-8a43-2b72217edfac | trentmkelly/LessWrong-43k | LessWrong | Explanations as Hard to Vary Assertions
Update: After some investigation, I found out that The Beginning of Infinity by David Deutsch contains a few minor misquotes of Popper, Turing and others. Nevertheless, it is an excellent book.
----------------------------------------
Background
As I read through Rationality... |
a382f429-5f2b-4deb-84d6-3784d0b879a8 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Melbourne, practical rationality
Discussion article for the meetup : Melbourne, practical rationality
WHEN: 05 April 2013 07:00:00PM (+1100)
WHERE: 55 Walsh Street, West Melbourne VIC 3003, Australia
Practical rationality. This meetup repeats on the 1st Friday of each month and is distinct from our social ... |
5b39f56b-ac23-4709-a583-6a50a35ac5a9 | trentmkelly/LessWrong-43k | LessWrong | On AI and Compute
This is a post on OpenAI’s “AI and Compute” piece, as well as excellent responses by Ryan Carey and Ben Garfinkel, Research Fellows at the Future of Humanity Institute. (Crossposted on the EA Forum)
Intro: AI and Compute
Last May, OpenAI released an analysis on AI progress that blew me away. The ke... |
2944ba02-9ad5-4a68-8f3e-3fd54c791e2e | trentmkelly/LessWrong-43k | LessWrong | Some questions about free will compatibilism
Edit to add:
I've read more about compatibilism, including the comments on the original post, and I think it's all been quite helpful in enhancing my understanding. The post and comments here were particularly illuminating: https://whyevolutionistrue.com/2021/04/30/why-do-... |
8210d52b-623f-4ee5-b28e-0d1034ab7b5b | trentmkelly/LessWrong-43k | LessWrong | Strategies and tools for getting through a break up
Background:
I was very recently (3 weeks now) in a relationship that lasted for 5.5 years. My partner had been fantastic through all those years and we were suffering no conflict, no fights, no strain or tension. My partner also was prone to depression, and is/was go... |
a9d3856c-4b79-4911-b0cd-9270a64438fb | trentmkelly/LessWrong-43k | LessWrong | The questions and classes of SSA
Important summary: For SSA, the reference class to be used is defined in the question asked.
I said that SSA derives from the following question:
* Q: What proportion of universes, where versions of me exist, will be large?
And it's sort of true, that sort of is the SSA question. B... |
f96ca60d-c651-44a7-90b6-6383e9724aef | trentmkelly/LessWrong-43k | LessWrong | The Wrongness Iceberg
As soon as I got out of college I got a job at a restaurant. At the time I had never had a job at a restaurant, but my mom had known the owners and I felt obligated to avoid performing badly. Yet inevitably I did perform badly, and how this performance was evaluated would greatly affect my way of... |
fba0bae2-ce36-4c65-80bb-3395a95539aa | trentmkelly/LessWrong-43k | LessWrong | [Link] St. Paul: memetic engineer
New post by Federico on his new blog that I mentioned earlier. Worth reading for those interested in the memetics of religion, politics, Christianity or Islam. The cited material also led my mind to some related questions.
> AnnoDomini suggests I write about “St. Paul the social engi... |
9ef2a123-759c-40ff-8095-f3121c73d58d | trentmkelly/LessWrong-43k | LessWrong | Resources For the Redwood Research Triplebyte Test
After working for a few years as a web developer, I am now applying to the MLAB2 AI Alignment Boot Camp. I'm very excited about this, and studying for the required coding test, the Triplebyte assessment for Redwood Research. Does anyone have any suggestions for good... |
7e2dab6e-ca31-4101-9f7b-d87685af2529 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Incident reporting for AI safety
*Zach Stein-Perlman, Sebastian Lodemann, Štěpán Los, and Moritz von Knebel*
*We made this (quickly) because a resource on* incident reporting as a tool for AI safety *should exist; we are not experts; suggestions are welcome.*
*Zach recommends engaging with this post by reading §1 an... |
197c5f87-b4c8-4a0f-aaa6-737c910ff023 | trentmkelly/LessWrong-43k | LessWrong | Bayesian Charity
Cross-posted from Living Within Reason
> In philosophy, the Principle of Charity is a technique in which you evaluate your opponent’s position as if it made the most amount of sense possible given the wording of the argument. That is, if you could interpret your opponent’s argument in multiple ways, ... |
5936bdc4-b6ee-47a0-849f-4c360245dcdb | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Toward a New Technical Explanation of Technical Explanation
A New Framework
===============
*(Thanks to Valentine for a discussion leading to this post, and thanks to CFAR for running the CFAR-MIRI cross-fertilization workshop. Val provided feedback on a version of this post. Warning: fairly long.)*
Eliezer's *A [Te... |
6ef31b16-879d-453b-b9fe-81ef16ad0943 | trentmkelly/LessWrong-43k | LessWrong | Should you try to do good work on LW?
I used to advocate trying to do good work on LW. Now I'm not sure, let me explain why.
It's certainly true that good work stays valuable no matter where you're doing it. Unfortunately, the standards of "good work" are largely defined by where you're doing it. If you're in academi... |
2f2cc73e-f64a-4266-bf22-00d2109a8342 | LDJnr/LessWrong-Amplify-Instruct | LessWrong | "Let us say you are a doctor, and you are dealing with a malaria epidemic in your village. You are faced with two problems. First, you have no access to the drugs needed for treatment. Second, you are one of two doctors in the village, and the two of you cannot agree on the nature of the disease itself. You, having car... |
85a5a58d-8a0c-4088-92a3-3453ab1ab243 | trentmkelly/LessWrong-43k | LessWrong | Lightweight minimal speech recognition?
I currently control my rhythm stage setup with a USB keyboard. This works, except that my hands are often busy playing instruments when I want to change something. I already have a microphone in front of my mouth running to a computer, which I use for my whistle-controlled bass ... |
ca02fc56-68ad-459c-996c-a8d9d88b83b4 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Underappreciated points about utility functions (of both sorts)
(**Edit** May 9th: I've gone and added a quick addendum to the end.)
In this post I'd basically like to collect some underappreciated points about utility functions that I've made in the comments of various places but which I thought were collecting int... |
204d6def-ebf2-4778-8383-c2b6d6437d0b | StampyAI/alignment-research-dataset/lesswrong | LessWrong | The idea of an "aligned superintelligence" seems misguided
For a while I’ve been reading LessWrong, and there’s something that doesn’t make much sense to me, which is the idea that it is possible to align an AGI/superintelligence at all. I understand that it’s probably not even a majority of discussion on LW that is o... |
58bc142b-b102-4dcd-8847-8af87a69a2d3 | trentmkelly/LessWrong-43k | LessWrong | Short Timelines Don't Devalue Long Horizon Research
Short AI takeoff timelines seem to leave no time for some lines of alignment research to become impactful. But any research rebalances the mix of currently legible research directions that could be handed off to AI-assisted alignment researchers or early autonomous A... |
e85593ed-b23b-49fe-8dd3-8b65ace7e1a8 | trentmkelly/LessWrong-43k | LessWrong | Notes on Dwarkesh Patel’s Podcast with Sholto Douglas and Trenton Bricken
Dwarkesh Patel continues to be on fire, and the podcast notes format seems like a success, so we are back once again.
This time the topic is how LLMs are trained, work and will work in the future. Timestamps are for YouTube. Where I inject my o... |
bfee51df-d092-4bdc-9325-67ceb041158b | StampyAI/alignment-research-dataset/blogs | Blogs | “Non-Player Character”
Rilanya: “You’re not like the others, are you?”
Darin: “What do you mean?”
Rilanya: “I… do you know why I first fell in love with you?”
Darin: “For my good looks?”
Rilanya: “My whole life I’ve felt so alone. The people around me… they just seemed to be going through the motions. Like the... |
e97b608c-b138-4049-8cf9-12b60692cb90 | trentmkelly/LessWrong-43k | LessWrong | Enemies vs Malefactors
Status: some mix of common wisdom (that bears repeating in our particular context), and another deeper point that I mostly failed to communicate.
Short version
Harmful people often lack explicit malicious intent. It’s worth deploying your social or community defenses against them anyway. I rec... |
820beb27-da7a-4ac6-8008-c009170d12aa | trentmkelly/LessWrong-43k | LessWrong | Work culture creep
epistemic status: This is not my field of study. I am working through some ideas here that I think might resonate with the community.
I am a fun, happy, goofy dad. I can play music and perform magic tricks. I sing silly songs and make stupid jokes.
However, for 8 out of the 16 hours that I am awak... |
1b7ec0fb-cc22-4269-9c61-568420c26bc0 | trentmkelly/LessWrong-43k | LessWrong | the devil's ontology
imagine you’re playing a game where some rules encoded in words are so special that no one is allowed to touch or change them. these special rules have special properties, so no one questions them. the devil, in this case, is like a sneaky player who hides behind those rules, using them to confuse... |
8886ebd1-cca7-4163-83cf-7ffa6796cbf0 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Full-time AGI Safety!
Hi everyone!
I'm Steve Byrnes, a professional physicist in Boston. Y'all may know me as a regular blogger and commenter on this site. ([Example post](https://www.alignmentforum.org/posts/wcNEXDHowiWkRxDNv/inner-alignment-in-salt-starved-rats), [Full list](https://sjbyrnes.com/agi.html).)
Well, ... |
fcabf024-3720-4e73-a236-421c8b57c486 | trentmkelly/LessWrong-43k | LessWrong | 2023 Alignment Research Updates from FAR AI
TL;DR: FAR AI's science of robustness agenda has found vulnerabilities in superhuman Go systems; our value alignment research has developed more sample-efficient value learning algorithms; and our model evaluation direction has developed a variety of new black-box and white-... |
f9a0513c-49f9-4507-8c99-93f84c307c42 | trentmkelly/LessWrong-43k | LessWrong | Weekly LW Meetups
This summary was posted to LW main on June 21st. The following week's summary is here.
Irregularly scheduled Less Wrong meetups are taking place in:
* Atlanta: Self-Hacking: 23 June 2013 06:00PM
* Bratislava Meetup IV.: 24 June 2013 06:00PM
* Brussels meetup with HEALES: 13 July 2013 01:00PM
* ... |
d184b158-fa7b-4fbc-8e6d-749afef61a04 | trentmkelly/LessWrong-43k | LessWrong | On silence
It has become common knowledge that some things—food, porn, outrage-share inducing articles—exploit a mismatch between our evolutionary heritage and the modern world, and do so at our expense. It seems, however, that we’ve collectively neglected one particular contemporary, omnipresent superstimulus: sound.... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.