id stringlengths 36 36 | source stringclasses 15
values | formatted_source stringclasses 13
values | text stringlengths 2 7.55M |
|---|---|---|---|
b5bd2005-49f1-476e-9f25-3c3138604812 | trentmkelly/LessWrong-43k | LessWrong | Pragmatism and Completeness
Philosophers often get a bad wrap, not entirely without justification. The common complaint is something like "come on, do we really need to analyze this thing in so much detail? Can't we just get on with life since the world keeps working whether we understand it or not?"
Without digging ... |
1926b06f-0bdb-4c5f-bd08-465afe256f5b | trentmkelly/LessWrong-43k | LessWrong | Meetup : Meetup Announcement: IRAD II and Future Events
Discussion article for the meetup : Meetup Announcement: IRAD II and Future Events
WHEN: 18 March 2015 07:00:00PM (-0700)
WHERE: 11066 Santa Monica Boulevard, Los Angeles, CA 90025
How to Find Us: Go into this Del Taco. We will be in the back room if possible.... |
6cc787c6-afe8-4730-b811-bd1c76a4faeb | StampyAI/alignment-research-dataset/arxiv | Arxiv | Fractal AI: A fragile theory of intelligence
Fractal AI
A Fragile Theory of Intelligence
Sergio Hernández Cerezo
Guillem Duran Ballester
Fr
AGI
le
BOOK #1
“Forward Thinking”
Version: V4.1
Fractal AI Book #1: Forward Thinking
Main researchers
Sergio Hernández Cerezo (
@EntropyFarmer
)
Guillem Duran Ballester (
@Miau_D... |
fca29e45-97b5-4674-892d-44faaeed7cae | trentmkelly/LessWrong-43k | LessWrong | The funnel of human experience
[EDIT: Previous version of this post had a major error. Thanks for jeff8765 for pinpointing the error and esrogs in the Eukaryote Writes Blog comments for bringing it to my attention as well. This has been fixed. Also, I wrote FHI when I meant FLI.]
The graph of the human population ove... |
3f663257-7be5-4e72-a6ca-9a06575c5c71 | LDJnr/LessWrong-Amplify-Instruct | LessWrong | "When you look at a paper, what signs cause you to take it seriously? What signs cause you to discard the study as too poorly designed to be much evidence one way or the other?I'm hoping to compile a repository of heuristics on study evaluation, and would love to hear people's tips and tricks, or their full evaluation-... |
87f86f52-6398-4a4a-a95f-4b71d951e9c4 | trentmkelly/LessWrong-43k | LessWrong | Interpreting the Learning of Deceit
One of the primary concerns when attempting to control an AI of human-or-greater capabilities is that it might be deceitful. It is, after all, fairly difficult for an AI to succeed in a coup against humanity if the humans can simply regularly ask it "Are you plotting a coup? If so, ... |
bdb8f44a-e38c-4c13-bf98-8913f385b6df | trentmkelly/LessWrong-43k | LessWrong | Rationality Quotes: May 2011
It looks like, this month, I get to be the one to start the quotes thread.
* Please post all quotes separately, so that they can be voted up/down separately. (If they are strongly related, reply to your own comments. If strongly ordered, then go ahead and post them together.)
* Do ... |
e63cbd99-7f5d-4480-aa70-59ebe20d279d | trentmkelly/LessWrong-43k | LessWrong | How to decide to get a nosejob or not?
The affect heuristic suggests that being attractive has many subtle advantages, both in social and professional life. These advantages might be too subtle to notice at any one interaction, but they are important in aggregate. If getting rhinoplasty made me 5 percentile more attra... |
f4f421d1-e550-417d-b667-2ddc2ae9a0a2 | trentmkelly/LessWrong-43k | LessWrong | Human-level Diplomacy was my fire alarm
Epistemic status: AI is my hobby, not my job.
Meta just released their new Diplomacy AI, Cicero. Sure, it was several different modules kludged together in a somewhat clever way, with no fundamentally new ideas or massive scaling, but it was damn impressive.
What surprises me... |
12916fa6-9174-4b27-82ac-b9caecefd39f | trentmkelly/LessWrong-43k | LessWrong | Harry Potter and the Methods of Rationality discussion thread, March 2015, chapter 116
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 116.
There is a site dedicated to the story at hpmor.com, w... |
33c6d21c-ac9c-4f4c-9e44-d38235a517e4 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Superintelligence 20: The value-loading problem
*This is part of a weekly reading group on [Nick Bostrom](http://www.nickbostrom.com/)'s book, [Superintelligence](http://www.amazon.com/Superintelligence-Dangers-Strategies-Nick-Bostrom/dp/0199678111). For more information about the group, and an index of posts so far s... |
e5e1ba6a-de97-4475-8333-81ab021116a2 | trentmkelly/LessWrong-43k | LessWrong | Can you MRI a deep learning model?
In an MRI scan you see which parts of a brain light up in response to a stimulus. This has proven invaluable in understanding brains.
Is there an equivalent thing you can do with deep learning models, where you can see which parts light up in response to stimuli? And does there exis... |
331836db-9e83-4502-9b72-4e10849ad951 | trentmkelly/LessWrong-43k | LessWrong | Social ethics vs decision theory
It seems to me that usually, when someone says "ethics" on lesswrong, ey usually means something along the lines of decision theory. When an average person says "ethics", ey is usually referring to a system of intuitions and social pressures designed to influence the behavior of member... |
3d909cf6-b72c-44ff-a800-890593a747a0 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Jeffrey Ding: Re-deciphering China’s AI dream
---
*In the realm of AI governance, China is an indispensable nation. In this talk, Jeffrey Ding, a D.Phil researcher for the Governance of AI at the Future of Humanity Institute, distills his opinions on the landscape of AI development in China. He draws mainly from 70+ ... |
9843f94f-e225-4bdf-bcff-65ffb82e8341 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | What if AGI is near?
Consider the following observations:
* The [scaling hypothesis](https://www.gwern.net/newsletter/2020/05#scaling) is probably true: we will likely continue to see great improvements in AI capabilities as model sizes increase.
+ Sutskever [mentioned](https://www.reddit.com/r/ControlProblem/commen... |
de6744c3-9b41-4f9d-a941-9a011aedb9ea | trentmkelly/LessWrong-43k | LessWrong | A pragmatic metric for Artificial General Intelligence
Current intelligence definitions are either partial or do not provide an actionable evaluation of intelligence, something tangible which can be used to quantify the potential of a given AGI. In absence of this definition we commonly define intelligence as the resu... |
63038db4-28ca-49f7-baff-061c3e9317ec | trentmkelly/LessWrong-43k | LessWrong | What are some good pieces on civilizational decay / civilizational collapse / weakening of societal fabric?
I've often heard rationalist and EAs talk about civilizational collapse, or civilizational decay: the set of trends whereby society seems to be becoming increasingly less functional, and might at some point fai... |
bf231972-63ff-480f-b241-47474880c4a0 | trentmkelly/LessWrong-43k | LessWrong | English is a Terrible Programming Language—And other reasons AI won't displace programmers
TL;DR It doesn't matter that ChatGPT can generate boilerplate that almost works, and it doesn't matter that, feasibly, a hypothetical model in the (near) future could write code that reliably works and might be a little bit more... |
b07a2aee-c875-4cb7-8e78-fb7c9c1fae6c | trentmkelly/LessWrong-43k | LessWrong | "Who I am" is an axiom.
I guess most of us, in moments of existential crisis, have asked the question of "Why am I me?". It is not about grammar: "because 'I' and 'me' are both first-person singular pronouns." Nor is it a matter of tautology: "of course Bill Gates is Bill Gates. duh" It's the urge to question why amo... |
9e8b940a-f6d0-4a02-81ab-81359e72d0b3 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Gato as the Dawn of Early AGI
*Written in a hurry today at the EA UCLA AI Timelines Workshop. Long and stream-of-thought, and a deliberate intellectual overreach as an epistemic exercise. My first foray into developing my own AGI timelines model* [*without deferring!*](https://www.lesswrong.com/posts/KGLJeeF6psTzHiuiR... |
45ed07c6-452a-44f1-b3b3-ad2fd463d3d8 | trentmkelly/LessWrong-43k | LessWrong | AI Alignment [Incremental Progress Units] this Week (10/22/23)
Sorry for getting this one out a day late.
Before getting started, one minor update. I liked David Orr’s suggestion that stars ⭐are too generic. So to emphasize that I am rating the innovativeness of a topic (and not necessarily its quality), I am going t... |
2a190286-547a-41d9-b815-c1e1201b3e35 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Multiple AIs in boxes, evaluating each other's alignment
Summary
-------
Below I describe a variation of the classical [AI box experiment](https://www.lesswrong.com/tag/ai-boxing-containment), in which *two* AIs in boxes are created, and asked to determine whether the other is [aligned with the values of humanity](h... |
90d25871-4a35-48b6-b8ef-52b2ac342814 | trentmkelly/LessWrong-43k | LessWrong | Call for evaluators: Participate in the European AI Office workshop on general-purpose AI models and systemic risks
I am sharing this call from the EU AI Office for organizations involved in evaluation. Please take a close look: among the selection criteria, organizations must be based in Europe, or their leader must ... |
3edaf4eb-26e4-4c07-8a70-dd13d3cd2efe | StampyAI/alignment-research-dataset/youtube | Youtube Transcripts | The Model That Changes Everything: Alpaca Breakthrough (ft. Apple's LLM, BritGPT, Ernie and AlexaTM)
a little on the 72 hours ago a language
model was released that could end up
being as consequential as gpt4 now I
know you were thinking that's a bowl
claim but let's see if you agree with it
after watching what ha... |
1fe4f61c-8842-49d5-96c9-e5c00c31dcc4 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | The biological intelligence explosion
Summary: Human genetic engineering could lead to intelligence enhancement that leads to genetic engineers who are better at genetic engineering (and research on pathways to improving intelligence). Which leads to a continuing process of greater and greater intelligence. This ite... |
3cf29311-07fe-4316-b12f-d4994f6417e5 | trentmkelly/LessWrong-43k | LessWrong | My intellectual influences
Prompted by a friend's question about my reading history, I've been thinking about what shaped the worldview I have today. This has been a productive exercise, which I recommend to others. Although I worry that some of what's written below is post-hoc confabulation, at the very least it's fo... |
5a7cded5-5eca-40ca-84db-20d9b706a344 | trentmkelly/LessWrong-43k | LessWrong | Meiosis is all you need
Cross-posted from my Substack.
Note: this post is still a rough draft but I think this idea is important enough to publish it anyway. I intend to revise it in the future. Also if you are a biologist who wants to collaborate with me on this project, or a philanthropist who wants to provide fund... |
0328591d-7222-4417-a7be-b00c0027acd8 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Fort Collins, Colorado Meetup Thursday 7pm *New Place*
Discussion article for the meetup : Fort Collins, Colorado Meetup Thursday 7pm *New Place*
WHEN: 27 September 2012 07:00:00PM (-0600)
WHERE: Mugs, 306 West Laurel, Fort Collins CO
We're meeting at a coffee shop on the north side of the university. What... |
102a1fdf-82e6-47c0-a98c-270d74e62b13 | trentmkelly/LessWrong-43k | LessWrong | What's your viewpoint on the likelihood of GPT-5 being able to autonomously create, train, and implement an AI superior to GPT-5?
If this seems unlikely, could you elaborate on the reasons? What essential capabilities would it lack in order to accomplish self-improvement? For instance:
- Enhanced Short-Term Memory: A... |
c7b7cf10-d498-41c9-a4a5-aea646b90027 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | CNAS report: 'Artificial Intelligence and Arms Control'
By Paul Scharre & Megan Lamberth, Center for a New American Security.
From the introduction:
> Advances in artificial intelligence (AI) pose immense opportunity for militaries around the world. With this rising potential for AI-enabled military systems, some a... |
5c676945-c8f4-40b4-ba9d-cb569ef4738f | StampyAI/alignment-research-dataset/lesswrong | LessWrong | MIRI announces new "Death With Dignity" strategy
tl;dr: It's obvious at this point that humanity isn't going to solve the alignment problem, or even try very hard, or even go out with much of a fight. Since survival is unattainable, we should shift the focus of our efforts to helping humanity die with with slightly ... |
fcb89837-1572-4dce-b97b-8899c3b88367 | trentmkelly/LessWrong-43k | LessWrong | [Link] TED- Cynthia Kenyon: Experiments that Hint at Longer Lives
TEDGlobal 2011- Cynthia Kenyon: Experiments that Hint at Longer Lives
Video Link: http://www.ted.com/talks/cynthia_kenyon_experiments_that_hint_of_longer_lives.html
Speaker's Bio (from TED):
> Cynthia Kenyon is revolutionizing our understanding of ag... |
aed7322e-7cce-4304-bcb9-830e314b1354 | trentmkelly/LessWrong-43k | LessWrong | Emmett Shear to be interim CEO of OpenAI
Update 2: this is kind of stale given more recent developments and I'm not planning to update it further; probably any further discussion should happen in OpenAI: Facts from a Weekend or other newer posts.
Update: un-paywalled article from the Verge: https://www.theverge.com/... |
5cc7c5bf-5d78-4d8a-83bb-ead7d9f9875b | trentmkelly/LessWrong-43k | LessWrong | Solomonoff Induction explained via dialog.
|
fa073bb7-d77a-41ee-99db-e56b5dbcb8c2 | trentmkelly/LessWrong-43k | LessWrong | What information on (or relevant to) modal immortality do you recommend?
Which information should I consume if I want to better understand the implications (or whether there are any implications) of modal immortality?
Modal immortality combines two premises:
* we live in a Big World: your brain configuration is ins... |
70858581-b073-4e6e-8fe9-e87477576e96 | trentmkelly/LessWrong-43k | LessWrong | The weak-to-strong generalization (WTSG) paper in 60 seconds
There's probably nothing new here if you've read the paper.
I am not affiliated with the paper authors. I took some figures from the paper.
You should really go read the paper.
Thanks to everyone who helped me with this post! Thanks to Justis for providin... |
fe0ace4c-2b4b-41a7-9d24-c28a29faef25 | StampyAI/alignment-research-dataset/youtube | Youtube Transcripts | A systems neuroscience approach to building AGI - Demis Hassabis, Singularity Summit 2010
[Music]
our next speaker up on the floor was the
highest rated chess player in the world
at the age of 13 he wrote the
best-selling video game theme park at
the age of 17 and then he got a double
first in computer science from Ca... |
fe41ae0e-7699-4dae-9d4c-c9aeafea07ea | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | An exploration of GPT-2's embedding weights
*I wrote this doc in December 2021, while working at Redwood Research. It summarizes a handful of observations about GPT-2-small's weights -- mostly the embedding matrix, but also the LayerNorm gain parameters -- that I found while doing some open-ended investigation of the ... |
00d074d5-7662-4b46-ad39-1b81fd0e68bd | trentmkelly/LessWrong-43k | LessWrong | AI Alignment in The New Yorker
The article is centered around the Singularity, that we might be relatively near it, and whether or not it is inevitable. On the way to tackling these questions, the author introduces and describes the field of AI Alignment:
> A growing area of research called A.I. alignment seeks to le... |
727bca97-e523-4436-bbc9-b16d633d2105 | trentmkelly/LessWrong-43k | LessWrong | [Link] (EA Podcast) Social Status: The Key to the Matrix Part I
A new episode of Global Optimum has been released! Global Optimum is a podcast aimed at making altruists more effective. This episode is about the psychology of social status. The desire for status is a fundamental human motive. Understanding status can h... |
bd44e189-dd66-409b-b863-6fe4648bec33 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Algorithms vs Compute
Two scenarios:
* I take a vision or language model which was cutting edge in 2000, and run it with a similar amount of compute/data to what's typically used today.
* I take a modern vision or language model, calculate how much money it costs to train, estimate the amount of compute I could have ... |
4f8270fb-5a4c-4baa-ac39-5e472bdeeae9 | StampyAI/alignment-research-dataset/arxiv | Arxiv | Improving Deep Reinforcement Learning in Minecraft with Action Advice
Introduction
------------
Training deep reinforcement learning agents requires significant time and computational resources. This is especially true in rich, virtual, 3D dimensional environments with large state spaces such as the game, Minecraft... |
5aa15280-f556-4712-907e-2035c665f4d4 | trentmkelly/LessWrong-43k | LessWrong | What can we learn from insecure domains?
Cryptocurrency is terrible. With a single click of a button, it is possible to accidentally lose all of your funds. 99.9% of all cryptocurrency projects are complete scams (conservative estimate). Crypto is also tailor-made for ransomware attacks, since it makes it possible ... |
f7094511-d49a-4195-8c2a-635d00e43855 | awestover/filtering-for-misalignment | Redwood Research: Alek's Filtering Results | id: post3460
An apparently aligned AI system may nevertheless behave badly with small probability or on rare “bad” inputs. The reliability amplification problem is to reduce the failure probability of an aligned AI. The analogous security amplification problem is to reduce the prevalence of bad inputs on which the fai... |
054a6d7d-4197-4802-a42e-33715f8e2435 | StampyAI/alignment-research-dataset/arxiv | Arxiv | IRLAS: Inverse Reinforcement Learning for Architecture Search
1 Introduction
---------------
The past several years have witnessed the remarkable success of convolutional neural networks in computer vision applications. Thanks to the advances in network architectures, *e.g*. ResNet [[11](#bib.bib11)], Inception [[2... |
9c62a045-70bb-493d-bb20-15878cfee9a3 | trentmkelly/LessWrong-43k | LessWrong | Dating profiles from first principles: heterosexual male profile design
Update: I improved the profile of someone who reached out to me from this article. They went from zero matches in a year to ~2/week.
I think this is roughly the effect size one should expect from following this advice: it's not going to take you ... |
32622487-5988-4331-a132-7df97cf587c7 | trentmkelly/LessWrong-43k | LessWrong | Hiring decisions are not suitable for prediction markets
In a recent edition of The Diff ($), Byrne Hobart pointed out a talk given by Robin Hanson at Manifest. The talk is well worth listening to. For example, Robin says:
> I've always thought [Prediction Markets] was the best idea I've had in my life [...] My bigge... |
4e9fb553-b7b4-4b70-a8f1-f30712334317 | trentmkelly/LessWrong-43k | LessWrong | Hints about where values come from
TsviBT
Intro written a day later:
Spiracular and I discuss the nature and origin of values. The dialogue doesn't get into much of a single clear thread, but I had fun writing it and hope it has some interesting trailheads for others.
(You may wish to skip the methodological exchan... |
8b8b892c-1315-4f08-ac6e-2904aa7b632c | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Paper: Understanding and Controlling a Maze-Solving Policy Network
Mrinank, Austin, and Alex wrote a paper on the results from [Understanding and controlling a maze-solving policy network](https://www.lesswrong.com/posts/cAC4AXiNC5ig6jQnc/understanding-and-controlling-a-maze-solving-policy-network), [Maze-solving agen... |
f2331ee9-8754-41fe-983e-7da4d71f2379 | trentmkelly/LessWrong-43k | LessWrong | Consider having sparse insides
It's easier to seek true beliefs if you keep your (epistemic) identity small. (E.g., if you avoid beliefs like "I am a democrat", and say only "I am a seeker of accurate world-models, whatever those turn out to be".)
It seems analogously easier to seek effective internal architectures i... |
9c21f5c1-ab8c-45bf-903b-b32761fed932 | awestover/filtering-for-misalignment | Redwood Research: Alek's Filtering Results | id: post1445
This was originally posted on Aligned AI's blog ; it was ideated and designed by my cofounder and collaborator, Rebecca Gorman. EDIT : many of the suggestions below rely on SQL-injection style attacks, confusing ChatGPT as to what is user prompt and what is instructions about the user prompt. Those do wor... |
dd5e243c-5aad-473b-bda3-aeb787a54bb5 | trentmkelly/LessWrong-43k | LessWrong | Atlantis: Berkeley event venue available for rent
Many events in and around the rationality community are run in Berkeley and might want event space. This is an announcement that there’s a venue in Berkeley, called Atlantis, that’s very well-suited to these kinds of events. It’s a former sorority house, so it fits lot... |
b28a3533-027e-4068-b906-fcef289a63f5 | trentmkelly/LessWrong-43k | LessWrong | False thermodynamic miracles
A putative new idea for AI control; index here. See also Utility vs Probability: idea synthesis.
Ok, here is the problem:
* You have to create an AI that believes (or acts as if it believed) that event X is almost certain, while you believe that X is almost impossible. Furthermore, you ... |
06e8c56f-ecec-4148-93c4-cb5ce92ee9b1 | StampyAI/alignment-research-dataset/aisafety.info | AI Safety Info | What are the main sources of AI existential risk?
While we can't predict the full story of how AI will affect the future, there are [several broad dynamics](https://www.alignmentforum.org/posts/WXvt8bxYnwBYpy9oT/the-main-sources-of-ai-risk) that seem like plausible building blocks of an existential catastrophe.
For e... |
342b88b6-1636-4f9b-bb4d-c617ef8cc6ec | trentmkelly/LessWrong-43k | LessWrong | Training Regime Day 16: Hamming Questions
Note: Over the past 3 days, COVID-19 preparations have taken precedence over writing. There is still more work to do, but I will likely resume on March 7th.
Edit: March 7th is a lie. Probably will be another week or two until time permits.
Later Edit: Turned out to be closer... |
f91bcd2c-0063-401b-9bc3-efbe58992d5b | trentmkelly/LessWrong-43k | LessWrong | Thoughts on The Replacing Guilt Series — pt 1
I’m currently going through Nate’s series on caring. He starts with the issue of listless guilt. First post is about the stamp collector, the second one is about allowing yourself to fight for something.
I want to log my thoughts on this, because it seems that there’s ... |
a8857ee6-f3e2-4af0-b4ab-71cc0df8a928 | trentmkelly/LessWrong-43k | LessWrong | Meetup : San Jose Meetup: Park Day (IX), 3pm
Discussion article for the meetup : San Jose Meetup: Park Day (IX), 3pm
WHEN: 06 November 2016 04:00:00PM (-0700)
WHERE: 438 Coleman Ave, San Jose, CA 95110
[PLEASE NOTE: The meetup is at 3pm PST, even though for me it's being displayed as 4pm PDT. It is not at 4pm PST.]... |
b5223d30-a4a7-4e39-9eac-fc7e4bd0c445 | trentmkelly/LessWrong-43k | LessWrong | Hacking on LessWrong Just Got Easier
TrikeApps has done a great job running LessWrong and adding new features, but they could use a little help. Have you thought about improving the LessWrong website but haven't done it because you weren't sure how? Or had installation issues? Well, now is a great time to start, becau... |
393828df-c416-49b1-aa69-d89e44b42a11 | trentmkelly/LessWrong-43k | LessWrong | Meetup :St.Louis MO: Psycology of memorization, Thiel Fellow James Koppel, and games
Discussion article for the meetup : Psycology of memorization, Thiel Fellow James Koppel, and games
WHEN: 24 November 2012 02:30:00PM (-0600)
WHERE: St.Louis Mo, Kaldi's Coffee Clayton branch
I'm planing to do a presentation with... |
6988ded8-a1de-4bc3-8d4c-37bdf957672b | trentmkelly/LessWrong-43k | LessWrong | [Link] "First Is Best" - The serial position effect / primacy effect
"First is Best"
> Abstract
> We experience the world serially rather than simultaneously. A century of research on human and nonhuman animals has suggested that the first experience in a series of two or more is cognitively privileged. We report thr... |
6210c8e6-27ad-460b-88e4-3e6b08f72328 | trentmkelly/LessWrong-43k | LessWrong | Proposal: Community Curated Anki Decks For MIRI Recommended Courses
Spaced repetition is optimal for recalling factual information. It won't necessarily teach you anything that you haven't already learned. It helps you retain knowledge, and won't necessarily help you develop skills. But, within the domain of factual i... |
9b46c5de-f272-434c-884d-b733d73dcca2 | trentmkelly/LessWrong-43k | LessWrong | Bucket Brigade: Video
The biggest pain point with Bucket Brigade has been that you generally wanted to be running a video call in parallel. You kept needing to switch windows, and you needed to remember to mute and unmute in the call at the beginning and end of each song. Solstice used Twilio Video, which packages up ... |
7004a91e-bbed-48d8-ba80-874ddcd27c72 | trentmkelly/LessWrong-43k | LessWrong | EC2 Scripts
I do a lot of work on ec2 instances, and one thing that makes this easier is having a few scripts for manipulating them. These are very small, small enough to go in my .bashrc, but still make my life a lot easier.
The first one is start_ec2:
alias start_ec2='aws ec2 start-instances --instance-ids i-NNN... |
85c7a095-be39-4ebe-8902-0366af1ee79c | trentmkelly/LessWrong-43k | LessWrong | Experiential Pica
tl;dr version: Akrasia might be like an eating disorder!
When I was a teenager, I ate ice. Lots of ice. Cups and cups and cups of ice, constantly, all day long, when it was freely available. This went on for years, during which time I ignored the fact that others found it peculiar. ("Oh," I would... |
a05a08e1-f8bc-4814-97f0-a6980d971354 | trentmkelly/LessWrong-43k | LessWrong | Dennett's "Consciousness Explained": Prelude
I'm starting Dennett's "Consciousness Explained". Dennett says, in the introduction, that he believes he has solved the problem of consciousness. Since several people have referred to his work here with approval, I'm going to give it a go. I'm going to post chapter summa... |
9714983d-a1ff-43eb-9f58-07456a9cef9b | trentmkelly/LessWrong-43k | LessWrong | LessWrong Help Desk - free paper downloads and more (2014)
Over the last year, VincentYu, gwern and others have provided many papers for the LessWrong community (87% success rate in 2012) through previous help desk threads. We originally intended to provide editing, research and general troubleshooting help, but artic... |
869abe05-7992-4930-afe0-8df4cb1e2bc5 | trentmkelly/LessWrong-43k | LessWrong | Collecting and hoarding crap, useless information
I am realizing something that many, many intelligent people are guilty of - collecting and hoarding and accumulating crap, useless information. This is dangerous, because it feels like you're doing something useful, but you're not.
However, speaking personally - once ... |
00cea54c-55e4-45f8-8b21-3c1a80cdee12 | trentmkelly/LessWrong-43k | LessWrong | Stop buttons and causal graphs
Edit: This work is known to be obsolete. The design will not make correct side bets on whether the button will be pressed.
Suppose that all problems relating to epistemic rationality have been solved. We create an AI with an epistemic module and an instrumental module. The epistemic mod... |
280e148c-53b2-4636-a0bf-67f5760be660 | trentmkelly/LessWrong-43k | LessWrong | SIA doomsday: The filter is ahead
The great filter, as described by Robin Hanson:
> Humanity seems to have a bright future, i.e., a non-trivial chance of expanding to fill the universe with lasting life. But the fact that space near us seems dead now tells us that any given piece of dead matter faces an astronomicall... |
37124bae-7560-4e36-9e2d-097999c997e7 | trentmkelly/LessWrong-43k | LessWrong | (US only) Donate $2 to charity (bing rewards)
Probably not worth most of your time, but if you already have a Windows Live account, log in here and you earn enough Bing reward points (whatever those are) to donate $2 to charity - http://www.bing.com/rewards/signup/web
Unfortunately, the charity selection is limited t... |
9a6a5ae8-e1e8-4508-9d05-7cc1a1f57edc | StampyAI/alignment-research-dataset/distill | Distill Scientific Journal | Curve Detectors
### Contents
[A Simplified Story of Curve Neurons](#a-simplified-story-of-curve-neurons)[Feature Visualization](#feature-visualization)[Dataset Analysis](#dataset-analysis)[Visualizing Attribution](#visualizing-attribution)[Human Comparison](#human-comparison)[Joint Tuning Curves](#joint-tuning-curves... |
50dc58fc-d3b0-450b-a216-5608429d7993 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Immanuel Kant and the Decision Theory App Store
*[Epistemic status: About as silly as it sounds.]*
Prepare to be astounded by this rationalist reconstruction of Kant, drawn out of an unbelievably tiny parcel of Kant literature![[1]](#fn9lk7ht49za)
Kant argues that all rational agents will:
* “Act only according to ... |
7f3894a5-2c6a-4ede-818f-126bb804fc65 | trentmkelly/LessWrong-43k | LessWrong | It's OK to be biased towards humans
Let's talk about art.
In the wake of AI art generators being released, it's become pretty clear this will have a seismic effect on the art industry all across - from illustrators, to comic artists, to animators, many categories see their livelihood threatened, with no obvious "high... |
25556432-0b32-4636-bfc9-b53ce727993e | trentmkelly/LessWrong-43k | LessWrong | Meetup : Urbana-Champaign: The löbstacle for picking good actions.
Discussion article for the meetup : Urbana-Champaign: The löbstacle for picking good actions.
WHEN: 22 September 2013 02:00:00PM (-0500)
WHERE: 40.111240, -88.227370
Let's have a slightly more technical meetup! When: 2:00 PM this Sunday. Where: The ... |
7d600c7d-1dc2-40f2-b98e-95efdfa62df2 | trentmkelly/LessWrong-43k | LessWrong | Value extrapolation vs Wireheading
Talk given by Rebecca Gorman and Stuart Armstrong at the CHAI 2022 Asilomar Conference. We present an example of AI wireheading (an AI taking over its own reward channel), and show how value extrapolation can be used to combat it.
https://www.youtube.com/watch?v=REUanSy0SgU |
87bebcde-272d-4a19-85c9-6b0dbcafacfc | trentmkelly/LessWrong-43k | LessWrong | My experience getting funding for my biological research
Linked from my Substack
Research is expensive, especially these days. Even though grad students and postdocs are underpaid, it still costs a lot to employ them.[1]
And every year, lab suppliers increase their prices. Overall, it continually gets more expensive... |
09275e17-d02e-448e-b1db-c1d4638edf49 | StampyAI/alignment-research-dataset/arxiv | Arxiv | Delegative Reinforcement Learning: learning to avoid traps with a little help
1 Introduction
---------------
A *reinforcement learning agent* is a system that interacts with an unknown environment in a manner that is designed to maximize the expectation of a utility function that can be written as a sum of rewards ... |
b1e23177-6265-47fd-9e81-0ea6a6ce55cd | trentmkelly/LessWrong-43k | LessWrong | How does late-2022 COVID transmissibility drop over time?
This seems like the kind of thing LW folks might have good information on (better than my random googling).
I tested positive 7 days ago, and have been on Paxlovid for 5 days; symptoms are mostly gone, and we'll see if they bounce back when I finish the course... |
c1ab69db-4ac5-45be-88fc-5cbc91ff417e | StampyAI/alignment-research-dataset/arxiv | Arxiv | Improving Language Model Prompting in Support of Semi-autonomous Task Learning
1 Introduction
---------------
Large language models (LLMs; Brown et al., [2020](#bib.bib5); Bommasani et al., [2021](#bib.bib2)) are a potential knowledge source for autonomous agents, especially those that learn new tasks within a perf... |
6006f375-6b20-4f9d-82cd-c8815ee1747f | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | [AN #71]: Avoiding reward tampering through current-RF optimization
Find all Alignment Newsletter resources [here](http://rohinshah.com/alignment-newsletter/). In particular, you can [sign up](http://eepurl.com/dqMSZj), or look through this [spreadsheet](https://docs.google.com/spreadsheets/d/1PwWbWZ6FPqAgZWOoOcXM8N_t... |
cec78d36-2951-4c01-82b1-a4c867863079 | trentmkelly/LessWrong-43k | LessWrong | [Linkpost] Jan Leike on three kinds of alignment taxes
Jan Leike leads the safety team at OpenAI. This post is from a Substack series titled "Musings on the Alignment Problem". I've found his posts helpful for understanding how the OpenAI safety team is thinking about alignment & for developing my own models of alignm... |
abec8846-f862-4b5f-a72d-9c1b48405618 | trentmkelly/LessWrong-43k | LessWrong | Nonprofit Boards are Weird
> Note: anything in this post that you think is me subtweeting your organization is actually about, like, at least 3 organizations. (I'm currently on 4 boards in addition to Open Philanthropy's; I've served on a bunch of other boards in the past; and more than half of my takes on boards are ... |
0ecbf5af-6c30-4d88-99a6-c62a50c6666d | trentmkelly/LessWrong-43k | LessWrong | AIS Hungary Operations Officer role, Deadline: 2024 October 6th
Apply & read more here.
Please note that EA Hungary is also hiring for an Operations Officer role. If you are interested in both roles, please apply to that role as well here.
(If you are interested in both, we recommend making the two applications in p... |
71d132b9-8f1b-4c85-aef4-76a388152903 | trentmkelly/LessWrong-43k | LessWrong | Friendly to who?
At
http://lesswrong.com/lw/ru/the_bedrock_of_fairness/ldy
Eliezer mentions two challenges he often gets, "Friendly to who?" and "Oh, so you get to say what 'Friendly' means." At the moment I see only one true answer to these questions, which I give below. If you can propose alternatives in the co... |
2cff1d2d-1b41-4bce-9c76-0fc69b039737 | trentmkelly/LessWrong-43k | LessWrong | Binary Search UDT
|
656bfa3e-b9ff-4389-97eb-18ef82b4146a | trentmkelly/LessWrong-43k | LessWrong | Social incentives affecting beliefs
Having weird ideas relative to your friends and associates means paying social costs. If you share your weird ideas, you'll have more arguments, your associates will see you as weird and you'll experience some degree of rejection and decreased status. If you keep your weird ideas ... |
6bfc9188-ed68-4163-835e-8fa785ef42a2 | trentmkelly/LessWrong-43k | LessWrong | Weekly LW Meetups
This summary was posted to LW Main on August 8th. The following week's summary is here.
Irregularly scheduled Less Wrong meetups are taking place in:
* Atlanta - Filtering Ideas: 09 August 2014 07:00PM
* Bratislava: 18 August 2014 06:00PM
* Buenos Aires LessWrong meetup: 09 August 2014 04:00PM
... |
4381b614-e8e8-4a09-b513-d65a8983cee9 | StampyAI/alignment-research-dataset/arxiv | Arxiv | A Reflection on Learning from Data: Epistemology Issues and Limitations
A Reflection on Learning from Data: Epistemology Issues and Limitations
Ahmad Hammoudeh a1,SaraTedmori a,NadimObeid b
a Princess Sumaya University for Technology
b The University of Jordan
1 Corresponding author : E-mail: at.hammoud... |
13932205-be15-4932-9070-a12fed9a3ce5 | trentmkelly/LessWrong-43k | LessWrong | How to develop a photographic memory 2/3
As you may have guessed, these first two (2) posts give you the building blocks of mnemonics. The final post will polish some theories I've developed, refine the process in building personal mnemonic techniques, and finally: how to acquire a photographic memory.
As was the ca... |
ce516c8e-bf30-4122-a04d-30c90ef617ff | trentmkelly/LessWrong-43k | LessWrong | Some thoughts on automating alignment research
As AI systems get more capable, they may at some point be able to help us with alignment research. This increases the chance that things turn out ok.[1] Right now, we don’t have any particularly scalable or competitive alignment solutions. But the methods we do have might... |
c9512fad-6b16-4a25-a9b7-4a01279800cb | trentmkelly/LessWrong-43k | LessWrong | Computer-mediated communication and the sense of social connectedness
IMHO there is little chance that an online-only community could replicate the successes (many friendships among the members, very high levels of enjoyment, motivation and engagement) of LW NYC.
Why not? Well, at the risk of putting off those reader... |
8c529889-39aa-4444-9f00-8a47db36cd1c | trentmkelly/LessWrong-43k | LessWrong | Why no Roman Industrial Revolution?
Why didn’t the Roman Empire have an industrial revolution?
Bret Devereaux has an essay addressing that question, which multiple people have pointed me to at various times. In brief, Devereaux says that Britain industrialized through a very specific path, involving coal mines, steam... |
010bd6a6-2103-4562-88a0-95a983ca7fae | trentmkelly/LessWrong-43k | LessWrong | How to trade money and time
Time has a monetary value to you. That is, money and time can be traded for one another in lots of circumstances, and there are prices that you are willing to take and prices you are not. Hopefully, other things equal, the prices you are willing to take are higher than the ones you aren’t.
... |
1cfa3ddc-e5d1-43ca-aa34-27bf2396c5e6 | StampyAI/alignment-research-dataset/blogs | Blogs | Michael Fisher on verifying autonomous systems
 Michael Fisher is a professor of Computer Science, specialising in logical methods and automated formal verification, and Director of the multi-disciplinary [Centre for Autonomo... |
032c2e8a-46ee-417e-ad87-b53ead1dd9f6 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Vancouver CFAR Materials: Bayes practice!
Discussion article for the meetup : Vancouver CFAR Materials: Bayes practice!
WHEN: 20 April 2013 03:00:00PM (-0700)
WHERE: 2505 W broadway, vancouver
CFAR's core materials have become organized to the point of being print-out-and-teachable, so I'm going to print o... |
47b630ec-6230-40c7-87e4-295da91528c0 | trentmkelly/LessWrong-43k | LessWrong | Link: "Health Care Myth Busters: Is There a High Degree of Scientific Certainty in Modern Medicine?"
A feature in Scientific American magazine casts some light on the troubled state of modern medicine.
Health Care Myth Busters: Is There a High Degree of Scientific Certainty in Modern Medicine?
Short excerpt:
> We c... |
fc673b2b-6392-4952-a191-1683b3d5c5b7 | trentmkelly/LessWrong-43k | LessWrong | Cryonics: peace of mind vs. immortality
I wrote a blog post arguing that people sign up for cryo more for peace of mind than for immortality. This suggests that cryo organizations should market towards the former desire than the latter (you can think of it as marketing to near mode rather than far mode, in Hansonian t... |
d23995af-a01d-4874-a654-a6f7c18d8183 | trentmkelly/LessWrong-43k | LessWrong | Meta: Frontier AI Framework
Meta just published its version of an RSP (blogpost, framework).
(Expect several more of these in the next week as AI companies try to look like they're meeting the Seoul summit commitments before the Paris summit deadline in a week.)
----------------------------------------
No details o... |
2ee29dea-4983-4259-8a7b-a238695ffd63 | trentmkelly/LessWrong-43k | LessWrong | Feedback for learning
I wish there were better (or more known to me) quantitative tests of skills that are good for me to have. I often find learning things when there is tight feedback pretty fun. For instance, I play geography and history quizzes on Sporcle with an addictive vigor, and enjoy learning various languag... |
66fadd44-b2ae-421d-b283-dc1f904b19b1 | trentmkelly/LessWrong-43k | LessWrong | Zeno's Paradoxes
Dichotomy Paradox
Imagine the following situation:
> A runner is trying to run 1km. Before they can run 1km, then must run 0.5km. Before they can run 0.5km, they must run 0.25km. Before they can run 0.25km, they must run 0.125km and so on. This would require completing an infinite number of tasks, wh... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.