id stringlengths 36 36 | source stringclasses 15
values | formatted_source stringclasses 13
values | text stringlengths 2 7.55M |
|---|---|---|---|
64991462-48f1-4db5-b3eb-da74068e8205 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Durham HPMoR Discussion, chapters 30-33
Discussion article for the meetup : Durham HPMoR Discussion, chapters 30-33
WHEN: 26 January 2013 11:00:00AM (-0500)
WHERE: Foster's Market, 2694 Durham-Chapel Hill Blvd., Durham, NC
Our regular HPMoR reading meetup. Please feel free to come, even if you haven't done... |
6620b15f-f3ac-4c9b-89a2-c5edaa55b418 | trentmkelly/LessWrong-43k | LessWrong | Open thread, Dec. 15 - Dec. 21, 2014
If it's worth saying, but not worth its own post (even in Discussion), then it goes here.
Previous Open Thread
Next Open Thread
----------------------------------------
Notes for future OT posters:
1. Please add the 'open_thread' tag.
2. Check if there is an active Open Threa... |
5aab11ae-4d3d-4976-9f87-e49544017406 | trentmkelly/LessWrong-43k | LessWrong | Idle Speculations on Pipeline Parallelism
Epistimic status: Pure speculation; I have 1 GPU, not dozens, so I have no reason to try and implement this stuff myself.
When training large models, we might have many GPUs we'd like to use to speed the process up. This requires a way to parallelize training somehow. Some op... |
4635b00d-3521-4594-bfd2-e5d80146960e | trentmkelly/LessWrong-43k | LessWrong | Marvel Snap: Phase 1
Usual game approach. I’ll start off spoiler-free, then proceed to the semi-spoiled section, then to the Here Be Spoilers section.
MAXIMALLY SPOILER-FREE (1 BIT OF INFO): SHOULD YOU PLAY?
No.
SPOILER-FREE (3 BITS OF INFO): WHAT TIER IS IT?
Tier 3. It is a good game.
I want to be able to put th... |
a6108033-e32e-411f-9ff4-026480f77e4e | trentmkelly/LessWrong-43k | LessWrong | Rationality Outreach: A Parable
This post grew out of a very long discussion with the New York Less Wrong meetup group. The question was, should a group dedicated to rationality be explicitly atheist? Or should it make an effort to be respectful to theists in order to make them feel welcome and spread rationality fa... |
15f769a6-a84a-49c6-8f3d-d8d40478b640 | StampyAI/alignment-research-dataset/special_docs | Other | Safe Exploration Techniques for Reinforcement Learning – An Overview
Lecture Notes in Computer Science: Safe
Exploration Techniques for Reinforcement
Learning { An Overview
Martin Pecka1?, and Tomas Svoboda1;2??
1Center for Machine Perception, Dept. of Cybernetics, Faculty of Electrical
Engineering, Czech Technical Un... |
09b27944-a879-441a-848e-a962b67087c4 | StampyAI/alignment-research-dataset/arxiv | Arxiv | Building Human Values into Recommender Systems: An Interdisciplinary Synthesis.
Building Human Values into Recommender Systems:
An Interdisciplinary Synthesis
Jonathan Stray
1
, Alon Halevy
2
, Parisa Assar
2
, Dylan
Hadfield-Menell
3
, Craig Boutilier
4
,
Amar
Ashar
5
, Lex Beattie
5
, Michael Ekstrand
6
, Claire Lei... |
5b121230-6f56-4f25-8da2-d2010617a1da | trentmkelly/LessWrong-43k | LessWrong | European Community Weekend 2015 Impressions Thread
The European Community Weekend in Berlin is over and was plain awesome.
This is no complete report of the event but a place where you can e.g. comment on the event, link to photos or what else you want to share.
I'm not the organizer of the Meetup but I have been th... |
cb4e91ac-c21a-4058-804d-15620ef01da8 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | A potential problem with using Solomonoff induction as a prior
There's a problem that has occurred to me that I haven't seen discussed anywhere: I don't think people actually wants to assign zero probability to all hypotheses which are not Turing computable. Consider the following hypothetical: we come up with a theo... |
4aaa3690-24d1-411a-94e2-69282cdb7f33 | awestover/filtering-for-misalignment | Redwood Research: Alek's Filtering Results | id: post1131
Although this appears to be some sort of fictional role-playing server, I quite like the explanation of corrigibility given. It is clearly a direct reference to the AI alignment problem. They even included a tl;dr, which I included at the bottom of the post . It covers a lot of aspects of the alignment pr... |
d4f01075-4db6-47b5-82a0-53baf89e3c44 | trentmkelly/LessWrong-43k | LessWrong | Estimating Tail Risk in Neural Networks
Machine learning systems are typically trained to maximize average-case performance. However, this method of training can fail to meaningfully control the probability of tail events that might cause significant harm. For instance, while an artificial intelligence (AI) assistant ... |
45a72101-a816-45e2-9ebf-f233996b5a6b | trentmkelly/LessWrong-43k | LessWrong | A Novice Buddhist's Humble Experiences
This is an introduction and description of vipassana meditation [edit: actually, anapanasati, not vipassana as such] more than Buddhism. Nonetheless I hope it serves as some testament to the value of Buddhist thought outside of meditation.
One day I hope more people take up the ... |
f9d2f89e-5453-4201-bcab-f47f579a2ab2 | trentmkelly/LessWrong-43k | LessWrong | Causal vs Predictive Models, and the Causal Taboo
(I wrote this post in April 2020 for a non-LW audience)
Causation is pretty cool. Even cooler than causation, causal models! If you haven't heard the news, the past few decades have produced big leaps in understanding causality and how to reason about it. There's also... |
26c5774a-a72e-41f5-bc5b-ac41dcc2753c | trentmkelly/LessWrong-43k | LessWrong | [Survey] Expectations of a Post-ASI Order
I put together a survey to explore what people think will happen after the emergence of Artificial Super Intelligence, assuming that humans survive and there is something resembling human society afterwards.
Take the survey here. Thanks!
Twenty four people from /r/singularit... |
bbf9e66f-be7e-42db-8a58-abf2bf67347f | trentmkelly/LessWrong-43k | LessWrong | [AN #124]: Provably safe exploration through shielding
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.
... |
f88ac763-f3e9-40dc-81b0-8a4e56947fa7 | trentmkelly/LessWrong-43k | LessWrong | A Game About AI Alignment (& Meta-Ethics): What Are the Must Haves?
I'm working on a pc & mobile game about metaethics, ethics & AI alignment. (Our steam page + announcement teaser will be up in 1-2 weeks.) It's important to me that we nail the AI alignment part and give people a good idea about why AI alignment is ha... |
567c741b-cc6b-45e3-bc92-1f8e9e7dc637 | trentmkelly/LessWrong-43k | LessWrong | Natural Latents: The Math
Natural latents are a relatively elegant piece of math which we figured out over the past year, in our efforts to boil down and generalize various results and examples involving natural abstraction. In particular, this new framework handles approximation well, which was a major piece missing ... |
9d534fd5-8797-435b-b849-1fdad92992f7 | trentmkelly/LessWrong-43k | LessWrong | ask me about technology
I have a lot of opinions about technology - what's feasible, what isn't, what the best ways to do things are, and so on. Sometimes I write them down on my blog, but usually I don't. I also know some people who are good at designing technology.
If you want my personal opinion on some technology... |
683ce1c2-217c-4805-b826-5ffd424065ea | trentmkelly/LessWrong-43k | LessWrong | The alignment problem from a deep learning perspective
This report (now available on arxiv) is intended as a concise introduction to the alignment problem for people familiar with machine learning. It translates previous arguments about misalignment into the context of deep learning by walking through an illustrative ... |
2c908ac4-ce9f-4177-a686-3da6e0e95b25 | trentmkelly/LessWrong-43k | LessWrong | How to estimate confidence intervals for fermi estimate?
Suppose you want to take a guess at the number of candyfloss sold over a month in some area and you would like a 90% confidence interval(CI) intead of a point estimate. You fermi estimates for two central subcomponents are:
(1) a 90% CI of the number of candyfl... |
32814808-0590-4436-8a9c-530261b2747e | trentmkelly/LessWrong-43k | LessWrong | Rationalist position towards lying?
What is the rationalist position towards lying? Is it uniformly wrong and reprehensible? Lying can sometimes be a good way to optimize efficiency in achieving morally commendable, broadly altruistic goals. Not all people are necessarily equipped to handle the truth properly, in the ... |
64873811-31ed-4873-b201-26f26d01dbc1 | trentmkelly/LessWrong-43k | LessWrong | Another AI Winter?
Robin Hanson has been suggesting recently that we've been experiencing an AI boom that's not too different from prior booms.
At the recent Foresight Vision Weekend, he predicted [not exactly - see the comments] a 20% decline in the number of Deepmind employees over the next year (Foresight asked al... |
2d097e5d-f333-4743-9d10-a8543c242c04 | trentmkelly/LessWrong-43k | LessWrong | Open Problems Create Paradigms
A few weeks ago I had a fascinating conversation with Ruby about models of the research process and how to improve it. This post outlines one particular model which I took away from that conversation: open problems as the primary factor which create a paradigm.
There’s a cluster of thin... |
fffead25-c029-41f3-bd01-3dcfef5ac22a | trentmkelly/LessWrong-43k | LessWrong | Benefits of adding poison to your DMT
DMT might be one of the best psychedelics out there, its effects are short-lived, it can be quite intense, and it allows for careful modulation of the dosage through trial and error.
With most psychedelics, working towards a dosage that won’t induce a difficult experience (read: ... |
323b5f20-96d4-479d-b13d-77bd814b2c28 | trentmkelly/LessWrong-43k | LessWrong | The World in 2029
Links are to prediction markets, subscripts/brackets are my own forecasts, done rapidly.
I open my eyes. It’s nearly midday. I drink my morning Huel. How do I feel? My life feels pretty good. AI progress is faster than ever, but I've gotten used to the upward slope by now. There has perhaps recentl... |
d1682396-effc-4269-8ac4-a1a599597918 | trentmkelly/LessWrong-43k | LessWrong | My Elevator Pitch for FAI
This is a short introduction to the idea of FAI and existential risk from technology that I've used with decent success among my social circles, which consist mostly of mathematicians or at least people who have taken an introductory CS class.
I'll do my best to dissect what I think is ef... |
30045d4b-c57c-4216-9a30-3d0afdb71453 | trentmkelly/LessWrong-43k | LessWrong | How Might an Alignment Attractor Look like?
It is a generally accepted view in MIRI and MIRI-adjacent circles that the odds are high that an eventual self-improving AGI will not be nice to humans, to put it mildly. It is actually becoming a bit of a zeitgeist, gaining more and more widespread support. There are plenty... |
b2337b0e-4fcb-4c34-8c9f-052d7ea558b7 | trentmkelly/LessWrong-43k | LessWrong | How much do markets value Open AI?
Summary: A BOTEC indicates that Open AI might have been valued at 220-430x their annual recurring revenue, which is high but not unheard of. Various factors make this multiple hard to interpret, but it generally does not seem consistent with investors believing that Open AI will capt... |
c78a3c60-2b7e-469d-8d67-c56b33216259 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Donation Decision Day
Discussion article for the meetup : Donation Decision Day
WHEN: 26 December 2015 01:00:00PM (-0800)
WHERE: 425 Harvard Ave E, Seattle, Washington, 98102
Come and work on your personal donation decisions; with a partner or by yourself. Do research, run the numbers, and if you run into ... |
f491e6f1-78ce-4898-88af-03a03ae56250 | trentmkelly/LessWrong-43k | LessWrong | Proposal: Using Monte Carlo tree search instead of RLHF for alignment research
Currently the most powerful techniques for getting a language model to act as an agent is via RLHF and similar approaches. For example, ChatGPT was trained to be an agent that tries to give humans answers that they want. Another approach is... |
754815af-003a-4f61-902b-74e660325f0c | trentmkelly/LessWrong-43k | LessWrong | Sorry for being less around, will be back soon
Hey everyone,
Some of you might have noticed that the development speed was a bit slower over the last two weeks. I had a bunch of not so great things happen to my family that have caused me a good amount of stress. They've also caused to me fly on fairly short notice ba... |
76483c00-28d6-460c-af22-68e4c4a620d6 | trentmkelly/LessWrong-43k | LessWrong | AI Disclosure Ballot Initiative (and voting method)
I shared that I'm helping to raise funds for a ballot initiative earlier (it also includes approval voting), but some of you may be interested that this initiative has a component requiring disclosure for AI generation in campaign ads. This could be the first time AI... |
ffe7be7e-05be-4624-9eaf-b228e9f2f1dd | trentmkelly/LessWrong-43k | LessWrong | Sam Altman's sister claims Sam sexually abused her -- Part 3: Timeline, continued
Previous posts (which you should read first)
This post is the 3rd post in a series of 11 posts about the claims of Sam Altman's sister, Annie Altman. Annie has claimed that Sam sexually abused her for about 9 years as a child, and that s... |
c548f5bc-ecc0-4393-b68e-098b38a15535 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | AI Safety For Dummies (Like Me)
Summary
=======
* **Progress in AI Safety can reduce current risks** ***and*** **existential risks.**
+ Ex: current risk - automated military drones.
+ Ex: existential risk - automated bioweapons research.
* **Risk = probability x severity x vulnerability x exposure**. (rule of thumb... |
6bd3bfad-a002-4d0b-a33c-9830dc71c029 | trentmkelly/LessWrong-43k | LessWrong | Mechanism Design for AI
None |
486dc5e7-48d3-430c-944d-35d35371db70 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | OpenAI Launches Superalignment Taskforce
In their announcement [Introducing Superalignment](https://openai.com/blog/introducing-superalignment), OpenAI committed 20% of secured compute and a new taskforce to solving the technical problem of aligning a superintelligence within four years. Cofounder and Chief Scientist ... |
75d7cbc8-a696-48a1-96ef-7e603f1cec0c | trentmkelly/LessWrong-43k | LessWrong | Christiano (ARC) and GA (Conjecture) Discuss Alignment Cruxes
The following are the summary and transcript of a discussion between Paul Christiano (ARC) and Gabriel Alfour, hereafter GA (Conjecture), which took place on December 11, 2022 on Slack. It was held as part of a series of discussions between Conjecture and p... |
a45ab3ae-277c-4639-955c-c135ac63286a | trentmkelly/LessWrong-43k | LessWrong | Understanding the Evidence for Killer Supplements
Related to: Even if You Have a Nail Supplements Kill
> So what test, exactly, did the authors perform? And what do the results mean? It remains a mystery to me - and, I'm willing to bet, to every other reader of the paper.
I'm willing to take that bet. In Marc... |
18823981-006e-468d-a2ac-7c1a57692a2e | trentmkelly/LessWrong-43k | LessWrong | OpenAI Claims Nonprofit Will Retain Nominal Control
Your voice has been heard. OpenAI has ‘heard from the Attorney Generals’ of Delaware and California, and as a result the OpenAI nonprofit will retain control of OpenAI under their new plan, and both companies will retain the original mission.
Technically they are no... |
a61caa07-dca2-4c4b-be40-2027805dbb59 | trentmkelly/LessWrong-43k | LessWrong | Is IQ what we actually need to know?
I've never heard of anyone saying "I thought that person was really intelligent, but they turned out not to be", and when there are scandals about people with fake credentials, they don't seem to come from people with fake credentials making mistakes-- instead, someone checks the h... |
4a8e07a7-9880-493c-9867-05e87bc05a2e | trentmkelly/LessWrong-43k | LessWrong | Catastrophic Risks from AI #4: Organizational Risks
This is the fourth post in a sequence of posts giving an overview of catastrophic AI risks.
4 Organizational Risks
In January 1986, tens of millions of people tuned in to watch the launch of the Challenger Space Shuttle. Approximately 73 seconds after liftoff, the ... |
7a644d8b-36ae-4399-b447-8dbb7a62fc91 | StampyAI/alignment-research-dataset/agentmodels | Tutorial: Modeling Agents with Probabilistic Programs | Modeling Agents with Probabilistic Programs
---
layout: chapter
title: Joint inference of biases and preferences I
description: Assuming agent optimality leads to mistakes in inference. Procrastination and Bandit Examples.
---
### Introduction
Techniques for inferring human preferences and beliefs from their behavi... |
77775dbf-5333-4585-a942-c9e22f60cada | trentmkelly/LessWrong-43k | LessWrong | Reflections on the PIBBSS Fellowship 2022
Last summer, we ran the first iteration of the PIBBSS Summer Research Fellowship. In this post, we share some reflections on how the program went.
Note that this post deals mostly with high-level reflections and isn’t maximally comprehensive. It primarily focusses on informat... |
ff7df2e1-2d45-4e99-8e16-df4df1e865a2 | trentmkelly/LessWrong-43k | LessWrong | Open Thread Summer 2024
If it’s worth saying, but not worth its own post, here's a place to put it.
If you are new to LessWrong, here's the place to introduce yourself. Personal stories, anecdotes, or just general comments on how you found us and what you hope to get from the site and community are invited. This is a... |
ae2b6d8e-a1ef-4090-a1a1-426b05c6a98b | LDJnr/LessWrong-Amplify-Instruct | LessWrong | "Related to: Extreme Rationality: It's Not That Great
A while back, I said provocatively that the rarefied sorts of rationality we study at Less Wrong hadn't helped me in my everyday life and probably hadn't helped you either. I got a lot of controversy but not a whole lot of good clear examples of getting some use out... |
00e4fb25-428f-4713-9b6a-5d0d2b5e4ba6 | trentmkelly/LessWrong-43k | LessWrong | Weekly LW Meetups: Berlin, Buenos Aires, Cambridge MA, Pittsburgh
There are upcoming irregularly scheduled Less Wrong meetups in:
* Pittsburgh: Making Beliefs Pay Rent: 01 June 2012 06:00PM
* Third Buenos Aires Meetup: 2 June 2012 4:00PM: 02 June 2012 04:00PM
* First Berlin meetup: 05 June 2012 07:30PM
* Phoenix,... |
cda35555-1978-4baf-a2cf-a89f53012b49 | trentmkelly/LessWrong-43k | LessWrong | Imitation is the Sincerest Form of Argument
I recently gave a talk at Chicago Ideas Week on adapting Turing Tests to have better, less mindkill-y arguments, and this is the precis for folks who would prefer not to sit through the video (which is available here).
Conventional Turing Tests check whether a programmer ca... |
ef75522f-69d2-473a-afac-8e587a5dea04 | trentmkelly/LessWrong-43k | LessWrong | D&D.Sci Divination: Nine Black Doves Evaluation & Ruleset
This is a follow-up to last week's D&D.Sci scenario: if you intend to play that, and haven't done so yet, you should do so now before spoiling yourself.
There is a web interactive here you can use to test your answer, or you can read on.
RULESET
Map
Provin... |
32e05228-642f-44ce-9b4d-dff2a1e7514d | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Web AI discussion Groups
After the success of Web Taisu, I have decided to organize a similar event.
Here are the rules, this document is an attempt to create common knowledge among the participants.
If you have an Idea for a topic to discuss
-------------------------------------------
Post it to reddit on <https... |
b7095373-d6b7-46c4-803a-54d45a69f0c8 | trentmkelly/LessWrong-43k | LessWrong | Markets are a kind of electrochemical cell
There are two processes taking place: adding and using up value from units. When everything is mixed together and these processes are happening in one place, they happen slowly (think of subsistence production by consumers). Separate them to their own containers and they happ... |
62276f83-5ef7-4e98-be74-fa7bc5334dcc | trentmkelly/LessWrong-43k | LessWrong | Useful Questions Repository
See also: Boring Advice Repository, Solved Problems Repository, Grad Student Advice Repository, Useful Concepts Repository, Bad Concepts Repository
I just got back from the July CFAR workshop, where I was a guest instructor. One useful piece of rationality I started paying more attention t... |
9787c2ff-d2e2-4ff9-b5a6-413ae3258838 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | A Double-Feature on The Extropians
Link-post for two pieces I just wrote on the Extropians.
The Extropians were an online group of techno-optimist transhumanist libertarians active in the 90s who influence a lot of online intellectual culture today. Prominent members include Eliezer Yudkowsky, Nick Bostrom, Robin... |
8cbb3d8a-0880-485c-b2b8-61f3d91fd4b6 | trentmkelly/LessWrong-43k | LessWrong | How likely is an attempted coup in the United States in the next four years?
Trump cares a lot about personal power. He does not (to put it lightly) seem to have much respect for tradition or the rule of law. He has 'joked' about a third term before.
Many of the traditional safeguards are disappearing: Trump has fire... |
08ae1ff6-f0f6-478e-9bde-faba7afdeb26 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Null-boxing Newcomb’s Problem
The leering statue of the unnamed trickster god stared down at him from its sacred alcove, motionless as always.
Maxwell sighed. He had been praying at the base of the statue for the past hour, and like every time previous, he hadn’t felt anything. The priests had promised salvation from ... |
22063f0b-b14d-4df3-bd18-3fd5ead1e7fd | LDJnr/LessWrong-Amplify-Instruct | LessWrong | "Epistemic status: pretty much factual.
CW: diets, calories
One of the odd things about working on longevity is that now people ask me for lifestyle advice.
Or they ask me what I, personally do to live longer.
Mostly my response has been a lame “um, nothing?”
There are, as of now, no interventions shown to make humans ... |
6fcaba80-3618-427e-af2e-0a0c6a421d47 | trentmkelly/LessWrong-43k | LessWrong | What life will be like for humans if aligned ASI is created
Let's assume computationalism and the feasibility of brain scanning and mind upload. And let's suppose one is a person with a large compute budget.
In this post I'll attempt to answer these quetions: How should one spend one's compute budget? How many uploa... |
dfae73b1-a426-44cb-854d-b1350acfd905 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Sydney Meetup - March
Discussion article for the meetup : Sydney Meetup - March
WHEN: 26 March 2014 06:30:00PM (+1100)
WHERE: Sydney City RSL, 565 George St, Sydney, Australia 2000
So far so good. Our last two meetups have been great - so lets do it one more time.
6:30 PM for early discussion 7PM general ... |
2f27cb37-f9d9-4a78-a488-e9af0ac68dad | StampyAI/alignment-research-dataset/arbital | Arbital | Countability
The [set](https://arbital.com/p/-3jz) of *counting numbers*, or of *positive integers*, is the set $\mathbb{Z}^+ = \{1, 2, 3, 4, \ldots\}$.
A set $S$ is called *countable* or *enumerable* if there exists a [surjection](https://arbital.com/p/4bg) from the counting numbers onto $S$.
### Example: The ratio... |
38b80918-ae38-4a34-a7e0-4744eba53062 | StampyAI/alignment-research-dataset/arxiv | Arxiv | The Conditional Entropy Bottleneck
1 Introduction
---------------
Despite excellent progress in classical generalization (e.g., accuracy on a held-out set), the field of Machine Learning continues to struggle with the following issues:
* [leftmargin=1em]
* Vulnerability to adversarial examples.
Most machine-learn... |
83e9070f-33bb-4ecc-ba6b-036ad9fa7252 | trentmkelly/LessWrong-43k | LessWrong | [Exercise] Four Examples of Noticing Confusion
Confusion is a felt sense; a bodily sensation you can pay attention to and notice! So here's my question for you:
What's this Song about?
Think About Things by Daði Frey
Youtube Link, Spotify Link
(recommended to listen to the song first, before reading the lyrics)
... |
a857a5e6-20e9-4d79-91a8-1cc83cfa9d14 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | The AI guide I'm sending my grandparents
*This is intended to be a comprehensive guide to getting up to speed in the AI landscape for somebody with little or no background. It walks through AI history, how it works, capabilities, potential impacts, and the general current landscape. I’ve been sending it to my friends ... |
c5176e07-7fb5-4ad8-ba73-5eaa28027b2f | trentmkelly/LessWrong-43k | LessWrong | Game-theoretic Alignment in terms of Attainable Utility
Acknowledgements:
This article is a writeup of research conducted through the SERI program under the mentorship of Alex Turner. It extends our research on game-theoretic POWER and Alex's research on POWER-seeking.
Thank you to Alex for being better at this than... |
bffbdc43-fef1-4f22-b08a-a6d1180c76c7 | trentmkelly/LessWrong-43k | LessWrong | Why does the mind wander?
By Joshua Shepherd, in Neuroscience of Consciousness (forthcoming)
I found this paper interesting. The paper is annoying trapped inside a Word document, which is about as bad as the standard PDF situation but bad in different ways, so I've included here the abstract, the conclusion, and a ch... |
49d23c77-c3e0-4d8e-ade3-76a3a41338f1 | StampyAI/alignment-research-dataset/arxiv | Arxiv | Optimal Cost Design for Model Predictive Control.
1 Introduction
---------------
Robotics planning algorithms often sacrifice optimality for tractability. One of the most well known and widely used planning paradigms in robotics is nonconvex model predictive control (MPC) (Lee, [2011](#bib.bib9); García et al., [19... |
049d6f26-6f56-45f6-a7b4-2699977ff2fa | trentmkelly/LessWrong-43k | LessWrong | Virtue signaling is sometimes the best or the only metric we have
Subtitle: Costly virtue signaling is an irreplaceable source of empirical information about character.
The following is cross-posted from my blog, which is written for a more general audience, but I think the topic is most important to discuss here on ... |
85e15b5d-81cf-4253-9a28-bb4425df114e | trentmkelly/LessWrong-43k | LessWrong | Corrigibility through stratified indifference
A putative new idea for AI control; index here.
Corrigibility through indifference has a few problems. One of them is that the AI is indifferent between the world in which humans change its utility to v, and world in which humans try to change its utility, but fail.
Now ... |
031a7e29-84e5-440c-a04f-3378b838563d | trentmkelly/LessWrong-43k | LessWrong | Four ways I've made bad decisions
I was in the process of making a decision the other day, and thought about the times I felt like I either made an incorrect decision or followed an incorrect decision making process. Turns out, there was actually many such cases.[1]
I categorized those into four different failure mod... |
15b9c42a-3a73-4927-b43e-5288a5d5b066 | trentmkelly/LessWrong-43k | LessWrong | Why "Changing the World" is a Horrible Phrase
Steve Jobs famously convinced John Scully from Pepsi to join Apple Computer with the line, “Do you want to sell sugared water for the rest of your life? Or do you want to come with me and change the world?”. This sounds convincing until one thinks closely about it.
Steve... |
402233a6-cec9-4be7-ab79-5dbb8ba741a8 | trentmkelly/LessWrong-43k | LessWrong | Shell, Shield, Staff
None |
ae44cb6c-d3b3-4967-8ecf-b180b363d4c1 | trentmkelly/LessWrong-43k | LessWrong | Terminal Preferences
I do most of my coding on the command line, so: some views on terminals.
I like having lots of terminals side-by-side, to see many different things simultaneously. Most places I've worked have kept code to 80 columns, with two-space indentation, which is a good fit for my approach. Here's how I u... |
68c72dae-ac38-4cbd-9272-6544773e6d6f | StampyAI/alignment-research-dataset/arxiv | Arxiv | A Closer Look at Deep Policy Gradients
1 Introduction
---------------
Deep reinforcement learning (RL) is at the core of some of the
most publicized achievements of modern machine
learning [[19](#bib.bib19), [9](#bib.bib9), [1](#bib.bib1), [10](#bib.bib10)]. To many, this
framework embodies the promise of the real-... |
08d1e3ac-7727-453a-a273-c245e778b5e3 | trentmkelly/LessWrong-43k | LessWrong | True Optimisation
Hello less wrong community! This is my first post here, so I know that my brain has not (obviously) been optimised to its fullest, but I've decided to give posting a try.
Recently, someone very close to me has unfortunately passed away, leading to the invitable inner dilemma about death. I don't k... |
772a5261-5841-42f0-a67c-3f77b3922846 | trentmkelly/LessWrong-43k | LessWrong | Embedding Ethical Priors into AI Systems: A Bayesian Approach
Abstract
Artificial Intelligence (AI) systems have significant potential to affect the lives of individuals and societies. As these systems are being increasingly used in decision-making processes, it has become crucial to ensure that they make ethically so... |
fd2ad5fd-5eab-4c8c-a30d-9377011500ab | trentmkelly/LessWrong-43k | LessWrong | What is space? What is time?
This is an attempt to describe space and time without much physics baggage. Then I can refer back to this post when I want to add more physics baggage.
It is surprisingly difficult to explain what I mean by “time” and “space” colloquially. The assumptions of time and space permeate langua... |
f8f32234-c9d7-4b70-a956-c3fe7cbf004c | trentmkelly/LessWrong-43k | LessWrong | The Milton Friedman Model of Policy Change
One-line summary: Most policy change outside a prior Overton Window comes about by policy advocates skillfully exploiting a crisis.
In the last year or so, I’ve had dozens of conversations about the DC policy community. People unfamiliar with this community often share a ... |
ebe012f9-2c8a-4850-b0c5-ca2121391310 | trentmkelly/LessWrong-43k | LessWrong | What makes a probability question "well-defined"? (Part I)
See also: Probability is in the Mind, You Are Never Entitled to Your Opinion, and “I don’t know.”
(Cross-posted from a new blog I'm trying to start. Sorry if some of the material is a bit basic for these parts, there's a larger point I'm trying to build up to... |
e70ec665-eebd-4a0e-b869-bb6cbc563c0b | trentmkelly/LessWrong-43k | LessWrong | Two methods on akrasia / unhealthy habits / addictions
1. Ignoring goals, tabooing / identifying with methods - the opposite of Beeminder, even the opposite of SMART
Suppose I want to lose weight. Have a general idea of how much and how fast, and decide on the method. Then I ignore the goal and focus on the method. I... |
d951a247-b315-40a2-9e9b-c66ec9b972e1 | StampyAI/alignment-research-dataset/blogs | Blogs | AGI outcomes and civilizational competence
> The [latest IPCC] report says, “If you put into place all these technologies and international agreements, we could still stop warming at [just] 2 degrees.” My own assessment is that the ... |
5c7e1098-d6f1-4b28-8914-f14b39b644e5 | trentmkelly/LessWrong-43k | LessWrong | Thin Alignment Can't Solve Thick Problems
> "Plurality is the law of the earth."
> —Hannah Arendt, The Human Condition (1958)
Abstract
Contemporary AI alignment practices overwhelmingly rely on thin ethical concepts — formal, universal notions such as "honesty," "harmlessness," and "helpfulness" — while neglecting t... |
693d698e-61ce-4af3-a63a-67e5d8bae16f | trentmkelly/LessWrong-43k | LessWrong | Second order logic, in first order set-theory: what gives?
With thanks to Paul Christiano
My previous post left one important issue unresolved. Second order logic needed to make use of set theory in order to work its magic, pin down a single copy of the reals and natural numbers, and so on. But set theory is a first ... |
450d6274-fcff-4e99-a53d-fd327e95f735 | trentmkelly/LessWrong-43k | LessWrong | Transparency and Accountability
[Added 02/24/14: After writing this post, I discovered that I had miscommunicated owing to not spelling out my thinking in sufficient detail, and also realized that it carried unnecessary negative connotations (despite conscious effort on my part to avoid them). See Reflections on a Per... |
b9fdea06-bb8c-46d1-aeea-9a7c91412495 | trentmkelly/LessWrong-43k | LessWrong | Give Neo a Chance
(To learn more about Place AI and other things mentioned here, refer to the first post in the series. Sorry for the rough edges—I’m a newcomer, non‑native speaker, and my ideas might sound strange, please steelman them and share your thoughts. The safe future is counterintuitive and takes a book to e... |
6d21b8a7-82b9-4901-987d-03ffccd34c46 | trentmkelly/LessWrong-43k | LessWrong | Statement on AI Extinction - Signed by AGI Labs, Top Academics, and Many Other Notable Figures
Today, the AI Extinction Statement was released by the Center for AI Safety, a one-sentence statement jointly signed by a historic coalition of AI experts, professors, and tech leaders.
Geoffrey Hinton and Yoshua Bengio hav... |
24bca632-fc65-4a93-b0ed-8b2ddfba7274 | trentmkelly/LessWrong-43k | LessWrong | Updates from Leverage Research: History and Recent Progress
As many LessWrong readers may have seen there was recently this post ostensibly aimed at creating common knowledge about Leverage Research.
While Leverage Research is no longer actively involved in the Rationality community, from the comments and upvotes it... |
9fd5954e-4df7-4898-9b19-f6dfd5b505b5 | trentmkelly/LessWrong-43k | LessWrong | [Fiction] Unspoken Stone
Author's note: I got the seed of an idea for a story. As an exercise, I wrote it down. I have no expectation of returning to this; I have a larger writing project I care about more. But, it was fun to write, so maybe you'll find it fun to read.
"Dr. Passo, thank you for coming."
The young ma... |
b053e527-86fa-4307-b879-b279e2185862 | trentmkelly/LessWrong-43k | LessWrong | Life, Death, and Finance in the Cosmic Multiverse
I published this report a while ago and would love to get some feedback from the experts here.
https://www.bcaresearch.com/public/content/GIS_SR_2021_12_21.pdf
Highlights
• The idea of parallel worlds seems like the stuff of science fiction. Yet, most leading cosmo... |
1f805356-06bf-4df9-a966-7307de580291 | trentmkelly/LessWrong-43k | LessWrong | Have frontier AI systems surpassed the self-replicating red line?
Introduction
A few weeks ago, researchers at Fudan University published a paper claiming frontier AI systems had surpassed the self-replicating red line. Notably, they were able to do this with Llama31-70B-Instruct and Alibaba’s Qwen25-72B-Instruct, two... |
f9a2c914-fadd-4fa3-8de2-90b3ed95537a | trentmkelly/LessWrong-43k | LessWrong | How I Think, Part Two: Distrusting Individuals
I've talked about movies before, so I guess I'll continue with that. I spoke with someone who seemed to distrust the people that decide which movies win the Academy Awards (at the time of writing, it's a group of over 7,000 people who have to have movie credits and be inv... |
46b9c333-b11d-443e-a981-f7d841ee6685 | trentmkelly/LessWrong-43k | LessWrong | 14+ AI Safety Advisors You Can Speak to – New AISafety.com Resource
Getting personalised advice from a real human can help newcomers to AI safety figure out how to contribute most effectively. For example, I (Bryce) ended up in my current role largely thanks to a call with 80,000 Hours.
There are a number of organisa... |
b4427aa1-c1f1-4c81-bd49-997c7e659d75 | trentmkelly/LessWrong-43k | LessWrong | 2015 New Years Resolution Thread
The new year is a popular Schelling point to make changes to your activities, habits, and/or thought processes. This is often done via the New Year's Resolution. One standard piece of advice for NYRs is to make them achievable, since they are often too ambitious and people end up givi... |
8e693b0e-0043-4001-b29f-f84fb4a0a2fa | trentmkelly/LessWrong-43k | LessWrong | Why 1-boxing doesn't imply backwards causation
I have written extensively on Newcomb's problem, so I assumed that I must have written up a clear explanation for this at some point. However, this doesn't seem to be the case, so I suppose I'm doing it now. The TLDR is that only one counterfactual is factual and the rest... |
94f20de8-5d57-45b2-8bd5-f386b5895f78 | trentmkelly/LessWrong-43k | LessWrong | Will bird flu be the next Covid? "Little chance" says my dashboard.
Rob and I built a bird flu risk dashboard.
You can check it out here:
birdflurisk.com
There is a box to be emailed if the index changes significantly. Hopefully this helps some people.
Happy to answer questions/make changes. |
ce84ecd9-1608-4528-9370-e61b782fa21f | trentmkelly/LessWrong-43k | LessWrong | Expanding HarmBench: Investigating Gaps & Extending Adversarial LLM Testing
Dear Alignment Forum Members,
We recently reached out to Oliver from Safe.ai regarding their work on HarmBench, an adversarial evaluation benchmark for LLMs. He confirmed that while they are not planning a follow-up, we have their blessing ... |
fb4041e5-4a47-4c89-a99f-7497cc447961 | trentmkelly/LessWrong-43k | LessWrong | Proof Explained for "Robust Agents Learn Causal World Model"
This post was written during Alex Altair's agent foundations fellowship program, funded by LTFF. Thank you Alex Altair, Alfred Harwood, Daniel C for feedback and comments.
This is a post explaining the proof of the paper Robust Agents Learn Causal World Mod... |
e2f58a02-843b-4063-bae2-04fc7826668c | trentmkelly/LessWrong-43k | LessWrong | Why "is evidence for" is a symmetric relation
None |
ac15f806-3f36-4dc9-a266-79e5240dd59b | trentmkelly/LessWrong-43k | LessWrong | Supplementary Alignment Insights Through a Highly Controlled Shutdown Incentive
My proposal entails constructing a tightly restricted AI subsystem with the sole capability of attempting to safely shut itself down in order to probe, in an isolated manner, potential vulnerabilities in alignment techniques and then impr... |
7a080c9c-d6ad-4095-8bd8-d64d16dba730 | trentmkelly/LessWrong-43k | LessWrong | My Research Process: Key Mindsets - Truth-Seeking, Prioritisation, Moving Fast
This is post 2 of a sequence on my framework for doing and thinking about research. Start here.
Before I get into what exactly to do at each stage of the research process, it’s worth reflecting on the key mindsets that are crucial througho... |
2d15c7b1-8012-4604-94fc-2db7be98cd69 | trentmkelly/LessWrong-43k | LessWrong | Don't die with dignity; instead play to your outs
In MIRI announces new "Death With Dignity" strategy, Eliezer makes a number of claims. As I understood them, they are:
1. Humanity is very likely doomed, because AI alignment is extremely hard, and no one has a strategy that could plausibly work (or is on track to fi... |
37276670-5c12-466f-a651-449b4a8d5e93 | StampyAI/alignment-research-dataset/blogs | Blogs | Luke at Quixey on Tuesday (Aug. 20th)
This coming Tuesday, MIRI’s Executive Director Luke Muehlhauser will give a talk at [Quixey](https://www.quixey.com/) titled **Effective Altruism and the End of the World**. If you’re in or near the Sout... |
b3f6a3ff-48e3-4a13-88f2-b07dbf83a09e | trentmkelly/LessWrong-43k | LessWrong | African Wild Dogs Vote By Sneezing - Can AI Help Us Do Better?
African wild dogs vote on whether or not to hunt by sneezing. As a social species, they institute a form of democracy. A tally is taken of those voting to hunt, with a vote being signaled by a sneeze. Have humans really transcended such one-dimensional vot... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.