id stringlengths 36 36 | source stringclasses 15
values | formatted_source stringclasses 13
values | text stringlengths 2 7.55M |
|---|---|---|---|
d3a7db27-2425-42bf-b488-f173f11a49c5 | trentmkelly/LessWrong-43k | LessWrong | Technical AI Alignment Study Group
Is anyone interested in joining a study group for studying technical AI alignment (specifically the MIRI course list)?
I've been lurking on this forum for quite some time, and I'm trying to get deep into technical Alignment by following the books on the MIRI course list. Following a... |
9bdbf826-d78f-4571-837c-227175324409 | StampyAI/alignment-research-dataset/blogs | Blogs | ML Safety Newsletter #7
Welcome to the 7th issue of the ML Safety Newsletter! In this edition, we cover:
* ‘Lie detection’ for language models
* A step towards objectives that incorporate wellbeing
* Evidence that in-context learning invokes behavior similar to gradient descent
* What’s going on with grokking?
* Troj... |
6e9b119b-3698-4559-86e7-72c54ca89f9c | trentmkelly/LessWrong-43k | LessWrong | Biological evolution as an analogy for the market economy
Both evolution and the free market are powerful forces in our universe. They can create both immense suffering and unimaginable success. We make moralistic judgments about the outcomes, but they are amoral.
By approaching economic policy in much the same way w... |
309fc325-9a36-472a-b76e-bc89ed915e3c | trentmkelly/LessWrong-43k | LessWrong | Additive and Multiplicative Subagents
This is the ninth post in the Cartesian frames sequence. Here, we refine our notion of subagent into additive and multiplicative subagents. As usual, we will give many equivalent definitions.
The additive subagent relation can be thought of as representing the relationship betwe... |
abe41734-acf5-44e7-be7c-31ce6772079b | StampyAI/alignment-research-dataset/arbital | Arbital | Bayes' Rule and its implications
Click below to start reading! |
65a91bca-93ee-46f2-bd0e-adec39fc6437 | trentmkelly/LessWrong-43k | LessWrong | How I buy things when Lightcone wants them fast
On occasion, for my work at Lightcone I have been able to buy things faster than their advertised lead times. For example, I once got…
* 10 sofas in 2 days when the website shipping speed said 5 days to 3 weeks
* 10 custom beds from Japan in 1 week when all suppliers... |
952c5091-1aa7-49cf-a9b9-39a30e149bd2 | trentmkelly/LessWrong-43k | LessWrong | Fixed point or oscillate or noise
Consider any property of any box of matter. Or consider any signal generated by a finite computer. Assume physics has true RNG. I claim that eventually either
* this signal will stop changing, or
* the system will reach a prior state and the signal will oscillate, or
* the system ... |
af5ad95b-92ce-4069-93b0-b24533ebb338 | StampyAI/alignment-research-dataset/arxiv | Arxiv | Parenting: Safe Reinforcement Learning from Human Input
1 Introduction
---------------
Within the next generation, autonomous learning agents could be regularly participating in our lives, for example in the form of assistive robots. Some variant of reinforcement learning (RL), in which an agent receives positive f... |
0bb1306e-3814-438d-b7e2-539dafebe00c | trentmkelly/LessWrong-43k | LessWrong | Working through D&D.Sci, problem 2 (solution)
None |
f876146f-e5b5-4042-abad-b6301cc5a9c7 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Uber Self-Driving Crash
*Content warning: discussion of death*
A year and a half ago an Uber self-driving car hit and killed Elaine
Herzberg. I [wrote](https://www.facebook.com/jefftk/posts/933323644812) at
the time:
>
> The dashcam video from the Uber crash has been released. It's really
> bad. The pede... |
052bcbb9-b367-440a-9243-b4855d7ec182 | trentmkelly/LessWrong-43k | LessWrong | Two Prosocial Rejection Norms
When I tell people that I'm shy, the universal response I get is "wow, you don't seem shy at all!"
Problem statement
Pop culture psychology has a single concept for "debilitatingly reluctant to approach people", and it's called "shyness" or "social anxiety" and is explained by "fear of ... |
be1a150c-1e96-4b7b-9a76-b42256303fed | trentmkelly/LessWrong-43k | LessWrong | Abstractions on Inconsistent Data
[I’m not sure this makes any sense – it is mostly babble, as an attempt to express something that doesn’t want to be expressed. The ideas here may themselves be an abstraction on inconsistent data. Posting anyway because maybe somebody else will prune it into something useful.]
I. AB... |
4dc2076d-f13e-4226-ad90-72bc687dc0b9 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Book review: Architects of Intelligence by Martin Ford (2018)
*Cross-posted from the [EA Forum](https://forum.effectivealtruism.org/posts/R5Drd54mRmSujdNyz/book-review-architects-of-intelligence-by-martin-ford-2018).*
The 2018 book *Architects of Intelligence* by Martin Ford is a collection of 23 interviews about pr... |
d58155e0-a526-4129-9933-71db4acc8db3 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Should we publish mechanistic interpretability research?
**TL;DR:**Multiple people have raised concerns about current mechanistic interpretability research having capabilities externalities. We discuss to which extent and which kind of mechanistic interpretability research we should publish. The core question we want ... |
3d2f94dc-c0f0-4cb2-9382-ea328ec9ace2 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | [video] Paul Christiano's impromptu tutorial on AIXI and TDT
[Paul Christiano](http://ordinaryideas.wordpress.com/) was about to give a tutorial on AIXI and TDT, so I whipped out my iPhone and recorded it. His tutorial wasn't carefully planned or executed, but it may still be useful to some. Note that when Paul writes... |
0a9d3d10-5be2-4c7e-8e7f-aac524983a35 | trentmkelly/LessWrong-43k | LessWrong | Arguing Orthogonality, published form
My paper "General purpose intelligence: arguing the Orthogonality thesis" has been accepted for publication in the December edition of Analysis and Metaphysics. Since that's some time away, I thought I'd put the final paper up here; the arguments are similar to those here, but thi... |
b7346606-a477-49d7-b766-62df21edb30f | trentmkelly/LessWrong-43k | LessWrong | New York OB Meetup (ad-hoc) on Monday, Mar 24, @6pm
> Correction: The giant Starbucks is at 13 Astor Place #25, not 51 Astor which is a smaller Starbucks. (Yes, there are two Starbucks a block apart, here.) The smaller Starbucks has metal steps and a ramp leading up; don't go here. The giant Starbucks, which is wh... |
893fe217-bfd6-49fa-af2e-e6f4c801fe55 | StampyAI/alignment-research-dataset/blogs | Blogs | August 2020 Newsletter
#### MIRI updates
* Three questions from MIRI's Abram Demski: [What does it mean to apply decision theory?](https://www.lesswrong.com/posts/wgdfBtLmByaKYovYe), [How “honest” is GPT-3?](https://www.lesswrong.com/posts/c3RsLTcxrvH4rXpBL), and [How should AI debate be judged?](https://www.lesswro... |
0a8c3575-11c3-487b-a858-f4eb49ca634e | trentmkelly/LessWrong-43k | LessWrong | Facing the Intelligence Explosion
FYI, the english site is working again: http://intelligenceexplosion.com/en/
I highly recommend reading this, or spreading it around if you're already fluent.
This is incredible work. Thanks Luke! |
516901c2-d8fa-4825-a8a5-5e8e7987bd95 | trentmkelly/LessWrong-43k | LessWrong | Comprehensive List of All Singularity Summit Talks and Video Links
Here is an index of all Singularity Summit speeches.
Summit 2006
* Todd Davies, Tyler Emerson, and Peter Thiel. Introduction to the Singularity Summit.
* Ray Kurzweil. The Singularity: a hard or soft takeoff?
* Douglas Hofstadter. Trying to muse... |
38568d04-2ac6-4f08-865c-7921433270cc | trentmkelly/LessWrong-43k | LessWrong | Halloween thread - rationalist's horrors.
This is a kind of "X files" thread.
Post experiences which spooked you, which made you doubt reality, mathematical or physical laws, your sanity, memory or perception. The more improbable the better, but no second-hand legends please, share only what you personally experience... |
c2f6013a-0c2e-4d06-8952-f294f25c488b | trentmkelly/LessWrong-43k | LessWrong | The Next ChatGPT Moment: AI Avatars
Epistemic Status: Speculative. Dependent on intuitions about near-term AI tech and human psychology.
Claim: Within the next 1-3 years, many people will have an interaction with an AI avatar that feels authentically human. This will significantly amplify the public perception of cur... |
6c92926f-50fa-4fec-b3ff-38f70a19c256 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Can AI solve climate change?
I've seen a lot on the EA and 80,000 Hours websites about the dangers of misaligned AI, which is fair enough. But I haven't seen much about how well-aligned AI can help us. In particular, climate change seems like an ideal problem to put AI to work on, because it's so complex and systemic.... |
72a62eb1-7748-43ca-930b-bf97d5c60b28 | StampyAI/alignment-research-dataset/youtube | Youtube Transcripts | Better together - DeepMind: The Podcast (S2, Ep3)
welcome back to deepmind the podcast
i'm hannah fry a mathematician who's
been following the remarkable progress
in artificial intelligence in recent
years
in this series i'm talking to scientists
researchers and engineers about how
their latest work is changing our wo... |
c4cfd593-9431-4052-b71f-c799fac7d3e2 | trentmkelly/LessWrong-43k | LessWrong | Can time preferences make AI safe?
This post exposes a concept I came across while brainstorming, while I am absolutely no expert in the field, the concept seemed interesting to me, a quick search revealed that it was never discuss in any previous research so I am sharing it here in case it might be of some relevance ... |
8aed1229-f77a-48f5-b3f1-2ced856579c2 | trentmkelly/LessWrong-43k | LessWrong | Career prospects for physics majors
Physics is attractive to many highly intellectually capable students, because
* Physical theories represent pinnacles of human achievement
* It's intellectually stimulating
* It has a reputation for being a subject that smart people do
See the comments on the post What attract... |
9422ac74-4fb5-402f-a5e5-9550736bee3f | trentmkelly/LessWrong-43k | LessWrong | Food Preparedness for Disasters
Author: Finan Adamson
Last Updated: 03//2022
Overview
This guide goes into detail on food storage, how much food people need, how to know if your food has gone bad, and how to keep your cold food cold. This is meant to give you a tool to reason about how to prepare for and act in e... |
905d38e3-d5db-4157-b827-880cd78406e8 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | The beautiful magical enchanted golden Dall-e Mini is underrated
[Showcasing some of the pictures that my daughter and I created with the open source generative model [Dalle-Mini](https://huggingface.co/spaces/dalle-mini/dalle-mini). No deeper insights, just an appreciation how well it works in some cases.]
I have be... |
4e39b774-606e-4d98-a5dd-8a231ad460e6 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Atlanta October Meetup (Second of Two)
Discussion article for the meetup : Atlanta October Meetup (Second of Two)
WHEN: 19 October 2013 06:00:00PM (-0400)
WHERE: 1314 Hosea L Williams Dr. NE, Atlanta
Members: Please note the change in location from our usual meetup location.
Come join us for the second me... |
4161aad0-921d-4bbe-8021-cb6f35b79a00 | trentmkelly/LessWrong-43k | LessWrong | [Link] Using Stories to Teach Human Values to Artificial Agents
Abstract:
> Value alignment is a property of an intelligent agent indicating that it can only pursue goals that are beneficial to humans. Successful value alignment should ensure that an artificial general intelligence cannot intentionally or unintention... |
fc26be92-6894-4282-9e89-89fc2c257c6d | trentmkelly/LessWrong-43k | LessWrong | Leading Discussions: Addressing Time Monopolizers
> "Having been that monologuist, I fully endorse this article." -David Gerard
>
> This is part of a series on leading discussions, aimed at teachers and event organizers.
Have you ever been in a discussion (in a class or a workshop) when another attendee delivered a ... |
02bb17c0-64fa-4769-9f88-f6c3481be5af | trentmkelly/LessWrong-43k | LessWrong | Is there a mu-like term that means something like "why is this any of my business?"
[Context: mu is a way to respond to a question phrased in a yes or no way but that can't be answered with a yes or a no.]
I'll quickly try to explain my predicament.
I'm two episodes in to the HBO series Allen v. Farrow, a document... |
d130e97a-bb99-47d3-9149-c8cfca1e2b3c | StampyAI/alignment-research-dataset/lesswrong | LessWrong | New report: Intelligence Explosion Microeconomics
**Summary**: [Intelligence Explosion Microeconomics](http://intelligence.org/files/IEM.pdf) (pdf) is 40,000 words taking some initial steps toward tackling the key quantitative issue in the intelligence explosion, "reinvestable returns on cognitive investments": what k... |
d36d55d9-fa9b-41f7-8ffd-37c175035023 | trentmkelly/LessWrong-43k | LessWrong | On Shifgrethor
A small number of terms are elevated from the pages of literature, up to the Mount Olympus of blog post vernacular. Moloch, as the dark god of failed coordination problems. The Dark Forest, as domains where there’s active incentive not to be noticed. And from Ursula K. Le Guin, the Omelas child, a perso... |
c97aa5f9-df05-4cb3-867c-2cce24543a24 | trentmkelly/LessWrong-43k | LessWrong | The real reason Futarchists are doomed
There are a lot of little ways political leadership is organized differently throughout the world. Looking just at outcomes, states themselves vary widely on a bunch of axes - you have command economies, like North Korea, and market economies, like Singapore and Switzerland. You ... |
4f676eb1-b50a-4900-9c6f-ae6232aad41e | trentmkelly/LessWrong-43k | LessWrong | in defense of Linus Pauling
Linus Pauling was a chemist who won a Nobel Prize in Chemistry in 1954. He later became well-known for advocating large doses of Vitamin C.
I've heard that advocacy referred to as a cautionary tale, but I've long had a bit of sympathy for Linus, and I'm writing this post to explain how and... |
57a7b425-a881-401f-85fc-a43b65812de4 | trentmkelly/LessWrong-43k | LessWrong | Immobile AI makes a move: anti-wireheading, ontology change, and model splintering
Research projects
I'm planning to start two research projects on model splintering/reward generalisation and learning the preferences of irrational agents.
Within those projects, I'm aiming to work on subprojects that are:
1. Posed i... |
93ac23ca-d5ae-4ef8-9697-123348e5d6a7 | trentmkelly/LessWrong-43k | LessWrong | When the Model Starts Talking Like Me: A User-Induced Structural Adaptation Case Study
Introduction: This is not a story — it's a behavioral structure that has yet to be formally modeled
Over the past month, I engaged in a series of deliberately structured, consistent interactions with a general-purpose large language... |
9a574d5a-d556-4ad2-becb-1aa2e976615c | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Three Biases That Made Me Believe in AI Risk
Below I describe three faulty thoughts that made me too confident that AI-related x-risk would be real. After I identified these thoughts, I grew a lot more sceptical of AI Safety as an effective cause area. This post consists of edited parts of [some](https://bethzero.com/... |
5ae809d3-73e4-4b65-89c4-390d07731341 | trentmkelly/LessWrong-43k | LessWrong | Identity in What You Are Not
We often talk about identity as if it is what we are. Yet, strangely, most of the trappings of identity are things that are not, strictly speaking, us.
Some examples of ways people identify themselves:
* I'm a rock/punk/folk/etc. music fan
* I'm an artist/writer/musician/intellectual/e... |
3f7b5935-2e7c-4fcf-b78f-0c5428db187c | trentmkelly/LessWrong-43k | LessWrong | COVID-19 transmission: Are we overemphasizing touching rather than breathing?
[SEE UPDATE AT BOTTOM]
Here's the CDC (yeah I know, not always a reliable source, but the Chinese government guide seems to back it up):
> The virus is thought to spread mainly from person-to-person.
>
> * Between people who are in close... |
7a2369d0-0678-4831-9331-a05059b61516 | StampyAI/alignment-research-dataset/blogs | Blogs | MIRI’s 2016 Fundraiser
**Update December 22**: Our donors came together during the fundraiser to get us most of the way to our $750,000 goal. In all, 251 donors contributed **$589,248**, making this our second-biggest fundraiser to date. Although we fell short of our target by $160,000, we have since made up this shor... |
3db42477-dde8-4ab8-acae-f020859ccf63 | trentmkelly/LessWrong-43k | LessWrong | AI companies are unlikely to make high-assurance safety cases if timelines are short
One hope for keeping existential risks low is to get AI companies to (successfully) make high-assurance safety cases: structured and auditable arguments that an AI system is very unlikely to result in existential risks given how it wi... |
56d18ce5-3456-4838-ba9b-7f39f1909f6d | trentmkelly/LessWrong-43k | LessWrong | A Philosophical Treatise of Universal Induction (Link)
> Abstract: Understanding inductive reasoning is a problem that has engaged mankind for thousands of years. This problem is relevant to a wide range of fields and is integral to the philosophy of science. It has been tackled by many great minds ranging from philos... |
9f5a5944-115d-4da2-8629-ff8019dce07b | trentmkelly/LessWrong-43k | LessWrong | Activation Engineering Theories of Impact
Below I summarize the thoughts of other people on what's the Theory of Impact for Activation Engineering. I mostly base it on the "discussion" parts of the papers and the answers under @Chris_Leong's post What's the theory of impact for activation vectors?
Alex Turner's posts... |
1296bfbf-2835-4968-8f5d-f93b58a12f16 | trentmkelly/LessWrong-43k | LessWrong | Calibration proverbs
Cross-post from Telescopic Turnip.
Every problem is a calibration problem. That’s why most advice is basically useless: it tells you what to do, but doesn’t tell you when to stop. Therefore, the best pieces of advice are calibration advice. What we need is a metric to know if we are doing well, o... |
9dc966cf-33a9-4d66-88e1-ffef0ee23768 | LDJnr/LessWrong-Amplify-Instruct | LessWrong | "Suppose we’re working on some delightfully Hard problem - genetically engineering a manticore, or terraforming Mars, or aligning random ML models. We need very top tier collaborators - people who are very good at a whole bunch of different things. The more they’re good at, and the better they are, the better the chanc... |
2c453bd3-51c5-48ab-a82d-9eb86a2a73df | LDJnr/LessWrong-Amplify-Instruct | LessWrong | "Open Philanthropy is planning a request for proposals (RFP) for AI alignment projects working with deep learning systems, and we’re looking for feedback on the RFP and on the research directions we’re looking for proposals within. We’d be really interested in feedback from people on the Alignment Forum on the current ... |
db6e1ee6-040c-454c-a3f7-106ddeb27677 | trentmkelly/LessWrong-43k | LessWrong | Notes on Mental Security
|
3582eadc-c064-4b24-a49f-43d162cb0b66 | trentmkelly/LessWrong-43k | LessWrong | 2020 predictions
EDIT: Ha, just noticed that Zvi has done something similar, I’ll be interested to check another source.
The need for comparison
A couple of posts (1, 2) recently have shown how difficult it is to judge predictions without a baseline to judge against - calibration testing is the only real option.
Ha... |
bc0fa093-2323-4bcf-9822-d212a1a6f453 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Urbana-Champaign: Decision Theory
Discussion article for the meetup : Urbana-Champaign: Decision Theory
WHEN: 20 October 2013 02:00:00PM (-0500)
WHERE: 40.111220,-88.227363
Meetup will be held here at 2pm Sunday October 20.
Topics we might discuss include (but are not limited to):
Pascal's mugging and pa... |
8784b520-62bb-47d2-8aa6-deb7be605ce4 | trentmkelly/LessWrong-43k | LessWrong | [EA relevant] Announcing "Everyday Heroes of Effective Giving" Series
We have so many great people involved in the EA movement, people who think hard and well about what cause to prioritize and who dedicate a significant portion of their money and time to advancing global flourishing in the most cost-effective manner... |
e804e4bc-d25e-4c4b-ba0e-0125d7e317e1 | trentmkelly/LessWrong-43k | LessWrong | Seeking advice on careers in AI Safety
Title sums it up. I am a mid-level software engineer (no CS degree*, but 4 years of full-stack dev work at mid-size local company).
I was recently laid off, and have been looking for new jobs. However, I am seriously considering taking this time to retool myself to work on AI sa... |
e1e335a1-6d69-4688-aa54-9d04e9ad1790 | trentmkelly/LessWrong-43k | LessWrong | Fiction advice
Hi all,
I want to try my hand at a story from the perspective of an unaligned AI (a ghost in the machine narrator kind of thing) for the intelligence in literature contest, which I think would be both cool and helpful to the uninitiated in explaining the concept.
I want a fairly simple and archetypa... |
260a6d11-177d-45a0-97d6-453949c98bd6 | trentmkelly/LessWrong-43k | LessWrong | Beyond the Board: Exploring AI Robustness Through Go
Last year, we showed that supposedly superhuman Go AIs can be beaten by human amateurs playing specific “cyclic” patterns on the board. Vulnerabilities have previously been observed in a wide variety of sub- or near-human AI systems, but this result demonstrates tha... |
29b476ab-d110-4844-b2dc-1b28bb642e5a | trentmkelly/LessWrong-43k | LessWrong | LW Frontpage Experiments! (aka "Take the wheel, Shoggoth!")
Update: June 20th
After a few rounds of adjustments and careful examination of the data, we've decided to make the Enriched tab be the default for all logged-in users[1]. Anyone is who has not already switched tabs ever will be set to Enriched tab. If you dis... |
900a3053-2765-4629-82bf-f24e02c8e1f5 | trentmkelly/LessWrong-43k | LessWrong | Alignment Newsletter #40
Find all Alignment Newsletter resources here. In particular, you can sign up, or look through this spreadsheet of all summaries that have ever been in the newsletter.
The Alignment Forum sequences have started again! As a reminder, treat them as though I had highlighted them.
Highlights
Ref... |
2d238d70-a998-4923-b8eb-3c32ad507dc1 | trentmkelly/LessWrong-43k | LessWrong | The Baby-Eating Aliens (1/8)
(Part 1 of 8 in "Three Worlds Collide")
This is a story of an impossible outcome, where AI never worked, molecular nanotechnology never worked, biotechnology only sort-of worked; and yet somehow humanity not only survived, but discovered a way to travel Faster-Than-Light: The past's Futu... |
3ab49753-9172-4c48-87b1-1565ce38f3cc | trentmkelly/LessWrong-43k | LessWrong | Curiosity: A Greedy Feeling
> But what does the phrase “scientifically explicable” mean? It means that someone else knows how the light bulb works. - Eliezar Yudkowsky
> In principle, I agree with the notion that it is unforgivable to not want to know, and not want to improve your map to match the territory. However... |
d31851c1-b698-422d-98be-2a6c89a759a6 | trentmkelly/LessWrong-43k | LessWrong | Un-optimised vs anti-optimised
A putative new idea for AI control; index here.
This post contains no new insights; it just puts together some old insights in a format I hope is clearer.
Most satisficers are unoptimised (above the satisficing level): they have a limited drive to optimise and transform the universe. T... |
62c5ad3b-e36a-4712-9c1f-8e7835d9390e | trentmkelly/LessWrong-43k | LessWrong | Mandatory Post About Monkeypox
It’s 2022, so of course enough people are worried about an epidemic of Monkeypox that it feels like a necessary public service to do a post about Monkeypox. I didn’t get to it especially fast exactly because I am not all that concerned, but I’m not entirely not concerned, so here we are.... |
5bb9c901-1669-4345-80ac-b92b7b877b41 | StampyAI/alignment-research-dataset/blogs | Blogs | Cost disease and civilizational decline
*Click lower right to download or find on Apple Podcasts, Spotify, Stitcher, etc. Note that this recording combines a [previous post](https://www.cold-takes.com/empowerment-and-stakeholder-management/) with this one.*
"Cost disease" is a term sometimes used to refer to the **... |
947263e8-f60b-446a-8395-588ba69460f3 | trentmkelly/LessWrong-43k | LessWrong | Bus-Only Bus Lane Enforcement
I think bus lanes are great: the faster you make buses the more people switch from cars, which means the more frequently you can run the buses, which means even more people switch from cars. But when there's traffic drivers are often tempted to sneak over. The chances of getting a ticket ... |
e9572506-0000-40ba-bace-460e802ed1bb | trentmkelly/LessWrong-43k | LessWrong | [LINK] AI risk summary published in "The Conversation"
A slightly edited version of "AI risk - executive summary" has been published in "The Conversation", titled "Your essential guide to the rise of the intelligent machines":
> The risks posed to human beings by artificial intelligence in no way resemble the popular... |
d1fff759-4321-46b0-bffc-b8ff4c51244d | trentmkelly/LessWrong-43k | LessWrong | Crypto-currency as pro-alignment mechanism
Abstract: can encourage pro-alignment actions, inclinations by crypto-currency linked to values of free-energy in the observed, so presumably “obtainable” universe.
Quite a nascent notion, but an interesting and perhaps useful one. To "peg" the value of a cryptocurrency, to ... |
b9115ea5-5727-4811-9d1b-8d452f19fa04 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Please provide feedback on AI-safety grant proposal, thanks!
Hello,
Disclaimer: This is my first post on the EA forum so please be gentle (but honest) with feedback :) In order to provide some accountability to myself that I actually post this, I am setting a 15 minute timer to write this post, and then a 5 minute ti... |
d5117f5e-be34-4d80-b701-1dd227360c2e | trentmkelly/LessWrong-43k | LessWrong | April Fools: Announcing LessWrong 3.0 – Now in VR!
On this April 1st, we at LessWrong face two problems.
1. According to Michael Nielsen, software is a way to develop transformative tools for thought. Things like Anki and collaborative editing allow us to think new thoughts. But at LessWrong, we argue that this does... |
ed149bde-8787-408e-8b62-29e863b154c7 | trentmkelly/LessWrong-43k | LessWrong | The end of Victorian culture, part II: people and history
The last post was about deep structural reasons for the end of the Victorian cultural system. Here, I’ll discuss the unpredictable role of people and events.
In every society, some people get listened to and copied. We listen to the rich and powerful, because ... |
c197f941-3428-4d6c-8f46-12bec9185389 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Research agenda: Formalizing abstractions of computations
*Big thanks to Leon Lang, Jérémy Scheurer, Adam Gleave, and Shoshannah Tekofsky for their feedback on a draft of this post, to Euan McLean (via FAR AI) for his feedback and a lot of help with editing, and to everyone else who discussed this agenda with me, in p... |
41d5a5e6-8ca3-47a2-8bf7-e83cbf8d6ecf | trentmkelly/LessWrong-43k | LessWrong | Open Thread - Jan 2022 [Vote Experiment!]
Experiment: Reaction-Ballot Voting
This open thread is using an new experimental voting system: reaction-ballot voting.
In addition to voting on a comment's overall quality, you can also vote separately on a number of axes, and apply a small set of emoji reactions. Try out a... |
e9438721-404c-48f8-a4d5-99cb766b62bb | trentmkelly/LessWrong-43k | LessWrong | Solstice 2014 - Kickstarter and Megameetup
Summary:
* We're running another Winter Solstice kickstarter - this is to fund the venue, musicians, food, drink and decorations for a big event in NYC on December 20th, as well as to record more music and print a larger run of the Solstice Book of Traditions.
* There... |
2f7f7162-2948-4cc0-8765-6db638137cf5 | trentmkelly/LessWrong-43k | LessWrong | Draft/wiki: Infinities and measuring infinite sets: A quick reference
EDIT: This is now on the Wiki as "Quick reference guide to the infinite". Do what you want with it.
It seems whenever anything involving infinities or measuring infinite sets comes up it generates a lot of confusion. So I thought I would write a ... |
5e8b1d37-2ad7-4c2f-a9e4-b2b125519d24 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | AI Safety Camp: Machine Learning for Scientific Discovery
### Applications for [AI Safety Camp](https://aisafety.camp/) 2023 have opened!
I'll be Research Lead for the project [Machine learning for scientific discovery: the present and future of science-producing AI models](https://docs.google.com/document/d/1J6aadV... |
0d55462e-7212-43c2-b5fe-7b5790a0581b | StampyAI/alignment-research-dataset/arxiv | Arxiv | Perceptual Adversarial Robustness: Defense Against Unseen Threat Models.
1 Introduction
---------------
Many modern machine learning algorithms are susceptible to adversarial examples: carefully crafted inputs designed to fool models into giving incorrect outputs (Biggio et al., [2013](#bib.bib1); Szegedy et al., [... |
fc74838a-b87c-43b2-9069-a106a9bfcdee | trentmkelly/LessWrong-43k | LessWrong | Applied Rationality podcast - feedback?
My goal is to help non-rationalists become rationalists, and provide exercises for existing rationalists to hone their skills, in the form of audio/video episodes.
'Rationally Speaking' and 'The Bayesian Conspiracy' are lovely, but they don't serve the function I have in mind.
... |
5a3eacf4-dd83-4160-bfda-0867dd72b544 | trentmkelly/LessWrong-43k | LessWrong | Bands And Low-stakes Dances
When I got started playing for contra dances in the Boston area there were several area dances you could consider "low stakes". Small dances, couldn't pay very much, less enthusiasm from established bands. And so they were relatively willing to book bands that were starting out and friendly... |
2b727785-7c9c-4c0a-b6c4-654bc0df6737 | trentmkelly/LessWrong-43k | LessWrong | Meetup : LessWrong Sao Paulo monthly meetup
Discussion article for the meetup : LessWrong Sao Paulo monthly meetup
WHEN: 13 February 2016 02:00:00PM (-0200)
WHERE: Rua Antonio Carlos, 452, São Paulo - SP, Brazil
Encontro mensal de LessWrongers em SP no café Vanilla. Como sempre, novatos são bem vindos!
Discussion... |
471f9eae-3809-4c75-a00d-5de522caade3 | trentmkelly/LessWrong-43k | LessWrong | Rational Romantic Relationships, Part 1: Relationship Styles and Attraction Basics
Part of the Sequence: The Science of Winning at Life. Co-authored with Minda Myers and Hugh Ristik. Also see: Polyhacking.
When things fell apart between me (Luke) and my first girlfriend, I decided that kind of relationship wasn't ide... |
f767b9b1-73e6-4a4a-b26e-4caca7714a98 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | An Oracle standard trick
An idea I thought I'd been mentioning to everyone, but a recent conversation reveals I haven't been assiduous about it.
It's quite simple: whenever designing an Oracle, you should, as a default, run it's output channel through a probabilistic process akin to the [false thermodynamic miracle]... |
1df83f86-b736-4fce-a2cd-ada3314d9d28 | trentmkelly/LessWrong-43k | LessWrong | Leaving LessWrong for a more rational life
You are unlikely to see me posting here again, after today. There is a saying here that politics is the mind-killer. My heretical realization lately is that philosophy, as generally practiced, can also be mind-killing.
As many of you know I am, or was running a twice-monthly... |
cddee56c-9927-4052-b9a4-99a8757d0c2a | trentmkelly/LessWrong-43k | LessWrong | Funding for work that builds capacity to address risks from transformative AI
None |
a39a560f-4c04-409c-af7b-1d17a8300adb | trentmkelly/LessWrong-43k | LessWrong | LW Update 2019-03-12 -- Bugfixes, small features
We've deployed an updated version of LessWrong, with a bunch of bug fixes and small features added since last time (March 1).
Features and Changes
* Auto-linkify URLs in posts and comments
* Coauthored posts appear on your user profile page
* Drafts that are shared... |
07ae0e60-96b8-46a4-bb98-6903f8aaf341 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | xAI announces Grok, beats GPT-3.5
Some highlights:
> *Grok is still a very early beta product – the best we could do with 2 months of training – so expect it to improve rapidly with each passing week with your help.*
>
>
I find it very interesting that they managed to beat GPT-3.5 with **only 2 months of training... |
d89fe284-5ce0-45c1-a90d-0531768758c0 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | AI Box Log
Here's the log of the AI Box experiment that just finished, with MileyCyrus as the AI and me as the Gatekeeper. The AI was not let out of the box.
(9:33:25 PM) Dorikka: I may need to get up for a minute while we're playing, but I'll keep it as short as possible. I'll just give you the time back on the end... |
a6c1d669-254a-40e6-acce-8d5cdd35a687 | trentmkelly/LessWrong-43k | LessWrong | Weekly LW Meetups
This summary was posted to LW main on September 20th. The following week's summary is here.
Irregularly scheduled Less Wrong meetups are taking place in:
* Atlanta Lesswrong September Meetup (2nd of 2): 28 September 2013 06:00PM
* Berlin: Fermi paradox discussion: 18 October 2013 07:00PM
* Brati... |
acf566e1-975c-4f4f-bc64-d3146fc75411 | trentmkelly/LessWrong-43k | LessWrong | West LA Biweekly Meetups
Next Meetup: Wednesday, February 29th.
----------------------------------------
When: 7pm - 9pm every other Wednesday.
Where: The Westside Tavern in the upstairs Wine Bar, located inside the Westside Pavillion on the second floor, right by the movie theaters.
Parking is free for 3 ... |
56c720db-112d-44a7-94d8-1b0e236a4fb4 | trentmkelly/LessWrong-43k | LessWrong | [Job ad] Lead an ambitious COVID-19 forecasting project [Deadline extended: June 10th]
In brief: I’m hiring for a project manager for Epidemic Forecasting, an independent project spun off from the Future of Humanity Institute that provides forecasting information to governments and other decision-makers during the cov... |
4d548d85-26db-46b9-91f9-3f3fe5ccff9f | StampyAI/alignment-research-dataset/arxiv | Arxiv | DropoutDAgger: A Bayesian Approach to Safe Imitation Learning
arXiv:1709.06166v1 [cs.AI] 18 Sep 2017DropoutDAgger: A Bayesian Approach to Safe Imitation Learn ing
Kunal Menda, Katherine Driggs-Campbell, and Mykel J. Koche nderfer
Abstract — While imitation learning is becoming com-
mon practice in robotics, this app... |
e88f931e-4ed0-4937-a81a-fe0b3c8f9ade | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Refine's Third Blog Post Day/Week
Yesterday was the third blog post day at Refine (technically it was announced on Monday for Friday, but people leveraged the week differently so it's in between a week and a day).
In the current phase of the incubator we're mostly pushing them to generate and iterate on ideas, with l... |
b9c9fad1-a21f-46e3-b28c-019eac167f4f | trentmkelly/LessWrong-43k | LessWrong | Hammertime Day 4: Design
This is part 4 of 30 in the Hammertime Sequence. Click here for the intro.
A central theme of Hammertime is that rationalists interact with reality first. We try for at least five minutes. We build habits to solve the bugs. We stick our necks out and ask reality for rapid feedback. Only after... |
eb80c90d-bab8-49f6-8436-0a8ea7948c86 | trentmkelly/LessWrong-43k | LessWrong | Cautions about LLMs in Human Cognitive Loops
soft prerequisite: skimming through How it feels to have your mind hacked by an AI until you get the general point. I'll try to make this post readable as a standalone, but you may get more value out of it if you read the linked post.
Thanks to Claude 3.7 Sonnet for giving... |
82e1f370-a055-4011-b788-bd8f16309938 | trentmkelly/LessWrong-43k | LessWrong | How to deal with non-schedulable one-off stimulus-response-pair-like situations when planning/organising projects?
When working in projects it often happens that one task (A) must be performed after something else (B) happens. If you know B is going to happen at a particular time in the future, this is straightforward... |
a986ad32-ec12-4339-868e-97e3090af4ba | trentmkelly/LessWrong-43k | LessWrong | Meetup : London Social - October 26th
Discussion article for the meetup : London Social - October 26th
WHEN: 26 October 2014 03:00:00PM (+0100)
WHERE: Holborn, London
The next London meetup will be Sunday October 26th from 2pm. No topic or special activity. Just turn up and talk about what interests you.
The venue... |
6d154138-6cd3-43c0-affc-8a9595336182 | StampyAI/alignment-research-dataset/special_docs | Other | Safely Interruptible Agents
Safely Interruptible Agents⇤
Laurent Orseau
Google DeepMind
5 New Street Square,
London EC4A 3TW, UK
lorseau@google.comStuart Armstrong
The Future of Humanity Institute
University of Oxford, UK
stuart.armstrong@philosophy.ox.ac.uk
Machine Intelligence Research Institute
Berkeley, CA 94704
A... |
a084903c-a180-485a-8938-cab684c85eae | StampyAI/alignment-research-dataset/lesswrong | LessWrong | How to Study Unsafe AGI's safely (and why we might have no choice)
**TL;DR**
A serious possibility is that the first AGI(s) will be developed in a Manhattan Project style setting before any sort of friendliness/safety constraints can be integrated reliably. They will also be substantially short of the intelligence r... |
3f3fa51f-c2b6-4e9a-8378-14ab2c5c48aa | StampyAI/alignment-research-dataset/youtube | Youtube Transcripts | 270. My Objections to Eliezer Yudkowsky 1
hello and welcome to session 270 in the
AI safety.com reading group tonight
we'll be discussing the first half of my
objections to we're all gonna die by
Elia saidkowski and this is written by
Quentin Pope
Quentin Pope
um is not publicly writing who he is
this is his image fro... |
9d7f1d7e-1cfc-48af-a0db-8908eff4f2c8 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Play Regrantor: Move up to $250,000 to Your Top High-Impact Projects!
**Summary.** We have collected expressions of interest for a total donation budget of about **$200,000** from donors who are interested in using our platform. How many of them will end up donating and how much they’ll donate depends on you and how m... |
d20505e1-986f-4f94-a96a-f5479aebd8cd | trentmkelly/LessWrong-43k | LessWrong | Meetup : San Francisco Meetup: Rationalist Taboo
Discussion article for the meetup : San Francisco Meetup: Rationalist Taboo
WHEN: 04 May 2015 06:15:00PM (-0700)
WHERE: 1061 Market St #4, San Francisco, CA 94103
We'll be meeting to practice rationalist taboo!
If you need help getting in, you can call (301)-458-076... |
3424f663-4fa9-4998-8141-e24ace0cdc26 | trentmkelly/LessWrong-43k | LessWrong | [SEQ RERUN] Against Modal Logics
Today's post, Against Modal Logics was originally published on 27 August 2008. A summary (taken from the LW wiki):
> Unfortunately, very little of philosophy is actually helpful in AI research, for a few reasons.
Discuss the post here (rather than in the comments to the original ... |
44821d04-4f56-4343-be34-f0a48233b00f | StampyAI/alignment-research-dataset/special_docs | Other | Test Time Robustification of Deep Models via Adaptation and Augmentation
Under review as a conference paper at ICLR 2022
TESTTIMEROBUSTIFICATION OF DEEPMODELS
VIAADAPTATION AND AUGMENTATION
Anonymous authors
Paper under double-blind review
ABSTRACT
While deep neural networks can attain good accuracy on in-distribution... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.