id stringlengths 36 36 | source stringclasses 15
values | formatted_source stringclasses 13
values | text stringlengths 2 7.55M |
|---|---|---|---|
51f2a113-25a2-4b98-ae81-4f442b346c34 | trentmkelly/LessWrong-43k | LessWrong | Logarithms and Total Utilitarianism
Epistemic status: I might be reinventing the wheel here
A common cause for rejection of total utilitarianism is that it implies the so-called Repugnant Conclusion, of which a lot has been written elsewhere. I will argue that while this implication is solid in theory, it does not ap... |
8f2bad96-2391-4623-aace-5a50c6e34eb8 | trentmkelly/LessWrong-43k | LessWrong | Who owns OpenAI's new language model?
OpenAI published about their very general language-writing & comprehension model (a) today.
Interestingly, they're not releasing the model itself:
> Due to concerns about large language models being used to generate deceptive, biased, or abusive language at scale, we are only re... |
7317a3c9-8c3c-4b76-8b38-a190fa5f223c | trentmkelly/LessWrong-43k | LessWrong | Link: Strong Inference
The paper "Strong Inference" by John R. Platt is a meta-analysis of scientific methodology published in Science in 1964. It starts off with a wonderfully aggressive claim:
> Scientists these days tend to keep up a polite fiction that all science is equal.
The paper starts out by observing th... |
d782cfad-cf82-42e3-ad16-d09b0b257ad6 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | “Pivotal Act” Intentions: Negative Consequences and Fallacious Arguments
**tl;dr:** I know a bunch of EA/rationality-adjacent people who argue — sometimes jokingly and sometimes seriously — that the only way or best way to reduce existential risk is to enable an “aligned” AGI development team to forcibly (even if nonv... |
a1bb8daf-8d7b-4d1e-b33e-bd1e0c5c9559 | StampyAI/alignment-research-dataset/youtube | Youtube Transcripts | Bing Chat (GPT 4) Tutorial: 12 Steps for Beginners to Advanced - Crash Course!
Bing chat is an extraordinary tool the
yes can make mistakes but which can also
do amazing things and give you a real
Edge in whatever you want to do this
crash course will take us from complete
beginners up to Advanced users following
12 s... |
03405dc9-2144-4639-88f1-f4d32884c041 | StampyAI/alignment-research-dataset/distill | Distill Scientific Journal | Computing Receptive Fields of Convolutional Neural Networks
While deep neural networks have overwhelmingly established state-of-the-art
results in many artificial intelligence problems, they can still be
difficult to develop and debug.
Recent research on deep learning understanding has focused on
feature visualiza... |
8b3d31a1-c246-49f5-ba70-46851eb52521 | trentmkelly/LessWrong-43k | LessWrong | ALBA requires incremental design of good long-term memory systems
Summary: ALBA is an approach to training aligned AGI. One problem with implementing ALBA is that some actions that must be overseen involve storing information that will be used in future episodes. To oversee these actions, it is necessary for the opera... |
775efd11-fb6d-4273-b1ca-13eb342d5e1f | StampyAI/alignment-research-dataset/aisafety.info | AI Safety Info | What are some objections to the importance of AI alignment?
[Søren Elverlin](https://aisafety.com/author/soeren-elverlin/) has compiled a list of counter-arguments and suggests dividing them into two kinds: "weak" and "strong".
Weak counter-arguments point to problems with the "standard" arguments (as given in, e.g.,... |
2ecfe20c-c73a-4da1-bd07-6ba48e09c003 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Antitrust-Compliant AI Industry Self-Regulation
Summary:
========
The touchstone of antitrust compliance is competition. To be legally permissible, any industrial restraint on trade must have sufficient countervailing procompetitive justifications. Usually, anticompetitive horizontal agreements like boycotts (includ... |
d61de9c4-734c-4f7e-9eb7-554b0fc1d142 | trentmkelly/LessWrong-43k | LessWrong | The Multiverse Interpretation of Quantum Mechanics [link]
http://arxiv.org/abs/1105.3796
http://www.newscientist.com/article/mg21028154.200-when-the-multiverse-and-manyworlds-collide.html?full=true |
d55a5843-6247-404c-892a-c2f56a626fea | trentmkelly/LessWrong-43k | LessWrong | Neutralizing Physical Annoyances
I have moved this post to my blog: http://squirrelinhell.blogspot.com/2016/09/neutralizing-physical-annoyances.html |
8021d966-1bd8-493b-a8b1-3ec614cc40cd | trentmkelly/LessWrong-43k | LessWrong | Making Sense of Coronavirus Stats
[Update: Some have pointed out the definition of mortality rate should be that of deaths to some defined population, typically a median estimate over a period of time (week, month, year...) and not limited only to those infected. (New update here. The ratios I have been considering ar... |
20cf23fc-829a-43a2-9b14-664f3b5e95ab | trentmkelly/LessWrong-43k | LessWrong | Burning Uptime: When your Sandbox of Empathy is Leaky and also an Hourglass
Epistemic status: idle musings
Some of us sometimes get so deeply caught up in a thought or understanding of someone else that it's difficult to switch back to our own context and experiences. This is what the mecha animation series Gen:Lo... |
fa0ec999-7aa2-4e6c-854d-9da71ebc3925 | trentmkelly/LessWrong-43k | LessWrong | Tradition is Smarter Than You Are
None |
321d784b-25fd-4bf2-93df-14d0e235a465 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Considerations on transformative AI and explosive growth from a semiconductor-industry perspective
This is a summary of my entry for the [Open Philanthropy AI Worldviews Contest](https://www.openphilanthropy.org/open-philanthropy-ai-worldviews-contest/), which you can view in full [here](https://muireall.space/pdf/co... |
b587db9a-02da-4e29-9015-fab76f76967c | trentmkelly/LessWrong-43k | LessWrong | A Framework of Prediction Technologies
Summary
This is the first piece of a blog post series that explores how AI prediction services affect the risks of war. It is based on my 10-week summer research project at Stanford Existential Risk Institute.
In this post, I will discuss different prediction technologies - how ... |
738f6847-9a51-4c9b-90b7-08e7a994bd97 | trentmkelly/LessWrong-43k | LessWrong | Book Review: Determined by Sapolsky
Determined: Life Without Free Will is a misguided attempt at moral reasoning based on scientific facts. Lacking a philosophical framework that can establish connections between morality and science, the author relied on his own rather lenient intuition without realizing it. One migh... |
81ce4117-c4bd-4c97-8092-49e633e15c19 | trentmkelly/LessWrong-43k | LessWrong | A Better Web is Coming
(Cross-posted from my blog)
It bums me out when I see how much of the Web is trolling, ads, and endless re-shares of yesteryear's memes. The cynic in me thinks that this is it, this is as good as it gets–we're stuck and cursed to never realize the full potential of this amazing, planet-spanning... |
0542e231-7622-45fe-9364-d2a8dcbb37bc | trentmkelly/LessWrong-43k | LessWrong | Behavioral psychology and buying a warranty at Menards
I just returned from buying a multimeter at Menards and wanted to post my thoughts while they were still fresh. I hardly ever have the need to use a multimeter. In diagnosing my non-heating microwave, I fried my 2-3 year old meter (don't ask) and went out for anot... |
4e5cd21d-1d96-419a-938f-6067509eb2fd | trentmkelly/LessWrong-43k | LessWrong | False vacuum: the universe playing quantum suicide
Imagine that the universe is approximately as it appears to be (I know, this is a controversial proposition, but bear with me!). Further imagine that the many worlds interpretation of Quantum mechanics is true (I'm really moving out of Less Wrong's comfort zone here, ... |
f70795ec-ba2e-4928-b96f-3ac7e6d88da3 | trentmkelly/LessWrong-43k | LessWrong | Meta: "Less Wrong" connotations?
I've always interpreted this site's title, "Less Wrong" (link is to description of origin of phrase), to name a goal that its members strive towards. After factoring in the slightly self-deprecating or human-deprecating connotations the title's message sorta feels like "We know it's im... |
1b1d84ea-3afe-417c-b9ea-b93a0f49584e | trentmkelly/LessWrong-43k | LessWrong | Welcome to Less Wrong!
If you've recently joined the Less Wrong community, please leave a comment here and introduce yourself. We'd love to know who you are, what you're doing, or how you found us. Tell us how you came to identify as a rationalist, or describe what it is you value and work to achieve.
If you'd like t... |
b59aeb6d-da8d-4f51-9148-31baf6e56190 | trentmkelly/LessWrong-43k | LessWrong | Melbourne Meetup May 6th, 6pm
Where: 55 Walsh St, West Melbourne, Australia
When: Friday May 6th, 6pm
We'll be discussing hacks for improving one's intelligence and energy levels (e.g. meditation, nootropics, paleo diet). If you know any tricks, or would like to learn some, come to the meetup!
|
7daf0c44-d681-4625-96bb-7a88194c91a8 | LDJnr/LessWrong-Amplify-Instruct | LessWrong | "The most important change between my game and Zvi's original is that bots can read each others' source code. They can simulate each other and predict each others' behavior. Within a day of the tournament launching—and an entire week before entries closed—Zack_M_Davis had already written a bot to simulate opponents and... |
eeee11d9-7304-4cec-be89-25b70788042c | trentmkelly/LessWrong-43k | LessWrong | Limitations on the Interpretability of Learned Features from Sparse Dictionary Learning
Overview
When I began my attempt to replicate Anthropic's Towards Monosemanticity paper, I had high expectations for how interpretable the extracted features should be. If I judged a feature to be uninterpretable, I attributed this... |
80eaa25b-ce72-4287-ad02-b59b09fd0990 | StampyAI/alignment-research-dataset/youtube | Youtube Transcripts | Reinforcement Learning 6: Policy Gradients and Actor Critics
okay let's get started basically I'm
going to continue where I left off last
lecture but we're also going to in some
sense dive into something a little bit
new so we touched upon some of the
topics in this lecture already in the
second lecture when we talked... |
5523569e-368b-4aaf-bc86-e21600915c95 | trentmkelly/LessWrong-43k | LessWrong | Open thread, Nov. 30 - Dec. 06, 2015
If it's worth saying, but not worth its own post (even in Discussion), then it goes here.
----------------------------------------
Notes for future OT posters:
1. Please add the 'open_thread' tag.
2. Check if there is an active Open Thread before posting a new one. (Immediately... |
00d78ab5-8646-4cec-9bf1-ea7b6cfbea2f | trentmkelly/LessWrong-43k | LessWrong | [SEQ RERUN] 31 Laws of Fun
Today's post, 31 Laws of Fun was originally published on 26 January 2009. A summary (taken from the LW wiki):
> A brief summary of principles for writing fiction set in a eutopia.
Discuss the post here (rather than in the comments to the original post).
This post is part of the Rerunn... |
4493c90e-6f37-466a-956d-65dad195e785 | trentmkelly/LessWrong-43k | LessWrong | In defence of epistemic modesty [distillation]
This is a distillation of In defence of epistemic modesty, a 2017 essay by Gregory Lewis. I hope to make the essay’s key points accessible in a quick and easy way so more people engage with them. I thank Gregory Lewis for helpful comments on an earlier version of this pos... |
da963a22-897a-42fb-a346-299a362afab8 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Team Shard Status Report
*Team Shard is a nebulous alignment research collective, on paper siloed under John Wentworth's SERI MATS program, but in reality extending its many tendrils far across the Berkeley alignment community. "Shard theory" -- a name spoken of in hushed, mildly confused tones at many an EA hangout. ... |
eea5f1aa-e693-4500-9997-52c75521ab8f | trentmkelly/LessWrong-43k | LessWrong | Welcome to Chicago Rationality
From January of 2017 through January of 2018, Chicago Rationality met weekly at either the University of Chicago or Harold Washington Library. The format of almost all events was to discuss assigned readings. We also held two large SSC meetups, which were more social, and a few applied r... |
9e607f4b-77fc-409f-a627-dad4af7edf4c | trentmkelly/LessWrong-43k | LessWrong | Take 5: Another problem for natural abstractions is laziness.
As a writing exercise, I'm writing an AI Alignment Hot Take Advent Calendar - one new hot take, written every day for 25 days. Or until I run out of hot takes.
Soundtrack.
Natural abstractions are patterns in the environment that are so convenient and so ... |
e20fe8ae-9aaf-4c05-9ff0-9b4c9f01bf52 | StampyAI/alignment-research-dataset/blogs | Blogs | Cold Links: misc
A long time ago, back in November of 2021, people made some interesting tweets about Cold Takes posts that I wanted to respond to, but I don't tweet. So I decided to try [this goofy thing](https://www.cold-takes.com/replies-to-ancient-tweets/) where I excerpt the tweets and give my replies on a Cold T... |
376cc763-4d8d-42b8-bf3a-c206acc5d1a5 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Atlanta: November Meetup (First of Two)
Discussion article for the meetup : Atlanta: November Meetup (First of Two)
WHEN: 03 November 2013 07:00:00PM (-0400)
WHERE: 1314 Hosea L Williams Dr. NE, Atlanta
Come join us! We'll be doing our normal eclectic mix of self-improvement brainstorming, educational mini... |
e50b8bc5-84a5-482e-add9-b1e7515a3ca4 | trentmkelly/LessWrong-43k | LessWrong | Rationality Reading Group: Part Q: Joy in the Merely Real
This is part of a semi-monthly reading group on Eliezer Yudkowsky's ebook, Rationality: From AI to Zombies. For more information about the group, see the announcement post.
----------------------------------------
Welcome to the Rationality reading group. Thi... |
608e1934-9860-4d0b-87f1-2af1aa5868eb | trentmkelly/LessWrong-43k | LessWrong | Investigating the rumors of OpenAI achieving AGI
There is rumor that OpenAI has achieved AGI internally. https://twitter.com/Simeon_Cps/status/1706078819617063304
If they do have AGI, i assume you would like to know that.
And i assume you would like to know more accurate AGI timeline.
Lets work together, to investi... |
8f02ef6d-e8a7-48c0-aca7-7299c8c0998e | trentmkelly/LessWrong-43k | LessWrong | Time-logging programs and/or spreadsheets
I currently log the total number of hours I work each day in an OpenOffice Spreadsheet. I input the start time, lunch/break time, and end time, and it calculates the total hours worked. I'm not savvy enough to create this type of spreadsheet myself, so I looked through a la... |
cec7092c-32ae-410b-87fd-e9f1c3e8e0d3 | trentmkelly/LessWrong-43k | LessWrong | Open Thread: November 2009
This is, hopefully, the usual monthly thread for discussion of Less Wrong topics that have not appeared in recent posts.
(My apologies and promise to fix if I'm doing this wrong; I have not previously submitted an article and am creating this thread so that I can post a comment in it.)
|
3c7b4378-e8aa-46b5-96b6-926b626e68fc | trentmkelly/LessWrong-43k | LessWrong | Trying Mastodon
One of the comments on yesterday's post about Mastodon was someone on LW inviting me to join their instance, and I now have an account: @jefftk@schelling.pt [EDIT: moved to @jefftk@mastodon.mit.edu after schelling.pt disappeared]. Right now everything is pretty empty: I don't have a good way to find pe... |
5b669b45-6864-403a-9b54-82e2b76bd59b | trentmkelly/LessWrong-43k | LessWrong | Literature regarding epidemics and political stability?
A story in today's news about COVID-19 and Iran got me wondering about the impact of such events on more centralized and totalitarian type government. Does anyone know of investigation into or theory about how an outbreak such as COVID-19, particularly if it take... |
865f0bb8-1184-4ed8-a03f-053e8bba38e6 | trentmkelly/LessWrong-43k | LessWrong | Depth-based supercontroller objectives, take 2
Thanks for all the helpful comments and discussion around this post about using logical depth as an objective function for a supercontroller to preserve human existence.
As this is work in progress, I was a bit muddled and stood duly corrected on a number of points. I'm ... |
2094b2d9-be3f-4940-8807-144c723cf09a | trentmkelly/LessWrong-43k | LessWrong | 2016 LessWrong Diaspora Survey Analysis: Part Four (Politics, Calibration & Probability, Futurology, Charity & Effective Altruism)
INTRODUCTION HERE
Politics
The LessWrong survey has a very involved section dedicated to politics. In previous analysis the benefits of this weren't fully realized. In the 2016 analysis ... |
54344552-2e24-4bd8-a3c4-148ad3d52dbc | StampyAI/alignment-research-dataset/youtube | Youtube Transcripts | How social science research can inform AI governance | Baobao Zhang | EAGxVirtual 2020
hello and welcome to the session on how
social science research can inform AI
governance with Baba Jeong my name is
more in the wing and I'll be an emcee
for the session thanks for tuning in
guys so we're first start with a
ten-minu... |
595e4ce3-2111-4b5c-be06-6fbc4d5859c1 | trentmkelly/LessWrong-43k | LessWrong | Is friendly-AI a good idea?
I know that to ask this question on this site is tantamount to heresy, and I know that the intentions are pure; to save uncountable human lives, but I would say that we are allowing ourselves to become blinded to what it is we are actually proposing when talk of building an FAI. The reas... |
019992cd-24fb-4bd0-95f2-3014b54e6a64 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Probabilities, weights, sums: pretty much the same for reward functions
.mjx-chtml {display: inline-block; line-height: 0; text-indent: 0; text-align: left; text-transform: none; font-style: normal; font-weight: normal; font-size: 100%; font-size-adjust: none; letter-spacing: normal; word-wrap: normal; word-spacing: n... |
4459e3c5-2fda-49db-b6ea-e6e869c8afd1 | trentmkelly/LessWrong-43k | LessWrong | The (not so) paradoxical asymmetry between position and momentum
(Cross-posted from my blog with light edits)
Related to: Which Basis is More Fundamental?
Epistemic status: Not an expert on physics, please point out any mistakes.
In physics, there appears to be a deep duality between position and momentum, in the s... |
3ffb726a-8593-4738-b8e6-bd8a7e390470 | trentmkelly/LessWrong-43k | LessWrong | Taboo "procrastination"
There are specific reasons for why you are procrastinating. There are specific solutions to these reasons. Bundling it all up into the "procrastination" black box, or offering solutions to the "procrastination" problem, is a problem. Specific diagnosis is not always straightforward, but it'll ... |
1040e9ac-3f6c-451a-8bd0-86c8a8453a4d | trentmkelly/LessWrong-43k | LessWrong | Meetup : Bratislava Meetup V.
Discussion article for the meetup : Bratislava Meetup V.
WHEN: 22 July 2013 06:00:00PM (+0200)
WHERE: Lanai, Dunajská 25, Bratislava
This time we try a new place: Cafe Lanai. We will have a lecture (in Slovak language) about rationality in human relations and education of children.
St... |
fd974fc1-4c1b-429e-82be-cf983d0562aa | StampyAI/alignment-research-dataset/arxiv | Arxiv | Robustness of different loss functions and their impact on networks learning capability
Robustness of different loss functions and their impact
on network’s learning capability
Vishal Rajput1,∗
1Computer Science Department, KU Leuven, Belgium
∗Corresponding author. Email: vishal.rajput@student.kuleuven.be
A... |
51983db9-d3e9-4506-944d-f6c28430b240 | trentmkelly/LessWrong-43k | LessWrong | What are some ways in which we can die with more dignity?
Eliezer's recent kidding not kidding death with dignity post suggests that our chances of survival are so low that we should just focus on going out with some semblance of dignity (ie. "at least we made an attempt that wasn't truly pathetic).
Or at least that'... |
9a2bf553-dae5-4704-bbac-d3d5e7f9a506 | trentmkelly/LessWrong-43k | LessWrong | A Pattern Language For Rationality
There's a lens to looking at the rationality project that I've been finding enlivening recently, and I think it's reached the point where more eyes and hands might be useful, while not being anywhere near complete yet. First, some background.
Christopher Alexander was a designer and... |
f63851d3-4f2b-43d0-843e-95146bc2020b | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Il panorama della governance lungoterminista delle intelligenze artificiali
*This is an Italian translation of* [***The longtermist AI governance landscape: a basic overview***](https://forum.effectivealtruism.org/posts/ydpo7LcJWhrr2GJrx/the-longtermist-ai-governance-landscape-a-basic-overview)
*Obiettivo: fornire un... |
71548812-dff5-41a6-bafd-1a4e2610c3b0 | trentmkelly/LessWrong-43k | LessWrong | Tao, Kontsevich & others on HLAI in Math
I found this 2015 panel with Terence Tao and some other eminent mathematicians to be interesting. The panel covered various topics but got into the question of when computers will be able to do research-level mathematics. Most interestingly, Maxim Kontsevich was alone in predic... |
30ec12d4-e9ab-4a5b-9696-64d3f0fd439a | trentmkelly/LessWrong-43k | LessWrong | Gradient descent might see the direction of the optimum from far away
Thanks to Justis, Arthur Conmy, Neel Nanda, Joseph Miller, and Tilman Räuker for their feedback on a draft.
I feel like many people haven't noticed an important result of mechanistic interpretability analysis of grokking, and so haven't updated how... |
78bfd8d9-3945-4a0b-882d-dd0bc14d51b8 | StampyAI/alignment-research-dataset/arxiv | Arxiv | Safe Transformative AI via a Windfall Clause
1 Introduction
---------------
Progress in artificial intelligence may lead to transformative artificial intelligence (TAI), here defined as artificial intelligence powerful enough to affect “practically irreversible change that is broad enough to impact most important a... |
1dc53d5f-72a0-4607-8726-ca46434f4dfe | trentmkelly/LessWrong-43k | LessWrong | Can we isolate neurons that recognize features vs. those which have some other role?
While I think about interpretability a lot, it's not my day job! Let me dive down a rabbit hole and tell me where I am wrong.
Intro
As I see it the first step to interpretability would be isolating which neurons perform which role. ... |
6d77fdfc-580a-4fcd-8e39-0cd69c6e44cf | trentmkelly/LessWrong-43k | LessWrong | AI acceleration, DeepSeek, moral philosophy
(Cross-posted from my website)
So, AI is accelerating. Does humanity have a future, and if so, what?
I consider myself a logical person, but this is not a logical piece of writing. It’s an attempt to share the contours of what I feel when I feel my thoughts about the futur... |
27708674-61a8-440e-80e9-ab96ddceb5a0 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | VLM-RM: Specifying Rewards with Natural Language
tl;dr
-----
We show how to use Vision-Language Models (VLM), and specifically CLIP models, as reward models (RM) for RL agents. Instead of manually specifying a reward function, we only need to provide text prompts like “a humanoid robot kneeling” to instruct and provi... |
8a650bbe-875b-4507-a88a-aa0d67145c0c | trentmkelly/LessWrong-43k | LessWrong | Group rationality diary, 10/29/12
This is the public group instrumental rationality diary for the week of October 29th. It's a place to record and chat about it if you have done, or are actively doing, things like:
* Established a useful new habit
* Obtained new evidence that made you change your mind about some be... |
13948bf2-34ef-4461-843d-cc3aab8eb730 | trentmkelly/LessWrong-43k | LessWrong | AI Needs Us? Information Theory and Humans as data
Assuming a superintelligent AI wanted to maximize its knowledge, I argue that a superintelligent AI must avoid destroying humanity.
I'll cover some information theory, how irreducible complexity means no perfect simulation of humans, and the "simming problem" from my... |
c0d4e504-3ec4-4e06-889f-e5cce52fe1df | trentmkelly/LessWrong-43k | LessWrong | Goal retention discussion with Eliezer
Although I feel that Nick Bostrom’s new book “Superintelligence” is generally awesome and a well-needed milestone for the field, I do have one quibble: both he and Steve Omohundro appear to be more convinced than I am by the assumption that an AI will naturally tend to retain its... |
9ea23094-d12a-43eb-bc10-64a0b456d3fc | trentmkelly/LessWrong-43k | LessWrong | Can we create self-improving AIs that perfect their own ethics?
If the process of self-improving AIs like described in an simple article by Tim Urban (below) is mastered, then the AI alignment problem is solved: "The idea is that we’d build a computer whose two-THREE major skills would be doing research on AI, ON ETHI... |
b0196c62-755c-48de-a27b-e8860d6555ae | StampyAI/alignment-research-dataset/special_docs | Other | Toward A Working Theory of Mind
Toward A Working Theory of Mind This is the first in a series of essays that are meant to lay out the models I’ve developed over the course of the last seven years of my practice. “My practice” is a coaching and guided self development technique, and much of its efficacy relies on the m... |
fa525019-f3e6-4b6a-8951-f6b6c38bce32 | trentmkelly/LessWrong-43k | LessWrong | AI Regulation is Unsafe
Concerns over AI safety and calls for government control over the technology are highly correlated but they should not be.
There are two major forms of AI risk: misuse and misalignment. Misuse risks come from humans using AIs as tools in dangerous ways. Misalignment risks arise if AIs take the... |
ae1932d8-82dd-47fc-b090-f43caa1f18f8 | trentmkelly/LessWrong-43k | LessWrong | We need to think more about Terminal Values
I just sent an email to Eliezer but it is also applicable to everyone on this website. Here it is:
Hi Eliezer,
I'm contacting you directly because I think I have an idea that could get us closer to obtaining our terminal value(s), and you are in a position of influe... |
bf1e9834-a779-498b-bd53-1664b5958fec | StampyAI/alignment-research-dataset/blogs | Blogs | AI Safety Highlights from NIPS 2016
This year’s Neural Information Processing Systems conference was larger than ever, with almost 6000 people attending, hosted in a huge convention center in Barcelona, Spain. The conference started off with t... |
09eed828-3567-44e4-8692-8fb303a1302d | trentmkelly/LessWrong-43k | LessWrong | How do you approach the problem of social discovery?
As in, how do you find ways to meet the right people you talk to? Presumably, they would have personality fit with you, and be high on both intelligence and openness. Furthermore, they would be in the point of their life where they are willing to spend time with you... |
fd766740-07f7-481e-9a67-405c39164cfd | trentmkelly/LessWrong-43k | LessWrong | Why is Mencius Moldbug so popular on Less Wrong? [Answer: He's not.]
I've seen several people on Less Wrong recommend Mencius Moldbug's writings, and I've been curious about how he became so popular here. He's certainly an interesting thinker, but he's rather obscure and doesn't have any obvious connection to Less Wro... |
b5277535-dc59-48ff-9923-f73d98b1163d | StampyAI/alignment-research-dataset/youtube | Youtube Transcripts | EleutherAI IRG 220507: Interpretability Research for the Most Important Century
gonna
i guess i should say welcome evan even
though he's a regular here at this point
um evan's been working on a really cool
post to try to motivate further research
and interpretability and communicate its
potential uh
impacts to others ... |
4d6050c8-fcd4-46be-be7f-1fb12f306ea4 | trentmkelly/LessWrong-43k | LessWrong | How to find a lost phone with dead battery, using Google Location History Takeout
Or, wow, it's crazy how much Google knows about me.
(I'm writing this in part because I tried to Google variations of "how to find a lost phone with dead battery" and among all the articles in the search results that I clicked on, didn'... |
c4a3e8cd-c8b6-46a6-a5e3-84ff79bac398 | trentmkelly/LessWrong-43k | LessWrong | Interactive model knob-turning
(Please discuss on LessWrong 2.0)
(Cross-posted from my medium channel)
When you are trying to understand something by yourself, a useful skill to check your grasp on the subject is to try out the moving parts of your model and see if you can simulate the resulting changes.
Suppose yo... |
c3327ead-fa5d-4ee7-bdc4-0b8aa35c2f6b | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Your Preferences
Hello, I am curious about your most basic preferences. I would like you to state them as many as possible.
You can use relation signs (<, >, ...), E. g.: true knowledge > ignorance > false knowledge, which would mean that you prefer true knowledge over ignorance, and ignorance over false knowledge. |
6c1c5fa9-62e0-4066-808b-8f84fff4191a | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Multimodal Neurons in Artificial Neural Networks
A paper investigating how individual neurons in a [CLIP model](https://openai.com/blog/clip/) (an image/text neural net combining a ResNet vision model with a Transformer language model) respond to various abstract concepts. This shouldn't be very surprising after GPT-3... |
52d1626b-1d91-496c-ba89-45c29893be1f | trentmkelly/LessWrong-43k | LessWrong | More art, less stink: Taking the PU out of PUA
Overview: This is a proposal for a LessWrong Pick Up Artist (PUA)-like sub-community; PUA without the PU (get it?)1. Members would focus on the deliberate practice of social artistry, but with non-mating goals. Origins and intent of the goal are discussed, possible topic... |
f62ff72c-6e79-464c-aecf-9a79965e3ca4 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Preface to the sequence on economic growth
On Lesswrong, when we talk about artificial intelligence, we tend to focus on the technical aspects, such as potential designs, specific developments, and future capabilities. From an engineering perspective, this focus makes sense. But most people here aren't interested in a... |
f66f8412-c3dc-48c5-81eb-39f1abcd7f49 | StampyAI/alignment-research-dataset/special_docs | Other | Information gathering actions over human internal state
I. introduction Imagine driving on the highway. Another driver is in the lane next to you, and you need to switch lanes. Some drivers are aggressive and they will never brake to let you in. Others are more defensive and would gladly make space for you. You don't ... |
9489576b-d2ba-4057-ae97-260dc67f0cbc | trentmkelly/LessWrong-43k | LessWrong | AGI systems & humans will both need to solve the alignment problem
Epistemic status: brainstorm-y musings about goal preservation under self-improvement and a really really bad plan for trading with human-level AGI systems to solve alignment.
When will AGI systems want to solve the alignment problem?
At some point, ... |
239b6ebb-de17-4ab2-9b8c-c24c3f3e3555 | trentmkelly/LessWrong-43k | LessWrong | How feasible is long-range forecasting?
Lukeprog posted this today on the blog at OpenPhil. I've quoted the opening section, and footnote 17 which has an interesting graph I haven't seen before.
----------------------------------------
How accurate do long-range (≥10yr) forecasts tend to be, and how much should we r... |
6a0c46d0-3ca0-48d6-b6d8-930c91353cc4 | trentmkelly/LessWrong-43k | LessWrong | What is the solution to the Alignment problem?
"The alignment problem is the tendency for text to appear blurry or misaligned on some monitors. The solution is to use a monitor with an anti-glare coating or a screen protector."
-GPT 3 |
8649a6d3-5b4e-4d91-a25b-af1b8a99e573 | trentmkelly/LessWrong-43k | LessWrong | Who invented knitting? The plot thickens
Last time on Eukaryote Writes Blog: You learned about knitting history.
You thought you were done learning about knitting history? You fool. You buffoon. I wanted to double check some things in the last post and found out that the origins of knitting are even weirder than I ... |
52eccaa4-8650-404f-ba5e-b0466e443d16 | trentmkelly/LessWrong-43k | LessWrong | The Importance of Those Who Aren't Here
One concept that people often miss when structuring groups, organizations, or social spaces is the importance of those who aren't here. People who aren't actively present can be just as important for the medium- and long-term future of your project as those who happen to be arou... |
308d7e1b-c678-4bd8-94f9-a8bf26d3a3dd | trentmkelly/LessWrong-43k | LessWrong | An elegant proof of Laplace’s rule of succession
(Already know what Laplace's rule of succession is and just want the proof? Scroll down.)
Suppose I give you a bag of marbles and tell you that all the marbles are either green or black. You repeatedly reach in, pick out a random marble, and then put it back. You find ... |
c55a0db5-ded3-4e88-8b21-26c648b7f2dc | trentmkelly/LessWrong-43k | LessWrong | Meetup : LW Sao Paulo - Meetup de maio
Discussion article for the meetup : LW Sao Paulo - Meetup de maio
WHEN: 21 May 2016 02:00:00PM (-0300)
WHERE: Vanilla café - 2o andar, Rua Antônio Carlos, 452. São Paulo, Brazil.
Encontro mensal de racionalistas, novatos são bem vindos! Teremos algumas lições de racionalidade ... |
cc37d2d8-a8f2-4e0a-9e1b-dc14f4577ffd | trentmkelly/LessWrong-43k | LessWrong | Attribution-based parameter decomposition
This is a linkpost for Apollo Research's new interpretability paper:
"Interpretability in Parameter Space: Minimizing Mechanistic Description Length with Attribution-based Parameter Decomposition".
We introduce a new method for directly decomposing neural network parameters... |
2831584f-ad2e-40f8-a356-2879959e7500 | trentmkelly/LessWrong-43k | LessWrong | How Does Cognitive Performance Translate to Real World Capability?
How does cognitive/intellectual performance (e.g. as measured by −log(error rate) translate to real-world capability? Do linear increases in cognitive performance result in linear increases in capability? I don't know. I did think of a way we could may... |
9c200f7e-5f62-4414-bb5b-a39de924b7c0 | trentmkelly/LessWrong-43k | LessWrong | What Motte and Baileys are rationalists most likely to engage in?
Motte and Bailey is a concept by Nicholas Shackel that Scott Alexander has popularised. He describes this as follows:
> The original Shackel paper is intended as a critique of post-modernism. Post-modernists sometimes say things like “reality is social... |
193c1e05-a36a-4dba-88ff-324c0a21278a | trentmkelly/LessWrong-43k | LessWrong | Examples of positive-sum(ish) games?
How many positive-sum(ish) games, or PSIGs, exist?
For discussion, I'll be using a broader definition than normally used of positive-sum games:
* Someone invents a magic free energy machine that outputs energy from nothing, breaking the laws of thermodynamics == PSIG
* Two peo... |
fd4c30d0-9788-4348-9058-808db6f5edec | trentmkelly/LessWrong-43k | LessWrong | AALWA: Ask any LessWronger anything
If you want people to ask you stuff reply to this post with a comment to that effect.
More accurately, ask any participating LessWronger anything that is in the category of questions they indicate they would answer.
If you want to talk about this post you can reply to my comment b... |
d65137de-0d43-4137-a4e3-2c9692afc2b2 | StampyAI/alignment-research-dataset/youtube | Youtube Transcripts | AI governance landscape | Carrick Flynn | EA Global: San Francisco 2018
In my previous career, I worked in global
poverty eradication, mostly doing field work.
This meant working on projects all around
the world, trying to find a place where I
felt I could do the most good and have the
largest impact.
A little over th... |
39ad09e6-4bef-4df4-a33b-2dfc956bc84e | StampyAI/alignment-research-dataset/special_docs | Other | The evolution of cognitive mechanisms in response to cultural innovations.
The evolution of cognitive mechanisms in response to
cultural innovations
Arnon Lotema, Joseph Y. Halpernb, Shimon Edelmanc, and Oren Kolodnyd,1
aDepartment of Zoology, Tel Aviv University, Tel Aviv 6997801, Israel;bDepartment of Computer Scien... |
7e15c39a-ee9b-43ff-b800-0af0bbbe5b7a | trentmkelly/LessWrong-43k | LessWrong | The new UK government's stance on AI safety
tl;dr: The new UK government will likely continue to balance encouraging AI innovation for public good against increasing regulation for public safety, with so-far rhetorical calls for stricter regulation than the previous government’s. Several reports have been published by... |
bbfbfab6-b138-49da-9613-9c4ca5f85140 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Is ELK enough? Diamond, Matrix and Child AI
Introduction
============
If we could solve [ELK](https://www.alignmentforum.org/posts/qHCDysDnvhteW7kRd/arc-s-first-technical-report-eliciting-latent-knowledge), would we always know where the diamond is at the end of a plan? That is, if we could get the direct translator,... |
1768797a-ae16-407b-b4f2-ea090e9b75ac | trentmkelly/LessWrong-43k | LessWrong | How much does the risk of dying from nuclear war differ within and between countries?
It seems that being in a major city such as London or New York makes it much more likely that one dies in case of nuclear war than in the countryside or in a smaller town.
Also, some people seem to think that certain countries are ... |
09c856a3-7e4d-432a-b4df-b44af8bb320a | trentmkelly/LessWrong-43k | LessWrong | Breaking the chain of akrasia
I'd like to share my specific motivation for writing Can the Chain Still Hold You?
I agree with Yvain that akrasia is probably a major reason that rationality alone doesn't create superheroes. You might be much better than average at making good decisions based on an accurate model of re... |
f53d2a20-3c9f-409f-a467-20ec3fc3a6c2 | StampyAI/alignment-research-dataset/blogs | Blogs | MIRI’s September Newsletter
| | |
| --- | --- |
|
| |
| --- |
| |
|
|
| | | | |
| --- | --- | --- | --- |
|
| | | |
| --- | --- | --- |
|
| | |
| --- | --- |
|
| |
| --- |
|
Greetings from the Executive Director
Dear friends,
With your help, we [finished](http://intelligence.org/2013/08/21/... |
739d02bd-8af8-43c8-8c8b-01249ab8b6ff | trentmkelly/LessWrong-43k | LessWrong | Countably Factored Spaces
A followup to Scott's "Finite Factored Sets", specifically the Applications part at the end where he talked about the infinite case.
As it turns out, there's a natural-seeming (at least to my mathematical tastes) way to generalize all the finite factored set stuff to the infinite case.
To b... |
4c396976-be5e-4c81-8b31-02d11a94dfed | trentmkelly/LessWrong-43k | LessWrong | Can moderators fix old sequences posts?
I'm re-reading sequences now and I'm noticing eye-opening things like so many have SEQ RERUN, useless copies that clutter up the link space while having very few comments that they seem to be intended for. Can the moderators do something about this? Remove links to them, maybe e... |
e93e59fe-1d97-4995-8d98-8310dc7bcd14 | trentmkelly/LessWrong-43k | LessWrong | LessWrong needs a sage mechanic
Most forums work in a similar way: if you reply to a post, it gets bumped to the top of the feed. This is usually good, from the perspective of the commenter; people generally want others to see their comments.
There are obvious problems, though. On every forum with enough volume, ther... |
c0d92765-af6d-4d4c-9d7b-299a04a83b45 | trentmkelly/LessWrong-43k | LessWrong | Perform Tractable Research While Avoiding Capabilities Externalities [Pragmatic AI Safety #4]
This is the fourth post in a sequence of posts that describe our models for Pragmatic AI Safety.
We argued in our last post that the overall AI safety community ought to pursue multiple well-reasoned research directions at o... |
765a777b-a14f-4609-a352-547b3f1a0062 | awestover/filtering-for-misalignment | Redwood Research: Alek's Filtering Results | id: post2115
This post was written under Evan Hubinger’s direct guidance and mentorship, as a part of the Stanford Existential Risks Institute ML Alignment Theory Scholars (MATS) program . Additional thanks to Michele Campolo, Oliver Zhang, Adam Shimi, and Leo Gao for their thoughts and feedback on this post. This wor... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.