id stringlengths 36 36 | source stringclasses 15
values | formatted_source stringclasses 13
values | text stringlengths 2 7.55M |
|---|---|---|---|
a2617cac-4884-4c80-9773-2f112d0355b0 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Robustness to Scale
I want to quickly draw attention to a concept in AI alignment: Robustness to Scale. Briefly, you want your proposal for an AI to be robust (or at least fail gracefully) to changes in its level of capabilities. I discuss three different types of robustness to scale: robustness to scaling up, robustn... |
8994878b-6f1f-47da-a5f2-8b4d8c5ea9ff | trentmkelly/LessWrong-43k | LessWrong | When Having Friends is More Alluring than Being Right (by Ferrett Steinmetz)
None |
b0d58d78-7268-476b-8258-e6f405a29b90 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Without a trajectory change, the development of AGI is likely to go badly
*This is a draft of my entry to the Open Philanthropy* [*AI worldviews contest*](https://forum.effectivealtruism.org/posts/NZz3Das7jFdCBN9zH/announcing-the-open-philanthropy-ai-worldviews-contest)*. It's also a good summary of my own current wor... |
6e35500e-2a79-4621-b4b0-30f7d417d380 | StampyAI/alignment-research-dataset/blogs | Blogs | September 2020 Newsletter
Abram Demski and Scott Garrabrant have made [a major update](https://www.lesswrong.com/posts/9vYg8MyLL4cMMaPQJ/updates-and-additions-to-embedded-agency) to "[Embedded Agency](https://www.lesswrong.com/s/Rm6oQRJJmhGCcLvxh/p/i3BTagvt3HbPMx6PN)", with new discussions of ε-exploration, Newcomblik... |
324da70d-05dc-4777-b027-5ebdcf9b7131 | trentmkelly/LessWrong-43k | LessWrong | How to Teach Students to Not Guess the Teacher’s Password?
As a teacher, I wonder if it is possible to instill this skill into students the skills of rationality and critical thinking. I teach the third grade, and it is not immediately apparent how to apply this with my own class.
The problems I foresee are as follow... |
f70eb5a9-e3eb-4da0-9713-8c0bffb94fb8 | trentmkelly/LessWrong-43k | LessWrong | On insecurity as a friend
There’s a common narrative about confidence that says that confidence is good, insecurity is bad. It’s better to develop your confidence than to be insecure. There’s an obvious truth to this.
But what that narrative does not acknowledge, and what both a person struggling with insecurity and ... |
a4ba7886-c1e6-4641-a12e-3e1bb51f68b1 | trentmkelly/LessWrong-43k | LessWrong | The Experimental Apparatus
Previous: The problem of other minds (Greaterwrong)
Greaterwrong // bearlamp (my blog)
Science really likes to isolate variables. From first grade, dropping a ball to measure gravity experiment. We talk about the wind resistance on the ball, and how it might affect a measurement. So we rep... |
d35afc81-e0f4-4943-9e83-f80b5946f083 | trentmkelly/LessWrong-43k | LessWrong | "Rationalist Discourse" Is Like "Physicist Motors"
Imagine being a student of physics, and coming across a blog post proposing a list of guidelines for "physicist motors"—motor designs informed by the knowledge of physicists, unlike ordinary motors.
Even if most of the things on the list seemed like sensible advice t... |
fe8a6fb1-7e5c-4439-bdab-e5780c80b0fc | trentmkelly/LessWrong-43k | LessWrong | Covid 9/22/22: The Joe Biden Sings
Joe Biden firmly declared on 60 Minutes that the Covid pandemic is over.
As you would expect, those in Public Health did not all quietly agree with this.
When Bob Wachter says it is a judgment call whether the pandemic is over, you know the pandemic is over.
When Caitlin Rivers sa... |
21fe5780-a131-4aa2-818b-0f5e402d4a79 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Washington, D.C.: Typical Mind Fallacy
Discussion article for the meetup : Washington, D.C.: Typical Mind Fallacy
WHEN: 05 February 2017 03:30:00PM (-0500)
WHERE: Donald W. Reynolds Center for American Art and Portraiture
We will be meeting in the courtyard to talk about the typical mind fallacy: the assum... |
7e1792a5-2346-4b5b-9c6f-99a65017ddb2 | StampyAI/alignment-research-dataset/special_docs | Other | Collaborative game specification: arriving at common models in bargaining
Collaborative game specification: arriving at common models in bargaining
=========================================================================
6 March 2021
by [Jesse Clifton](https://longtermrisk.org/author/jesse-clifton/ "Posts by Jesse... |
6cd6acbf-6f55-436d-a530-dc9d67f28a81 | trentmkelly/LessWrong-43k | LessWrong | The Cabinet of Wikipedian Curiosities
For some time, I’ve been collecting my favourite Wikipedia articles – presented here for your reading pleasure. I found some of these through @depthsofwikipedia, Aella, and various links aggregators, but I don’t remember which articles came from where.
---------------------------... |
614c684b-0c19-4cef-98da-b9150ec5c82e | trentmkelly/LessWrong-43k | LessWrong | The Schelling Choice is "Rabbit", not "Stag"
Followup/distillation/alternate-take on Duncan Sabien's Dragon Army Retrospective and Open Problems in Group Rationality.
----------------------------------------
There's a particular failure mode I've witnessed, and fallen into myself:
I see a problem. I see, what seems... |
3b6ca767-605b-4b52-8661-a7feee31ff47 | trentmkelly/LessWrong-43k | LessWrong | Bear Surprise Freedom Network
Here's some crazy nascent ideas I have for getting information in and out of Russia when other methods fail. From what I'm seeing, current Internet censorship efforts focus on... the Internet.
It's worth remembering that there are plenty of other ways to use technology to communicate th... |
8f7ba8fb-a260-4337-b3e2-2ab8e7e17364 | trentmkelly/LessWrong-43k | LessWrong | Fallacies of reification - the placebo effect
TL;DR: I align with the minority position that "there is a lot less to the so-called placebo effect than people tend to think there is (and the name is horribly misleading)", a strong opinion weakly held.
The following post is an off-the cuff reply to a G+ post of gwern's... |
4fff01ec-34db-49e0-97c9-268c1d33e7e0 | trentmkelly/LessWrong-43k | LessWrong | Open Thread – Winter 2023/2024
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. Th... |
a16b2352-3fe2-40f8-8113-4025d0de1daa | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Yudkowsky and Christiano discuss "Takeoff Speeds"
This is a transcription of Eliezer Yudkowsky responding to Paul Christiano's [Takeoff Speeds](https://sideways-view.com/2018/02/24/takeoff-speeds/) live on Sep. 14, followed by a conversation between Eliezer and Paul. This discussion took place after Eliezer's [convers... |
4dc5e6e6-3079-4ecf-b489-c2a040463139 | trentmkelly/LessWrong-43k | LessWrong | Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
Browse these SAE Features on Neuronpedia!
Update 1: Since we posted this last night, someone pointed out that our implementation of ghost grads has a non-trivial error (which makes the results a-priori quite surprising). We computed the gho... |
a09841be-3bd6-4ce1-8f4d-2e0e4e89f756 | trentmkelly/LessWrong-43k | LessWrong | Weekly LW Meetups
This summary was posted to LW main on May 9th. The following week's summary is here.
New meetups (or meetups with a hiatus of more than a year) are happening in:
* Houston, TX: 17 May 2014 11:31AM
Irregularly scheduled Less Wrong meetups are taking place in:
* Chicago Calibration Game: 10 May 2... |
a33c30be-a0ec-4a23-829e-d1af8e5e85f7 | trentmkelly/LessWrong-43k | LessWrong | What if AGI had its own universe to maybe wreck?
Is there any good thought to read about what if we used all this fancy compute to first build a hugely detailed, one-way-glass model of our universe to contain AGI?
From a naive standpoint, it seems like maybe we could sort of "incept" the intellectual work we want aut... |
b964c137-bb22-45b7-8f1a-df3e4ab5c99c | trentmkelly/LessWrong-43k | LessWrong | Limiting an AGI's Context Temporally
Okay, so I have a proposal for how to advance AI safety efforts significantly.
Humans experience time as exponential decay of utility. One dollar now is worth two dollars some time in the future, which is worth eight dollars even further in the future, and so forth. This is the pr... |
f00d3d22-c902-44fb-b186-de0a7c24f69a | trentmkelly/LessWrong-43k | LessWrong | Improving Our Safety Cases Using Upper and Lower Bounds
(Does anyone have the original meme? I can’t find it)
One key challenge in discussing safety cases is separating two distinct questions:
1. Would a particular safety measure be sufficient if we had it?
2. Is it technically possible to implement that measur... |
4044c372-cc59-42b7-aef7-fd8edbe5aec3 | trentmkelly/LessWrong-43k | LessWrong | Risks from Learned Optimization: Introduction
This is the first of five posts in the Risks from Learned Optimization Sequence based on the paper “Risks from Learned Optimization in Advanced Machine Learning Systems” by Evan Hubinger, Chris van Merwijk, Vladimir Mikulik, Joar Skalse, and Scott Garrabrant. Each post in ... |
6590d016-56a1-4818-874b-898a313f1ecf | trentmkelly/LessWrong-43k | LessWrong | Minimax forecasting
%&latex
% operators that are separated from the operand by a space
% operators that require brackets
% operators that require parentheses
% Paper specific
This post continues the research programme of attacking the grain of truth problem by departure from the Bayesian par... |
dfdcaf96-bb8f-48f0-968c-8351d2d05f5c | trentmkelly/LessWrong-43k | LessWrong | A bit of word-dissolving in political discussion
> I found Scott Alexander's steelmanning of the NRx critique to be an interesting, even persuassive critique of modern progressivism, having not been exposed to this movement prior to today. However I am also equally confused at the jump from "modern liberal democracies... |
f7cd1c7e-fca0-4789-a783-e2150f404ee2 | trentmkelly/LessWrong-43k | LessWrong | FAI PR tracking well [link]
This time, it's by "The Editors" of Bloomberg view (which is very significant in News world). Content is very reasonable explanation of AI concerns, though not novel to this audience.
http://www.bloombergview.com/articles/2014-08-10/intelligent-machines-scare-smart-people
Directionally th... |
83774fb2-68e3-4547-9e70-06e81afe6158 | trentmkelly/LessWrong-43k | LessWrong | LINK: Anders Sandberg, "An Overview of Models of Technological Singularity"
I just noticed that this useful paper has not been linked yet from Less Wrong:
Sandberg, "An Overview of Models of Technological Singularity" (2010) |
5ca8fe4b-c460-4e74-b5a6-2bb7f93641a1 | StampyAI/alignment-research-dataset/blogs | Blogs | New new blog location
I’ve switched the host of my blog to Ghost, so my blog is now located at <https://bounded-regret.ghost.io/>. This is different from the previous re-hosting on my personal website, which is no longer maintained.
Feedly users can subscribe via RSS [here](https://feedly.com/i/subscription/feed/htt... |
109a842f-c207-42f5-87c0-ac1de1a4d922 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Washington DC: Robin Hanson visits to talk about giving
Discussion article for the meetup : Washington DC: Robin Hanson visits to talk about giving
WHEN: 30 March 2014 03:00:00PM (-0400)
WHERE: National Portrait Gallery
Robin Hanson will be visiting to discuss the issues involved with giving now vs. giving... |
f2fd2897-e50a-48bc-9c14-82ba44afec86 | StampyAI/alignment-research-dataset/arxiv | Arxiv | NAS-Bench-Suite-Zero: Accelerating Research on Zero Cost Proxies
1 Introduction
---------------
Algorithms for neural architecture search (NAS) seek to automate the design
of high-performing neural architectures for a given dataset.
NAS has successfully been used to discover architectures with better
accuracy/laten... |
8ab2651a-8b82-4eb7-9fe4-d55833f99a4a | trentmkelly/LessWrong-43k | LessWrong | Basics of Animal Reinforcement
Behaviorism historically began with Pavlov's studies into classical conditioning. When dogs see food they naturally salivate. When Pavlov rang a bell before giving the dogs food, the dogs learned to associate the bell with the food and salivate even after they merely heard the bell . Whe... |
796d5453-4778-4b86-b91f-326394abbb66 | trentmkelly/LessWrong-43k | LessWrong | Schelling Day: A Rationalist Holiday
The Big Idea
—Holidays are awesome.
—Getting to know people is awesome.
—Schelling Day is a holiday about getting to know people.
—The Boston group will celebrate Schelling day.
—Hopefully other cities will too.
Why Are We Doing This?
Getting to know people—really, truly gettin... |
cce8b642-212d-4e9b-9fe7-7356557b11f8 | trentmkelly/LessWrong-43k | LessWrong | Change the labels, undo infinitely good improvements
Infinity is big. You just won't believe how vastly, hugely, mindbogglingly big it is. I mean, you may think it's a long way down the road to the chemist's, but that's just peanuts to infinity.
And there are a lot of paradoxes connected with infinity. Here we'll be ... |
288d5934-be5b-4072-889b-046411bb7c52 | trentmkelly/LessWrong-43k | LessWrong | Dissent Collusion
Summary: Take turns being confronted with a group that’s presenting a united front, and grow comfortable agreeing or disagreeing with them.
Tags: Medium, Repeatable
Purpose: Working with groups involves balancing between agreeing with the group consensus when the group is wrong and disagreeing with... |
30860cf0-0e78-4cf6-b1ee-e940062ca466 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Role playing game based on HPMOR in Moscow
Discussion article for the meetup : Role playing game based on HPMOR in Moscow
WHEN: 15 April 2017 03:00:00PM (+0300)
WHERE: Большая Дорогомиловская, 5к2
Welcome to our custom made role playing game "Hogwarts and methods of rationality", inspired by HPMOR story! I... |
e39fcd08-ceb4-4764-9894-55c8ac9e3268 | trentmkelly/LessWrong-43k | LessWrong | Proportional Giving
Executive summary: The practice of giving a fixed fraction of one's income to charity is near-universal but possibly indefensible. I describe one approach that certainly doesn't defend it, speculate vaguely about a possible way of fixing it up, and invite better ideas from others.
--------------... |
0fbcc076-e0c3-47a9-ae0d-14199ec312c3 | trentmkelly/LessWrong-43k | LessWrong | Elastomeric Fitting Session
A well-fitting elastomeric respirator is something everyone should have on hand for emergencies. [1] Primarily for infection prevention, since a future pandemic could be much worse than covid, though they're also useful for smoke. There are a lot of options, though, and which one is best fo... |
a2f854c9-1d41-43a7-9a5f-8a32f65b6336 | trentmkelly/LessWrong-43k | LessWrong | How to Make Your Article Change People's Minds or Actions? (Spoiler: Do User Testing Like a Startup Would)
Are you writing an article that’s supposed to change people’s minds and/or actions? For example, an article about career advice, productivity, or getting people to apply for a job? Then this post is for you!
TL;... |
f6032ee6-eea1-4f1e-ba23-d704051fc1d5 | trentmkelly/LessWrong-43k | LessWrong | I am Bad at Flirting; Realizing that by Noticing Confusion
This post is about applying rationality to my dating life. It is gooey and rich in self-disclosure. But it was a great triumph over motivated reasoning.
I notice my confusion
When I was 22 I that my romances began mainly when I was busy. They began most ofte... |
9492553d-477c-41bc-b87c-15604d25cd1a | trentmkelly/LessWrong-43k | LessWrong | Ideal Number of Parents
I'll often hear people say with varying levels of seriousness that the ideal number of parents to have is something large, like maybe five. Children, especially infants, can be an enormous amount of work, which is certainly easier spread out among more people. Kids also like getting a lot of at... |
3deb08b5-3bda-4d3a-b974-2a386499d98d | trentmkelly/LessWrong-43k | LessWrong | Which College Major?
N.B.: This discussion isn't up for mainstream article status, as far as I'm concerned (unless someone else wants to take it and run with it). I just didn't know how else to direct an important question to the LW community in general.
I'm currently a first-year university student in Vancouver, Can... |
a0c571dd-a65e-4d83-9340-a7f3eed30549 | trentmkelly/LessWrong-43k | LessWrong | The RAIN Framework for Informational Effectiveness
[Epistemic status: Medium-Uncertainty. I've only spent a few days thinking about this, but it seems to fit well for some specific problem so far.]
The following describes one possible framework for understanding the usefulness of different sources of information. It'... |
34f860c8-8ed9-479d-8d30-ab5ed9ca8417 | StampyAI/alignment-research-dataset/distill | Distill Scientific Journal | Activation Atlas
Introduction
------------
Neural networks can learn to classify images more accurately than any system humans directly design. This raises a natural question: What have these networks learned that allows them to classify images so well?
Feature visualization is a thread of research that tries to... |
da00b936-53a8-435f-85b1-bab39d022cbe | trentmkelly/LessWrong-43k | LessWrong | [Retracted] Newton's law of cooling from first principles
This post is wrong. See this comment.
A high-school physics textbook will tell you that, according to Newton's law of cooling, a hot thing will cool down according to
dQdt∝T∞−T
where Q is heat, t is time, T is temperature, and T∞ is the temperature of the en... |
3dc7b707-45e8-4cba-aec4-2ee1d284430e | trentmkelly/LessWrong-43k | LessWrong | Has anyone actually changed their mind regarding Sleeping Beauty problem?
https://en.wikipedia.org/wiki/Sleeping_Beauty_problem
If that happened what was the argument that did it for you?
I'm interested in situations where a person used to think that the correct answer was 1/2 and then, on a reflection, decided that... |
44bc5e90-c64a-4a42-8b38-0f11772988d3 | trentmkelly/LessWrong-43k | LessWrong | What Noise Does a Tambourine Make?
I recorded myself playing around with a tambourine, and I was somewhat surprised to see what it looked like:
(mp3)
I'm thinking of it as four notes, and I hear four notes, but there are really more like eight. As I start to move my wrist to play a note there's a small jingle, and ... |
4df4c161-7cb6-4e5b-a508-d6e168f73a2c | trentmkelly/LessWrong-43k | LessWrong | Epistemic security: example from experimental physics
I was reading the introduction to a textbook on electrodynamics (J.D. Jackson), and in there was a description of an experiment designed to measure the exponent of the inverse square force law that governs forces between charges (Coulomb’s law). So, the experiment ... |
a900dc1e-444f-44dc-bd96-108482f5bf59 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Another (outer) alignment failure story
Meta
----
This is a story where the alignment problem is somewhat harder than I expect, society handles AI more competently than I expect, and the outcome is worse than I expect. It also involves inner alignment turning out to be a surprisingly small problem. Maybe the story is... |
1622aae0-a29a-468a-99f8-3af90785b5e1 | trentmkelly/LessWrong-43k | LessWrong | The Gemini Incident Continues
Previously: The Gemini Incident (originally titled Gemini Has a Problem)
The fallout from The Gemini Incident continues.
Also the incident continues. The image model is gone. People then focused on the text model. The text model had its own related problems, some now patched and some no... |
c328b55a-38e2-453c-a5bd-2bf412930159 | trentmkelly/LessWrong-43k | LessWrong | Anthropic is being sued for copying books to train Claude
OpenAI faces 10 copyright lawsuits and Anthropic is starting to get sued as well. Whether or not you agree with copyright, this is worth looking into. Lawsuits could hinder AI companies from scaling further.
The recent filing against Anthropic is notable bec... |
3ff92d64-ad63-4445-865a-32b266bb4943 | trentmkelly/LessWrong-43k | LessWrong | LINK: Subculture roles (by Brad Hicks)
By apparently popular demand
These roles do not represent personality types as such. A person who serves as an Authenticity Cop in one community may serve as a Fun Maven or even theoretically a Dream Nazi elsewhere. |
4e851188-6b9a-4ac9-af28-c0cfe2c6e68c | trentmkelly/LessWrong-43k | LessWrong | We need to revisit AI rewriting its source code
It feels like community discussion has largely abandoned the topic of AGI having the self-modifying property, which makes sense because there are a lot of more fundamental things to figure out.
But I think we should revisit the question at least in the context of narrow... |
66f73cf2-a4e1-475a-9a32-4245182c82f3 | trentmkelly/LessWrong-43k | LessWrong | Activation Costs
Enter Wikipedia:
> In chemistry, activation energy is a term introduced in 1889 by the Swedish scientist Svante Arrhenius, that is defined as the energy that must be overcome in order for a chemical reaction to occur.
In this article, I propose that:
* Every action you take has an activation cost ... |
8811d53d-3182-41fd-9c65-6d4ae0e6b882 | trentmkelly/LessWrong-43k | LessWrong | What nation did Trump prevent from going to war (Feb. 2025)?
In his meeting with Zelenskyy in the Oval Office, Trump briefly said
> I could tell you right now there's a nation thinking about going to war on something that nobody in this room has ever even heard about. Two smaller nations—but big, still big—and I thin... |
25c07531-d55c-4b3f-b2e7-ae6420b0f890 | LDJnr/LessWrong-Amplify-Instruct | LessWrong | "We’ve talked a lot about preparations and prevention, but statistically some of us, or people we care about, are going to actually get sick. What do we do once that happens?"
The above is a chunk of text from a post created by a user on the website LessWrong.
Could you elaborate on the specific measures one should ta... |
6c776954-9840-4020-a362-48c3c0adac4d | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Alignment Newsletter #30
Highlights
**[Learning Complex Goals with Iterated Amplification](https://blog.openai.com/amplifying-ai-training/)** *(Paul Christiano et al)*: This blog post and the accompanying [paper](https://arxiv.org/abs/1810.08575) introduces iterated amplification, focusing on how it can be used to de... |
a09f0ec8-afe0-48df-911e-dc9feebe454a | trentmkelly/LessWrong-43k | LessWrong | The End of Anonymity Online
Hello, I was hesitant to post this so I asked permission first. If mods feel it shouldn't be here please remove it and I will be fine with that. Essentially it's just me arguing with ChatGPT. I enjoy the format, but I can't trust anything it says obviously. So I would appreciate it if peopl... |
a6e35ee2-e028-416a-8fe0-244271cd7e19 | trentmkelly/LessWrong-43k | LessWrong | SWE Automation Is Coming: Consider Selling Your Crypto
[legal status: not financial advice™]
Most crypto is held by individuals[1]
Individual crypto holders are disproportionately tech savvy, often programmers
Source: Well known, just look around you.
AI is starting to eat the software engineers market
Already e... |
d3f17059-88d4-470f-9aa0-6afeac9026de | StampyAI/alignment-research-dataset/arbital | Arbital | Prime number
A [https://arbital.com/p/-45h](https://arbital.com/p/-45h) $n > 1$ is *prime* if it has no [divisors](https://arbital.com/p/divisor_number_theory) other than itself and $1$.
Equivalently, it has the property that if $n \mid ab$ %%note:That is, $n$ divides the product $ab$%% then $n \mid a$ or $n \mid b$.
... |
b6974ae6-5b3e-4f5e-860b-7b6c7f46537d | trentmkelly/LessWrong-43k | LessWrong | State of the Solstice (NY Event Upcoming)
> Every year people mock the Rationalist Solstice Ritual. And every year I read the discussions, nod, and then go to New York and have a great time and meet amazing people and have fascinating discussions and get the songs stuck in my head for the next couple of weeks.
>
> (“... |
49725912-0d78-45dd-8e07-dd9c2c3fbc88 | trentmkelly/LessWrong-43k | LessWrong | Meetup : DC Planning Meetup
Discussion article for the meetup : DC Planning Meetup
WHEN: 01 July 2012 03:00:00PM (-0400)
WHERE: Portrait Gallery Courtyard
We're meeting to discuss the future of the group!
Discussion article for the meetup : DC Planning Meetup |
ccb39ddf-ebbe-470a-836f-f5b4aab7ff3e | trentmkelly/LessWrong-43k | LessWrong | How to Get Off the Hedonic Treadmill?
I've been on a hedonic treadmill my whole life w/ respect to education, career, significant other, etc. I've started to realize that this is unsustainable but also can't shake the feeling that I'm settling. Any tips and perspectives on how to deal with this? |
e44d4fc7-68db-4000-94d1-cc1e09e8adf3 | StampyAI/alignment-research-dataset/blogs | Blogs | December 2022 updates and fundraising
*Harlan Stewart and *Katja Grace*\**, *22 December, 2022*
News
----
### New Hires and role changes
In 2022, the AI Impacts team has grown from two to seven full time staff. Out of more than 250 applicants, we hired Elizabeth Santos as Operations Lead, Harlan Stewart as Rese... |
11b252f1-0754-43ca-b946-f7c9d9b32fd1 | trentmkelly/LessWrong-43k | LessWrong | Why We Can't Take Expected Value Estimates Literally (Even When They're Unbiased)
Note: I am cross-posting this GiveWell Blog post, after consulting a couple of community members, because it is relevant to many topics discussed on Less Wrong, particularly efficient charity/optimal philanthropy and Pascal's Mugging. Th... |
af1107ca-7bd9-4e60-b807-c2dfe0729a3b | trentmkelly/LessWrong-43k | LessWrong | Do yourself a FAVAR: security mindset
(We're always in search of a concise and actionable way to communicate security mindset. Here's my attempt.)
Just as a sprocket you don't use never fails and code you don't write has no bugs, an input constraint you don't require is never violated.
Want a secure system? Then DO ... |
17671357-da4b-480d-9711-dd94293b4bf9 | trentmkelly/LessWrong-43k | LessWrong | The application of the secretary problem to real life dating
The following problem is best when not described by me:
https://en.wikipedia.org/wiki/Secretary_problem
Although there are many variations, the basic problem can be stated as follows:
There is a single secretarial position to fill.
There are n appl... |
f02d8408-aea0-4c72-acf8-9e302938fbf2 | trentmkelly/LessWrong-43k | LessWrong | Is there some reason LLMs haven't seen broader use?
When GPT-3 first came out, I expected that people would use it as a sort of "common-sense reasoning module". That is, if you want to process or generate information in some way, then you can give GPT-3 a relevant prompt, and repeatedly apply it to a bunch of differen... |
2691b665-0e6a-4a10-86da-330610b56c8f | trentmkelly/LessWrong-43k | LessWrong | Meetup : Washington, D.C.: Autism
Discussion article for the meetup : Washington, D.C.: Autism
WHEN: 21 May 2017 03:30:00PM (-0400)
WHERE: Donald W. Reynolds Center for American Art and Portraiture
This week, we will be meeting in the courtyard to discuss the autism spectrum, including some of the ways autistic peo... |
de12baaa-ebd5-4171-8b64-cf65c8f84cb2 | trentmkelly/LessWrong-43k | LessWrong | If it looks like utility maximizer and quacks like utility maximizer...
...it's probably adaptation executer.
We often assume agents are utility maximizers. We even call this "rationality". On the other hand in our recent experiment nobody managed to figure out even approximate shape of their utility function, and we... |
346143bf-a13f-45dc-ac87-e2bf71204025 | trentmkelly/LessWrong-43k | LessWrong | A review of "Agents and Devices"
In the small group of people who care about deconfusing goal-directedness and like Daniel Dennett's intentional stance, there's one paper everyone mentions: Agents and Devices, by Orseau, McGill and Legg. It's also pretty much the only paper ever mentioned on that topic.
Given that th... |
d28394e9-aca3-489a-873e-85d5de3912fc | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | AI Forecasting Question Database (Forecasting infrastructure, part 3)
*This post introduces [an open-source database of 76 questions about AI progress, together with detailed resolution conditions, categorisations and several spreadsheets compiling outside views and data](https://airtable.com/shrE9dnTVMJnbaYqs), as we... |
9f0cd9c5-1d55-4751-91ce-567e2aff1c61 | trentmkelly/LessWrong-43k | LessWrong | FYI: I’m working on a book about the threat of AGI/ASI for a general audience. I hope it will be of value to the cause and the community
The TLDR is the title of this post.
Hi all, long time EA/Rationalist, first time poster (apologies if formatting is off). I’m posting to mention that I’m 30,000 words into a draft o... |
351a9573-9889-4248-a5b2-ab85f267be31 | trentmkelly/LessWrong-43k | LessWrong | What determines the balance between intelligence signaling and virtue signaling?
Lately I've come to think of human civilization as largely built on the backs of intelligence and virtue signaling. In other words, civilization depends very much on the positive side effects of (not necessarily conscious) intelligence an... |
70e33c3b-67ab-4e33-aa19-e40ca011f4d7 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Where are the red lines for AI?
*Thanks to Daniel Kokotajlo, Jan Hendrik Kirchner, Remmelt Ellen, Berbank Green, Otto Barten and Olaf Voß for helpful suggestions and comments.*
As [AI alignment remains terribly difficult](https://www.lesswrong.com/posts/3pinFH3jerMzAvmza/on-how-various-plans-miss-the-hard-bits-of-the... |
0767ef50-2d64-4162-8fef-95e494e1eef5 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | AI Rights: In your view, what would be required for an AGI to gain rights and protections from the various Governments of the World?
* Easily pass the Turing Test by most Humans?
* Easily pass the Turing Test by most AI Researchers?
* Prove the ability to experience pain and pleasure, as well as have preferences? ... |
125a6b7e-53c4-4df3-a208-70eb869baf16 | trentmkelly/LessWrong-43k | LessWrong | Google could build a conscious AI in three months
Summary: Theories of consciousness do not present significant technical hurdles to building conscious AI systems. Recent advances in AI relate to capacities that aren't obviously relevant to consciousness. Satisfying major theories with current technology has been and ... |
dbbbcba8-50df-4550-8ab1-d57af05f3e65 | trentmkelly/LessWrong-43k | LessWrong | Do you want a first-principled preparedness guide to prepare yourself and loved ones for potential catastrophes?
Many of us have heard how we should prepare for potential disasters. But is such advice outdated given increasing threats from both AI and biotechnology? This post is both a test for people’s appetite for a... |
0e79a93e-ab2e-4a14-bcc4-4dc3173d6db2 | trentmkelly/LessWrong-43k | LessWrong | Card Rebalancing and Economic Considerations in Digital Card Games
Previously: Artifact Embraces Card Balance Changes, Card Collection and Ownership, Card Balance and Artifact (good comments on that one)
VII. Card Rebalancing, Pack Value and Economics
What happens when a game periodically rebalances its cards?
Ther... |
6bdf35e3-884a-448e-8b1b-7a037259a264 | trentmkelly/LessWrong-43k | LessWrong | What specific decision, event, or action significantly improved an aspect of your life in a value-congruent way? How?
On LessWrong and elsewhere in the Rationalsphere, it seems like there's more focus on the theoretical and philosophical aspects of rationality than on the concrete, actionable, and applied aspects of r... |
8d9a6e64-7211-4a69-9e76-26b0c3b67ad3 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Language models surprised us
*Note: This post was crossposted from* [*Planned Obsolescence*](https://www.planned-obsolescence.org/language-models-surprised-us/) *by the Forum team, with the author's permission. The author may not see or respond to comments on this post.*
> Most experts were surprised by progress in ... |
be399115-df69-4b12-919c-cd1371de3fb2 | trentmkelly/LessWrong-43k | LessWrong | Direct Realism is probably false
Phenomenology is a philosophical discipline that seeks to understand and describe human experiences as they are perceived. It always was a difficult topic to explore as people struggle to even understand what is being discussed; it does not attempt to study conscious experience the way... |
31b420ad-2a90-49ec-b034-91b9b6c3d0ed | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | What AI could mean for animals
*I’ve been working in animal advocacy for two years and have a very amateur interest in AI. I don’t necessarily agree with all of the possible implications of AI for animals that I’ve set out below, or endorse all of the references that I’ve cited; they’re just a collection of feasible-s... |
82d76294-8152-41c3-bf44-f7237b21fa80 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | What's Happening in Australia
*EDIT: Since first publishing I've added the Animals and Longtermism network, Animal Ask, High Impact Medicine Australia. Thanks to those groups!*
Introduction
------------
Crikey! There's a lot going on in 'Straya right now!
There are a lot of new and exciting projects happening in A... |
3dc18001-112b-46a4-b2a9-dc963531f329 | trentmkelly/LessWrong-43k | LessWrong | The Good Try Rule
I didn't really try the Zettelkasten Method. I jotted a few notes on cards, complete with indexing designations, over the course of a day. Then the notecards migrated to a back corner of my desk, and eventually made their way to the bottom of the Big Bin Of Office Supplies. Such has always been the f... |
3f3d92a2-274f-4787-af3d-b4a4f2a5e92d | trentmkelly/LessWrong-43k | LessWrong | Is there a way to sort LW search results by date posted?
|
a0bcdb76-2a35-49a8-b4df-8b8c41b80673 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Introducing a New Course on the Economics of AI
After having researched the topic as part of the [Centre for the Governance of AI (GovAI)](https://www.governance.ai/) for several years, I have made a six-week course on ["The Economics of AI"](https://www.coursera.org/learn/economics-of-ai/) freely available on Courser... |
07c01a8a-a069-432f-a0de-30e992a4bfac | trentmkelly/LessWrong-43k | LessWrong | Virtue ethics and why the rationalist community might care about it.
This post introduces “virtue ethics” and explains why you might find it worth pursuing.
Virtue ethics is meant to be of practical help
The two main forks of the rationalist endeavor are clear thinking and effective action. Virtue ethics addresses b... |
2a50dcdb-8421-4f45-8b9b-baf27c365270 | trentmkelly/LessWrong-43k | LessWrong | WTF is with the Infancy Gospel of Thomas?!? A deep dive into satire, philosophy, and more
A few weeks ago, you may have seen sensationalist headlines like "A New Discovery Could Offer Some Clues About Jesus’ Childhood", discussing a find of a 4th-5th century manuscript fragment of the Infancy Gospel of Thomas, a stran... |
5de914a6-5c25-428f-aed9-77559e0e171b | trentmkelly/LessWrong-43k | LessWrong | Burst work or steady work?
Some people work at a steady pace, below their Epic Maximum Work Capacity. Others try to keep up their Maximum Capacity, sometimes(?) burning out, but getting more done in the short bursts.
Over a time of, say, 6-18 months, which practice will lead to more work in the long run? Are there da... |
8b7478d0-b2b4-4f8d-bbe1-00c69e2fc931 | trentmkelly/LessWrong-43k | LessWrong | Alignment has a Basin of Attraction: Beyond the Orthogonality Thesis
While the Orthogonality Thesis is correct, there is a lot more that one can say about what kinds of agent motivations are likely to be encountered. A simple analysis shows that living agents produced by evolution, and constructed agents that are the ... |
ff2bf386-d3c6-43c8-ba8f-dfd2ef194fa1 | LDJnr/LessWrong-Amplify-Instruct | LessWrong | "In Rational Choice in an Uncertain World, Robyn Dawes describes an experiment by Tversky:1Many psychological experiments were conducted in the late 1950s and early 1960s in which subjects were asked to predict the outcome of an event that had a random component but yet had base-rate predictability—for example, subject... |
62959a7b-5735-4a38-892b-7f1132c491e2 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Madison: Probability Calibration
Discussion article for the meetup : Madison: Probability Calibration
WHEN: 08 July 2012 07:00:00PM (-0500)
WHERE: Michelangelo's: 114 State St., Madison, WI
I'll run a probability calibration exercise. How confident are you? How confident should you be?
I'll bring my Zendo... |
fc105757-ba5d-4477-b359-53c45b956297 | trentmkelly/LessWrong-43k | LessWrong | Low-hanging fruit: improving wikipedia entries
Many people are likely stumble across the Wikipedia entry for topics of interest relevant to those of us who frequent LessWrong: rationality, artificial intelligence, existential risks, decision theory, etc. These pages often shape one’s initial impressions of how interes... |
45529425-93b8-4474-b1e9-4c10cba021b9 | trentmkelly/LessWrong-43k | LessWrong | Cryonics signup guide #1: Overview
This is the introduction to a sequence on signing up for cryonics. In the coming posts I will lay out what you need to do, concretely and in detail. This sequence is intended for people who already think signing up for cryonics is a good idea but are putting it off because they're no... |
6259b459-eb73-4580-8cdd-a38f3ee564ba | trentmkelly/LessWrong-43k | LessWrong | Open Thread for January 17 - 23 2014
If it's worth saying, but not worth its own post (even in Discussion), then it goes here. |
1592014e-f8b3-47d9-ab96-034502ad4d9a | trentmkelly/LessWrong-43k | LessWrong | Simple Rules of Law
Response To: Who Likes Simple Rules?
Epistemic Status: Working through examples with varying degrees of confidence, to help us be concrete and eventually generalize.
Robin Hanson has, in his words, “some puzzles” that I will be analyzing. I’ve added letters for reference.
* A] People are often ... |
3d764585-a7fd-44f1-894e-f420e7017789 | trentmkelly/LessWrong-43k | LessWrong | MIRI's September 2024 newsletter
MIRI updates
* Aaron Scher and Joe Collman have joined the Technical Governance Team at MIRI as researchers. Aaron previously did independent research related to sycophancy in language models and mechanistic interpretability, while Joe previously did independent research related to AI... |
506864d0-c34d-4923-9a7f-f80975a25d71 | trentmkelly/LessWrong-43k | LessWrong | "Cars and Elephants": a handwavy argument/analogy against mechanistic interpretability
TL;DR: If we can build competitive AI systems that are interpretable, then I argue via analogy that trying to extract them from messy deep learning systems seems less promising than directly engineering them.
EtA - here's a follo... |
7da6bb90-24e5-440a-a20f-7a181617b5b5 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | A moral backlash against AI will probably slow down AGI development
*Note: This is a submission for the 2023 Open Philanthropy AI Worldviews contest, due May 31, 2023. It addresses Question 1: “What is the probability that AGI is developed by January 1, 2043?”*
**Overview**
People tend to view harmful things as e... |
51d54c81-c105-48a0-8dde-e12ed85fd612 | StampyAI/alignment-research-dataset/agentmodels | Tutorial: Modeling Agents with Probabilistic Programs | Modeling Agents with Probabilistic Programs
---
layout: chapter
title: "MDPs and Gridworld in WebPPL"
description: Noisy actions (softmax), stochastic transitions, policies, Q-values.
---
This chapter explores some key features of MDPs: stochastic dynamics, stochastic policies, and value functions.
### Hiking in Gri... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.