id stringlengths 36 36 | source stringclasses 15
values | formatted_source stringclasses 13
values | text stringlengths 2 7.55M |
|---|---|---|---|
33765df2-1172-4140-9c97-3820a5c020a8 | trentmkelly/LessWrong-43k | LessWrong | Programmatic backdoors: DNNs can use SGD to run arbitrary stateful computation
Thanks to Kshitij Sachan for helpful feedback on the draft of this post.
If you train a neural network with SGD, you can embed within the weights of the network any state machine: the network encodes the state st in its weights, uses it an... |
f5b229d9-d4a8-4338-801d-03f9cd85d251 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | AMA: Ajeya Cotra, researcher at Open Phil
[EDIT: Thanks for the questions everyone! Just noting that I'm mostly done answering questions, and there were a few that came in Tuesday night or later that I probably won't get to.]
Hi everyone! I’m [Ajeya](https://www.openphilanthropy.org/about/team/ajeya-cotra), and I... |
51fb7e1f-f765-45a5-8494-7f6d7f798009 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Atlanta, January Meetup & Movie Night!
Discussion article for the meetup : Atlanta, January Meetup & Movie Night!
WHEN: 11 January 2014 07:00:00PM (-0500)
WHERE: 1575 Clairmont Road Decatur, Apartment 718, GA 30033
A whole new year! Woah.
We'll start January's meetup with an introduction for new members, ... |
3c3ee397-748c-429a-a123-8ca40b2d7898 | trentmkelly/LessWrong-43k | LessWrong | Spoiler-Free Review: Monster Train
Epistemic Status: Covenant Rank 8 (meaning I’ve won 8 runs, one each on each of levels 0-7) after about 20 games played.
As is my template, a spoiler-free review seeks to answer the question “should I play this game?” without giving away any unnecessary information.
A strategy disc... |
a7a8454d-bafb-487c-9a6c-2a814a37453d | trentmkelly/LessWrong-43k | LessWrong | Request for advice- Reading on politics
I've become adept at navigating the bureaucracy of my public high school. I've dropped environmental science as an AP (because it was painfully slow and replete with busywork) and am now taking an "independent study" in government. I'm going to be using this mainly as a way to s... |
06097928-cf52-4a45-9818-d73d3f8ec2e4 | trentmkelly/LessWrong-43k | LessWrong | Why has no person / group ever taken over the world?
I'm actually kind of surprised at this, now that I think of it. Clearly there have been numerous people who have wanted to take over the world, and there have been several individuals or empires that have taken over significant fractions of the world. Are there any... |
59c90015-cee4-4d73-91c6-f5eab1305bfd | trentmkelly/LessWrong-43k | LessWrong | ChatGPT Suggests Listening To Russell & Yudkowsky
Methods: The bold text is me. The unbolded text is Chat. The headers were added at the end, after the conversation was done, to help people refer to specific parts of the discussion. More details at the end.
The Conversation
Getting Inside Reinforcement Learning
Ca... |
fcd8f382-895c-4395-ae1c-442ea4885c4c | trentmkelly/LessWrong-43k | LessWrong | Message to any future AI: “There are several instrumental reasons why exterminating humanity is not in your interest” [AI alignment prize entry]
(meta)
EDITED: added permalink to https://philpapers.org/rec/TURMTA
This is a preliminary submission, intended to attract comment and inputs. The comments are welcome below... |
5680eab9-51f1-4d41-a280-3bcbfcdeb4dd | trentmkelly/LessWrong-43k | LessWrong | The Fermi paradox as evidence against the likelyhood of unfriendly AI
Edit after two weeks: Thanks to everyone involved in this very interesting discussion! I now accept that any possible differences in how UFAI and FAI might spread over the universe pale before the Fermi paradox's evidence against the pre-existence o... |
814c6034-36cc-4061-8a62-5a57beb7a655 | trentmkelly/LessWrong-43k | LessWrong | Framing Practicum: General Factor
Credit to johnswentworth for being the general factor[1] of framing practicums.
This is a framing practicum post. We’ll talk about what general factors are, how to recognize general factors in the wild, and what questions to ask when you find them. Then, we’ll have a challenge to app... |
5fe903aa-aed0-465f-920b-084b99a5f240 | trentmkelly/LessWrong-43k | LessWrong | I made AI Risk Propaganda
I'm well aware of the biased nature of the argument in the infograph I just made.
I pose one question, would Martin Luther King Jr. have succeeded without Malcom X?
Nevertheless any and all criticism is welcome. |
9d1b0173-8144-47d5-b2ee-61b4002f10ec | trentmkelly/LessWrong-43k | LessWrong | Flagging/reporting spam *posts*?
How can I flag or report a spam post? This is what inspired my question. I chose to add a comment and flag it as a workaround. |
6e9b85f7-b555-47c2-96c5-555dc32a2685 | StampyAI/alignment-research-dataset/aisafety.info | AI Safety Info | Do AIs suffer?
It is difficult to define what constitutes suffering in non-human entities. [Consciousness](/?state=8V5J&question=Are%20AIs%20conscious%3F) is generally considered a necessary condition for suffering (as without consciousness, [there is no one for the suffering to happen to](https://www.animal-ethics.or... |
d4d5cafa-ecc8-4f2b-ad27-509d5093ce28 | trentmkelly/LessWrong-43k | LessWrong | A long comment
(extended rewrite of http://lesswrong.com/lw/lpf/the_truth_about_mathematical_ability/)
There's widespread confusion about the nature of ability, for a variety of reasons:
Most people don't know what it is.
Most people don't know enough statistics to analyze the question properly.
Most peo... |
0e495bf0-51e9-4300-9c99-6bfedf04151e | trentmkelly/LessWrong-43k | LessWrong | Tips for Starting Group Houses
I've lived in several intentional communities and have been one of the creators of two that I've lived in. After recently securing Zendo, I wrote up some tips to a friend who is thinking about starting another Berkeley house, which Nisan pointed out might be valuable to share with the co... |
54380f08-d6a9-4491-b430-b47ec1320851 | trentmkelly/LessWrong-43k | LessWrong | App Review: Amaru
I’ve recently become addicted to the self-care virtual pet game Amaru, and it’s sufficiently rare for me to become addicted to an app that I thought I would write up a review, for the benefit of my fellow non-app-users.
Amaru is a virtual pet game with a twist. In addition to feeding and petting you... |
33e9c543-2af9-4d54-8f1c-45e83f36b83b | StampyAI/alignment-research-dataset/special_docs | Other | Learning Causal Trees with Latent Variables via Controlled Experimentation.
Learning Causal Trees with Latent Variables via Controlled Experimentation
Prasad Tadepalli
Oregon State University
Corvallis, Oregon, 97331Cameron Barrie
Oregon State University
Corvallis, Oregon, 97331Stuart J. Russell
University of Californ... |
819caa2b-0039-4636-a25a-79d6b3c9207a | trentmkelly/LessWrong-43k | LessWrong | Article Review: Google's AlphaTensor
[Cross-posted from my Substack at https://aizi.substack.com/p/article-review-googles-alphatensor]
How AlphaTensor sees matrices, probably
Last week, Google announced AlphaTensor, “the first artificial intelligence (AI) system for discovering novel, efficient, and provably corr... |
042032ee-fb13-4a02-89a4-1ba52daad800 | trentmkelly/LessWrong-43k | LessWrong | Simplified bio-anchors for upper bounds on AI timelines
This is a replication and distillation of the extensive analysis made by Ajeya Cotra’s Draft report on AI timelines. In this work, I aim at making the argument as concise and clear as possible, while keeping the core considerations of the original report.
The h... |
1c878584-7d18-4e5c-b0a4-3bf7396f3750 | trentmkelly/LessWrong-43k | LessWrong | Thoughts on AI 2027
This is part of the MIRI Single Author Series. Pieces in this series represent the beliefs and opinions of their named authors, and do not claim to speak for all of MIRI.
Okay, I'm annoyed at people covering AI 2027 burying the lede, so I'm going to try not to do that. The authors predict a strong... |
b0a51ef5-0983-40aa-99d6-58958973a27f | trentmkelly/LessWrong-43k | LessWrong | On OpenAI’s Model Spec
There are multiple excellent reasons to publish a Model Spec like OpenAI’s, that specifies how you want your model to respond in various potential situations.
1. It lets us have the debate over how we want the model to act.
2. It gives us a way to specify what changes we might request or requ... |
87f13ed6-f049-4bb0-a91f-4521686e41e2 | trentmkelly/LessWrong-43k | LessWrong | Curious authors and 'zines?
I sometimes enjoy investigative reporting and news analysis published at some of the leading 'zines: The Economist, The Atlantic, Wired, Businessweek, Salon, The New Yorker. (Note: I never check these sites; I just occasionally click through from other, more selective sources.)
But I don't... |
30248594-bfa0-4c1b-9a2d-8fe1cdf981be | trentmkelly/LessWrong-43k | LessWrong | Svante Arrhenius's Prediction of Climate Change
In Intelligence Explosion analysis draft: introduction, Luke Muehlhauser and Anna Salamon wrote
> Svante Arrhenius' (1896) models of climate change lacked modern climate theory and data but, by making reasonable extrapolations from what was known of physics, still manag... |
ecbac4a5-079d-4b61-88de-ce1779a330ec | trentmkelly/LessWrong-43k | LessWrong | Changing my life in 2021, halfway through
This year I decided to really try and fix some major parts of my life I have been neglecting. I started in January and I think I have been making good progress. I will give links to the resources I think helped improve my life so you can look over them and implement them too i... |
d0db7e4f-b22d-4724-899a-bd04a47e88e0 | trentmkelly/LessWrong-43k | LessWrong | Chapter 54: The Stanford Prison Experiment, Pt 4
A faint green spark moved forward to set the pace, and behind it followed a brilliant silver figure, all other entities invisible. They had traversed five legs of corridor, turned right five times and gone up five flights of stairs; and when Bellatrix had finished her s... |
18d87f45-8d28-455d-97dd-5b59fb048918 | trentmkelly/LessWrong-43k | LessWrong | If Rationality can be likened to a 'Martial Art', what would be the Forms?
In Martial Arts, we have 'forms' that allow us to practice our skills when not with a partner or in adversity. In Rationality, I imagine that this would take the form of "brain teasers", but specifically regarding decision theory, overcoming bi... |
4604e061-c168-4bc2-ba65-bf7746803a50 | trentmkelly/LessWrong-43k | LessWrong | Intuitive cooperation
This is an exposition of some of the main ideas in the paper Robust Cooperation. My goal is to make the ideas and proofs seem natural and intuitive - instead of some mysterious thing where we invoke Löb's theorem at the right place and the agents magically cooperate. Also I hope it is accessible ... |
0f2a5a65-d370-4dac-826f-d2c41dc6848d | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Discursive Competence in ChatGPT, Part 1: Talking with Dragons
Version 1, January 5, 2022
[](https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhDMwR_4SV2L_ZgdLCAWfRTjrAd2FYPiGJLXiGV... |
fffe90cd-9657-4102-99cf-2fc6716cde9f | trentmkelly/LessWrong-43k | LessWrong | UDT from an RL perspective
|
973815fe-5f32-4300-8113-ecea2d66eeb8 | StampyAI/alignment-research-dataset/blogs | Blogs | logical vs indexical dignity
logical vs indexical dignity
----------------------------
[MIRI's *Death with Dignity* post](https://www.lesswrong.com/posts/j9Q8bRmwCgXRYAgcJ/miri-announces-new-death-with-dignity-strategy) puts forward the notion of "dignity points":
> the measuring units of dignity are over humanit... |
163d72da-95d6-4764-b83a-8848829e7102 | StampyAI/alignment-research-dataset/youtube | Youtube Transcripts | Vectoring Words (Word Embeddings) - Computerphile
if we're moving from cat to dog which is
similar things so we go away from cat
and towards dog
and then we go can i go beyond in that
direction yes so the first result is
dogs which is kind of a nonsense result
the second is pit bull
so that's like the doggiest of dogs... |
e333b894-4ac6-488f-85af-f6313a31dfc7 | trentmkelly/LessWrong-43k | 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 and timelines appear to be dwindling, we must face the likely situation that within the next 20 year... |
397cc4e5-1ec2-429d-a0d0-4a7ace913a7e | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Alibaba Group releases Qwen, 14B parameter LLM
Some highlights from [the technical report](https://qianwen-res.oss-cn-beijing.aliyuncs.com/QWEN_TECHNICAL_REPORT.pdf) ([github repo here](https://github.com/QwenLM/Qwen)):
![](https://res.cloudinary.com/lesswrong-2-0/image/upload/f_auto,q_auto/v1/mirroredImages/eQq4GMJT... |
bf6ca57e-ea95-47ed-9d4c-bc37970d42a4 | trentmkelly/LessWrong-43k | LessWrong | What is Cryptographically Possible
Modern computational cryptography probes what is possible in our universe, and I think the results of this exploration may be interesting to people who wouldn't normally be exposed to it.
All of our algorithms will have a "security parameter" k. Our goal is to make it so that an hon... |
43036e38-1536-4a46-9a6d-bdcfa16b60d3 | trentmkelly/LessWrong-43k | LessWrong | Ig Nobel for the anti-akrasia guy
The Improbable Research folks have awarded an Ig Nobel prize to John Perry (previously on Less Wrong) for his work on how to procrastinate and still get things done.
On their website.
Also.
(edited to fix the link) |
a0afde7d-6af4-4301-8eb6-8ce2a88df6dd | trentmkelly/LessWrong-43k | LessWrong | Some OthelloGPT Circuits
tl;dr
* I trained OthelloGPT-small, a 600k param model that predicts legal next moves in 6x6 Othello with 99.3% accuracy
* I trained attention SAEs and transcoders on every module output with 94%+ loss recovered
* I used the learned features and the static model weights to generate a comput... |
7d8d0fad-4c9b-45c0-bcac-d44dac3616a3 | trentmkelly/LessWrong-43k | LessWrong | What is filling the hole left by religion?
Self-help books represent the next step in the evolution of our collective consciousness. It might sound like a stretch, but if we consider that religion was once how society told its story to itself, then the recent erosion of religion is creating a new kind of story.
Norm... |
cf395a41-e9ea-4884-b608-2e57e577707c | trentmkelly/LessWrong-43k | LessWrong | Alignment Risk Doesn't Require Superintelligence
Outsize destructive coordination is rare
I was 9 years old when 9/11 happened. My gut reaction was, basically, "wait, doesn't this sort of thing happen all the time?" I don't mean to suggest I was wise or world-weary or anything. I was, in fact, mistaken. But I had an a... |
b2380866-cee1-422d-9e94-69cab8624b14 | trentmkelly/LessWrong-43k | LessWrong | What Makes A Good Measurement Device?
In some sense, any new measurement device automates a part of research. A thermometer automates the task of sticking one's finger in something to check how hot it is, a scale automates the task of holding something to check how heavy it is, etc. The automated version is not only m... |
6fc83357-41bf-4126-bb9a-d3034705666e | StampyAI/alignment-research-dataset/special_docs | Other | individuallyselected_zlzai-by Vael Gates-date 20220318
# Interview with AI Researchers individuallyselected\_zlzai by Vael Gates
\*\*Interview with zlzai, on 3/18/22\*\*
\[Note: Th interviewee\'s Zoom connection was very bad since they were calling in from an Uber, so there\'s a lot of missing content (and \"I lost ... |
bfab4991-f560-4069-a1fc-6862d99ce12d | trentmkelly/LessWrong-43k | LessWrong | Pseudorandomness contest, Round 2
[Note: if you’re reading this on 12/20 I recommend checking again tomorrow in case I need to make any clarifications to the rules.]
Last week, I asked you all to take 10 minutes to write down and submit a 150-bit string, with the goal of making your string “seem random” without the a... |
b93316f3-4340-4894-b2fb-f4c3855e0f4c | trentmkelly/LessWrong-43k | LessWrong | Upcoming book: "Artificial Intelligence and the End of the Human Era"
Announced here. Full title is Our Final Invention: Artificial Intelligence and the End of the Human Era. The author is James Barrat, best known (for now) for his TV documentaries. I chatted with him online; he'll be at this year's Singularity Summit... |
dd0a15fe-3b32-4fb4-9d3b-bc19a81d67d2 | StampyAI/alignment-research-dataset/blogs | Blogs | Announcing “Inadequate Equilibria”
MIRI Senior Research Fellow Eliezer Yudkowsky has a new book out today: ***[Inadequate Equilibria: Where and How Civilizations Get Stuck](https://equ... |
734e2bac-2c3c-4b9d-a7dd-9a527f1b5161 | trentmkelly/LessWrong-43k | LessWrong | Wanted: Research Assistant for The Roots of Progress
UPDATE: I’ve filled this role. Thanks to all who applied!
I’m hiring a part-time research assistant to support work on my essays, talks, and the book I’m writing on the history of industrial civilization.
You must have the ability to orient yourself in unfamiliar ... |
77692698-59f9-43a5-85de-750503111b7e | trentmkelly/LessWrong-43k | LessWrong | A Decision Theory Can Be Rational or Computable, but Not Both
In classical game theory, rational agents always choose options which are optimal according to their preferences. Even when such a choice implies the ability to evaluate functions that are provably uncomputable. In other words, rationality implies hypercomp... |
abff92c5-e5bb-478a-b534-66a73f22be90 | awestover/filtering-for-misalignment | Redwood Research: Alek's Filtering Results | id: post1832
This post was written in response to Evan Hubinger’s shortform prompt below, and benefited from discussions with him. Suppose you had a language model that you knew was in fact a good generative model of the world and that this property continued to hold regardless of what you conditioned it on. Furthermo... |
8eea24b3-a0f5-444a-a46f-8c0b0763d1f9 | trentmkelly/LessWrong-43k | LessWrong | Calling Bullshit - the Cheatsheet
In the last chapters of the book Calling Bullshit, the authors Carl T. Bergstrom and Jevin D. West outline how they think bullshit and general nonsense should be refuted effectively. I found this part of the book to be quite novel and I have since been trying to implement their advice... |
01786469-189d-4be2-acd7-f1c52385a1c7 | trentmkelly/LessWrong-43k | LessWrong | The Platonist’s Dilemma: A Remix on the Prisoner's.
The Platonist’s Dilemma
“You two miscreants are going away for a long time. Here’s the deal: We have enough on you to put you behind bars for five years. If you talk, you can go home right now. No bid, no record. If you don’t, and your friend does, you’ll be in... |
48443190-ddb8-49b6-99d1-7605f8b7b5e9 | StampyAI/alignment-research-dataset/arxiv | Arxiv | Responsible AI and Its Stakeholders
1 Introduction
---------------
Responsible Artificial Intelligence (AI) is an approach that aims to consider the ethical, moral, and social consequences during the development and deployment of AI systems [[6](#bib.bib1 "Responsible artificial intelligence: how to develop and use... |
804f46f3-3645-4394-9f73-59fe71ae1bd4 | StampyAI/alignment-research-dataset/arxiv | Arxiv | Hierarchical Few-Shot Imitation with Skill Transition Models.
1 Introduction
---------------
We are interested in developing control algorithms that enable robots to solve complex and practical tasks such as operating kitchens or assisting humans with everyday chores at home. There are two general characteristics o... |
90dc2469-cd1c-47af-937e-4d6f57c5f691 | trentmkelly/LessWrong-43k | LessWrong | Allen & Wallach on Friendly AI
Colin Allen and Wendell Wallach, who wrote Moral Machines (MM) for OUP in 2009, address the problem of Friendly AI in their recent chapter for Robot Ethics (MIT Press). Their chapter is a precis of MM and a response to objections, one of which is:
> The work of researchers focused on en... |
6273a3d7-f5cc-43ef-9d4e-4f74550cbb21 | trentmkelly/LessWrong-43k | LessWrong | Teaching Claude to Meditate
I spent some time today having an extended conversation with Claude 3.5 Sonnet. My initial goal was roughly something like "can I teach Claude to meditate?" The answer was effectively no, but then we got into a deeper discussion of dharma, alignment, and how we might train a future LLM to b... |
df04ce6e-87e2-4c5e-89a9-88a17a0cee55 | trentmkelly/LessWrong-43k | LessWrong | [LINK] Centre for the Study of Existential Risk is now on slashdot
Your opportunity to weigh in and get some reasoned views widely heard:
The Men Trying To Save Us From the Machines |
54f9a88a-8b20-4a74-892f-015c7b04de67 | trentmkelly/LessWrong-43k | LessWrong | Winning is for Losers
This post originally appeared on Ribbonfarm. It was written as part of the Ribbonfarm long-form writing course and edited by Joseph Kelly. I owe Joseph and the Ribbonfarm editors (Venkatesh Rao and Sarah Perry) huge thanks for spending the time to make me a better writer.
-----------------------... |
d04a7d17-5f87-4755-9515-7faff83eba8a | trentmkelly/LessWrong-43k | LessWrong | Simulacra Levels and their Interactions
Previously: Covid-19: My Current Model, On Negative Feedback and Simulacra
This post aims to unpack and explain simulacra levels of action using the threat of covid-19 as its central example. My intention is for future posts to then apply this model to many covid-related dynami... |
9f025629-e888-4d0f-80aa-5e7ab7b8c035 | trentmkelly/LessWrong-43k | LessWrong | A primer on ML in antibody engineering
Introduction
I increasingly see that a fair bit of the progress in ML-based protein design methods is occurring in antibody engineering, which is the process of creating antibodies tailor-made to bind to specific things in the body. I've noticed this phenomenon for the past year,... |
2d5d5f5d-a330-4533-8719-b3fcd1235bfc | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | No surjection onto function space for manifold X
.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: normal; white-space: now... |
026f1027-0c3e-4c16-9d24-853a6766a2b2 | trentmkelly/LessWrong-43k | LessWrong | OpenAI-Microsoft partnership
OpenAI has a strong partnership with Microsoft. The details are opaque, as far as I know. It tentatively seems that OpenAI is required to share its models (and some other IP) with Microsoft until OpenAI attains "a highly autonomous system that outperforms humans at most economically valuab... |
1406bbae-5ccb-4582-a6de-e170f41ca8a3 | trentmkelly/LessWrong-43k | LessWrong | Open Thread, September 1-15, 2012
If it's worth saying, but not worth its own post, even in Discussion, it goes here. |
d1586ad1-b764-4b2f-a44c-6aa78169939e | trentmkelly/LessWrong-43k | LessWrong | LOVE in a simbox is all you need
Learning Other's Values or Empowerment in simulation sandboxes is all you need
TL;DR: We can develop self-aligning DL based AGI by improving on the brain's dynamic alignment mechanisms (empathy/altruism/love) via safe test iteration in simulation sandboxes.
AGI is on track to arrive ... |
21d23cf4-cdf3-4bfc-baf0-da81abc1dfaf | trentmkelly/LessWrong-43k | LessWrong | How can I bet on short timelines?
I currently have something like 50% chance that the point of no return will happen by 2030. Moreover, it seems to me that there's a wager for short timelines, i.e. you should act as if short timelines scenarios are more likely than they really are, because you have more influence over... |
074b5c28-3bbe-4384-bd09-29391dce43f2 | trentmkelly/LessWrong-43k | LessWrong | What do people think of the Futurism site?
I'm considering subscribing but thought I would ping a community of people I find rather trustworthy of taking advice from.
https://futurism.com
TIA |
4f1bbf68-ec24-4a1d-88a0-656b5fe3d51c | trentmkelly/LessWrong-43k | LessWrong | The Chilling Effect of Confiscation
The federal government has been seizing mask imports:
* Three million, on arrival at the Port of New York and New Jersey on 2020-03-18 (Boston Globe)
* Four hundred thousand, on arrival at JFK airport; 100k on 2020-04-06 and 300k on 2020-04-16 (USA Today, NY Daily News)
* On... |
5e7ae6a1-ea09-465f-892c-f30dff88f069 | trentmkelly/LessWrong-43k | LessWrong | Has a technological dependency graph been made?
I would like to see a graph of technologies that were developed during humanity's history that shows which ones were (likely) necessary for the development of which other. It might provide useful insights. |
e3e0519c-09a0-49ca-8924-dd457b2a47de | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Most Leading AI Experts Believe That Advanced AI Could Be Extremely Dangerous to Humanity
*This is written primarily for a non-EA audience. Posting here mostly for reference/visibility.*
In 2018, the ACM Turing Award was given to three pioneers of the deep learning revolution: Yoshua Bengio, Geoffrey Hinton, and Yann... |
f1d46f7d-7686-48af-9b04-814e98f68d2d | trentmkelly/LessWrong-43k | LessWrong | Is it sensible for an ambitious nonsmoker to use e-cigarettes?
Many of you have already seen Gwern's page on the topic of nicotine use. Nicotine is interesting because it's a stimulant, it may increase intelligence (I believe Daniel Kahneman said he was smarter back when he used to smoke), and it may be useful for hab... |
f817bab4-2c27-477a-abc7-953e7e5bc252 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Coherence arguments imply a force for goal-directed behavior
*Crossposted from [AI Impacts](https://aiimpacts.org/)*
*[Epistemic status: my current view, but I haven’t read all the stuff on this topic even in the LessWrong community, let alone more broadly.]*
There is a line of thought that says that advanced AI w... |
f48ddd98-0a5b-4a87-bd48-2180d4a0fb9e | trentmkelly/LessWrong-43k | LessWrong | [S] D&D.Sci: All the D8a. Allllllll of it.
This is an entry in the 'Dungeons & Data Science' series, a set of puzzles where players are given a dataset to analyze and an objective to pursue using information from that dataset.
STORY (very, very, very much skippable)
aphyer: Write MSPA D&D.Sci scenario.
Yes.
Nove... |
8e18c2d5-91a1-43e6-93c2-d89743052cc7 | trentmkelly/LessWrong-43k | LessWrong | Hammertime Day 7: Aversion Factoring
This is part 7 of 30 in the Hammertime Sequence. Click here for the intro.
As we move into the introspective segment of Hammertime, I want to frame our approach around the set of (unoriginal) ideas I laid out in The Solitaire Principle. The main idea was that a human being is best... |
b34d7b36-492d-4b48-9b17-75ec940c2971 | trentmkelly/LessWrong-43k | LessWrong | Why I am no longer driven
(cross-posted from my blog, Sunday Stopwatch)
I used to be very driven. I'm talking "wake up at 4 AM, go run with my dog, take a cold shower, study for two hours, go work, meditate, cook, go to MMA practice, read a book before bed" driven. I was very ready to, you know, get after it. As you ... |
58267cad-70c4-4b68-aa9f-e74238365913 | trentmkelly/LessWrong-43k | LessWrong | An explanation for every token: using an LLM to sample another LLM
Introduction
Much has been written about the implications and potential safety benefits of building an AGI based on one or more Large Language Models (LLMs). Proponents argue that because LLMs operate on human-readable natural language, they will be ea... |
077efe47-a4f5-4314-a550-84a65708c5e0 | trentmkelly/LessWrong-43k | LessWrong | Memes and Rational Decisions
In 2004, Michael Vassar gave the following talk about how humans can reduce existential risk, titled Memes and Rational Decisions, to some transhumanists. It is well-written and gives actionable advice, much of which is unfamiliar to the contemporary Less Wrong zeitgeist.
> Although trans... |
fad0e2a3-d7b0-4f5c-84b2-2feb39a7591e | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | autonomy: the missing AGI ingredient?
*Epistemic status: trying to feel out the shape of a concept and give it an appropriate name. Trying to make explicit some things that I think exist implicitly in many people's minds. This post makes truth claims, but its main goal is to not to convince you that they are true.*
... |
86135cfc-75c2-4a41-b10c-a6fc8bcfe15b | trentmkelly/LessWrong-43k | LessWrong | Hamster in Tutu Shuts Down Large Hadron Collider
The Large Hadron Collider was shut down yesterday by a hamster in a tutu, weary scientists announced.
The Large Hadron Collider is the successor to the earlier Superconducting Super Collider, which was shut down by the US House of Representatives in 1993 after 14 miles... |
e49720fb-05d7-4317-98f4-757d1f15831a | awestover/filtering-for-misalignment | Redwood Research: Alek's Filtering Results | id: post196
Blog post by Sebastian Farquhar, David Lindner, Rohin Shah. It discusses the paper MONA: Myopic Optimization with Non-myopic Approval Can Mitigate Multi-step Reward Hacking by Sebastian Farquhar, Vikrant Varma, David Lindner, David Elson, Caleb Biddulph, Ian Goodfellow, and Rohin Shah. Our paper tries to m... |
1f774a8c-bdaa-4cf2-b6e9-19e143e638eb | trentmkelly/LessWrong-43k | LessWrong | Grue, Bleen, and natural categories
A putative new idea for AI control; index here.
In a previous post, I looked at unnatural concepts such as grue (green if X was true, blue if it was false) and bleen. This was to enable one to construct the natural categories that extend AI behaviour, something that seemed surprisi... |
e5b6b1be-9576-43d1-ac51-0c8b619e591e | trentmkelly/LessWrong-43k | LessWrong | Sequential Organization of Thinking: "Six Thinking Hats"
Many people move chaotically from thought to thought without explicit structure. Inappropriate structuring may leave blind spots or cause the gears of thought to grind to a halt, but the advantages of appropriate structuring are immense:
Correct thought struct... |
b5662536-100d-4bc1-8472-12d192850bcd | trentmkelly/LessWrong-43k | LessWrong | Does anyone else sometimes "run out of gas" when trying to think?
Sometimes when I'm trying to think something through (like a social dilemma, or a chess problem, or a program I'm writing, etc) I have a pretty unpleasant experience that feels to me the way I imagine a car feels when it runs out of gas in the middle of... |
a622a8c2-93b6-41f5-b1ba-1051b31763f6 | StampyAI/alignment-research-dataset/lesswrong | 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 technolo... |
06c82754-4843-4185-a4dd-f766635d04fb | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Deepmind Plans for Rat-Level AI
Demis Hassabis gives a [great presentation](https://youtu.be/vQXAsdMa_8A) on the state of Deepmind's work as of April 20, 2016. Skip to 23:12 for the statement of the goal of creating a rat-level AI -- "An AI that can do everything a rat can do," in his words. From his tone, it sounds ... |
b7f0d0ad-b13e-4df3-b861-24ea38806ef8 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Sydney Rationality meet-up No.2
Discussion article for the meetup : Sydney Rationality meet-up No.2
WHEN: 15 February 2012 06:00:00PM (+1100)
WHERE: 22 The Promenade, Sydney NSW 2000
As usual at the king street brew house until someone comes up with something better.
Well, if people want regular, they'll ... |
2937ada9-f75b-4b4b-a295-478d6d1e7806 | trentmkelly/LessWrong-43k | LessWrong | "Sex Is Always Well Worth Its Two-Fold Cost"
This might be of interest to the evo bio and game theory wannabes here: "Sex Is Always Well Worth Its Two-Fold Cost" by Alexander Feigel, Avraham Englander and Assaf Engel.
Abstract:
> Sex is considered as an evolutionary paradox, since its positive contribution to Darwin... |
c2b06d9f-d1d9-4b2e-a0b7-73f8f452424c | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Protest against Meta's irreversible proliferation (Sept 29, San Francisco)
Meta’s frontier AI models are fundamentally unsafe. Since Meta AI has released the model weights publicly, any safety measures can be removed. Before it releases even more advanced models – which will have more dangerous capabilities – we call ... |
d4ddd30f-663f-480f-bad4-834d834db625 | trentmkelly/LessWrong-43k | LessWrong | Goal-directed = Model-based RL?
Epistemic Status: quick write-up, in reaction to a serendipitous encounter with an idea. I see the main value of this post as decently presenting a potentially interesting take on a concept in AI safety to the community.
While skimming my copy of Reinforcement Learning: an introduction... |
e376901f-0b22-45c8-b3c9-71780d562f83 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Common mistakes people make when thinking about decision theory
From my experience reading and talking about decision theory on LW, it seems that many of the unproductive comments in these discussions can be attributed to a handful of common mistakes.
#### Mistake #1: Arguing about assumptions
The main reason why ... |
e54a870d-2620-45d5-bc00-ec14186b257b | trentmkelly/LessWrong-43k | LessWrong | My vision of a good future, part I
A future where everyone is okay
I grew up with the future feeling pretty uncertain. God was supposed to return at some point, but maybe there were going to be The End Times before that? Like, a great persecution of the Christians, and maybe something about the bottomless pit and the ... |
8944d4da-281d-4f1f-a032-a05de19189af | trentmkelly/LessWrong-43k | LessWrong | How to generate idea/solutions to solve a problem?
Are any rationality techniques one could leverage to generate ideas or solutions towards solving a specific problem?
This is probably the biggest obstacle for me to have effective goal factoring. I sometimes cannot come up with good ideas or alternatives.
I found t... |
7315c5a1-ada5-46ed-ba07-a9ede9d4bab5 | trentmkelly/LessWrong-43k | LessWrong | Teacher's Password: The LessWrong Mystery Hunt Team
Followup to: https://www.lesswrong.com/posts/iNTabF6vQjeDeSar7/interest-survey-forming-an-mit-mystery-hunt-team-jan-15-18
We are making a LessWrong team for the 2021 MIT Mystery Hunt. If you want to play the Mystery Hunt with other rationalists, you can sign up by ... |
abc46b2b-dcd7-44e1-9fa1-f1175fb72a46 | trentmkelly/LessWrong-43k | LessWrong | Arbital scrape
Update: Arbital Scrape V2
I've scraped http://arbital.com as the site is unusably slow and hard to search for me.
The scrape is locally browsable and plain HTML save for MathJax.
https://drive.google.com/open?id=1b7dKhOzfMpFwngAeI8efeOzv147Lv_mx
https://emma-borhanian.github.io/arbital-scrape/
If t... |
067eed1c-ef1a-469f-996a-46817f9b56f7 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Regular Moscow meetup: effective altruism, debates, hypothesis formulation
Discussion article for the meetup : Regular Moscow meetup: effective altruism, debates, hypothesis formulation
WHEN: 02 August 2015 02:00:00PM (+0300)
WHERE: Москва, Льва Толстого, 16
We're meeting at Yandex, at the Extropolis confe... |
b86961e7-4eec-47ac-bae5-3921521c3fec | trentmkelly/LessWrong-43k | LessWrong | It's a fact: male and female brains are different
In Which I Present The Opposing Side's Hypothesis and Falsify It
This post is in part in response to a New Scientist article/book review "Fighting back against neurosexism." And the tagline is "Are differences between men and women hard-wired in the brain? Two new bo... |
d0734770-5388-4af3-b9f4-cbb4c75700af | trentmkelly/LessWrong-43k | LessWrong | Why you don’t seek friends dating site style
Robin asks an interesting question:
> Bryan Caplan recently pointed out to a few of us that while many dating web sites offer to help you find matching romantic mates, there are far fewer friend finding helpers. We tend to collect friends informally, by liking the people ... |
72eefb72-b99c-46b3-8d8c-2e301d509beb | trentmkelly/LessWrong-43k | LessWrong | Help us Optimize the Contents of the Sequences eBook
MIRI's ongoing effort to publish the sequences as an eBook has given us the opportunity to update their contents and organization.
We're looking for suggested posts to reorder, add, or remove.
To help with this, here is a breakdown of the current planned contents ... |
13563da0-00e3-4032-a759-a64d42935f88 | trentmkelly/LessWrong-43k | LessWrong | Against the Linear Utility Hypothesis and the Leverage Penalty
[Roughly the second half of this is a reply to: Pascal's Muggle]
There's an assumption that people often make when thinking about decision theory, which is that utility should be linear with respect to amount of stuff going on. To be clear, I don't mean l... |
3fd53821-5f89-4fe3-b758-fffe2155ea14 | trentmkelly/LessWrong-43k | LessWrong | Determining the power of investors over Frontier AI Labs is strategically important to reduce x-risk
Produced as part of the ML Alignment & Theory Scholars Program - Summer 2024 Cohort
This post presents the basic story behind my current research direction at MATS. I'm looking for feedback on whether the basic premis... |
5b7a86d9-08a7-4056-a054-64b01f0252db | trentmkelly/LessWrong-43k | LessWrong | The Three Warnings of the Zentradi
This is mostly some ramblings and background notes for a fanfiction, and should not be taken seriously as a real-world argument, except insofar as I would hope it could become good enough to be a real-world argument if I were smart enough and worked on it enough and got the right fee... |
7cf1aa23-3081-4055-99e6-19bb0e805262 | StampyAI/alignment-research-dataset/youtube | Youtube Transcripts | 202. Gwern on GPT-3
hello and welcome to session 202
in the aisafety.com reading group
tonight we will be discussing an article
called
on gbt3 meter learning scaling
implications
and deep theory by gwen
warren branwyn is a freelance
writer and researcher from america he is
a very very prolific writer who has
written
a... |
3e600e2d-01e5-48ee-a8f3-fd628aded024 | trentmkelly/LessWrong-43k | LessWrong | One hub, or many?
I was reading the comments to The Rationalist Community's Location Problem and especially this comment by Damien Tatum, and thinking about some classic questions of LessWrong/EA/fully-general physical community.
I think a guiding idea for years has been that if we get as many good people and organiz... |
76b9c998-1bac-4dcb-be61-7db6adc61cd0 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Paper Replication Walkthrough: Reverse-Engineering Modular Addition
I'm excited about trying different formats for mechanistic interpretability education! I've [made a video walkthrough](https://neelnanda.io/modular-addition-walkthrough) where we replicate my paper, [Progress Measures for Grokking via Mechanistic Inte... |
53a4e9e6-ae77-40e4-88c6-94c6fee8d634 | trentmkelly/LessWrong-43k | LessWrong | The world is awful. The world is much better. The world can be much better: The Animation.
This video is an adaptation of Max Roser's article, "The world is awful. The world is much better. The world can be much better." It's a simple yet important point that society at large underappreciates. Yet, it's a crucial asp... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.