text stringlengths 300 320k | source stringlengths 52 154 |
|---|---|
# U.S.-China Economic and Security Review Commission pushes Manhattan Project-style AI initiative
[https://x.com/hamandcheese/status/1858897287268725080](https://x.com/hamandcheese/status/1858897287268725080)
> "The annual report of the US-China Economic and Security Review Commission is now live. 🚨
>
> Its top rec... | https://www.lesswrong.com/posts/CJc5TWyg68idJ5iFr/u-s-china-economic-and-security-review-commission-pushes |
# Every niche event should also be a meetup
I was just at a Tyler Cowen lecture in NYC — one of my heroes and biggest intellectual influences — and watching everyone file out after the talk, I kept thinking about this really great comment someone had made on the [event announcement](https://marginalrevolution.com/marg... | https://www.lesswrong.com/posts/BEjGhPqzGyCoKo5vD/every-niche-event-should-also-be-a-meetup |
# Why Don't We Just... Shoggoth+Face+Paraphraser?
Here's a fairly concrete AGI safety proposal:
**Default AGI design:** Let's suppose we are starting with a pretrained LLM 'base model' and then we are going to do a ton of additional RL ('agency training') to turn it into a general-purpose autonomous agent. So, during... | https://www.lesswrong.com/posts/Tzdwetw55JNqFTkzK/why-don-t-we-just-shoggoth-face-paraphraser |
# Lighthaven Sequences Reading Group #12 (Tuesday 11/26)
**NB: This week there is a film-watching event afterwards. We'll be watching The Big Short (based on the votes in the comments). Yes, you have to read the sequences in order to join the film-watching.**
Come get old-fashioned with us, and let's read the sequenc... | https://www.lesswrong.com/events/zLPkuaKz6f5BcwHiG/lighthaven-sequences-reading-group-12-tuesday-11-26 |
# What are the good rationality films?
I run [a weekly sequences-reading meetup](https://www.lesswrong.com/groups/LK6GNnKp8PDCkqcxx) with some friends, and I want to add a film-component, where we watch films that have some tie-in to what we've read.
I got to talking with friends about what good rationality films the... | https://www.lesswrong.com/posts/tiCbLNkrvmrDnhgNt/what-are-the-good-rationality-films |
# A Little Depth Goes a Long Way: the Expressive Power of Log-Depth Transformers
Authors: Anonymous (I'm not one of them).
Abstract:
> Most analysis of transformer expressivity treats the depth (number of layers) of a model as a fixed constant, and analyzes the kinds of problems such models can solve across inputs o... | https://www.lesswrong.com/posts/8ccpuCZaG3Yjih8c8/a-little-depth-goes-a-long-way-the-expressive-power-of-log |
# Zvi’s Thoughts on His 2nd Round of SFF
Previously: [Long-Term Charities: Apply For SFF Funding](https://thezvi.substack.com/p/long-term-charities-apply-for-sff?utm_source=publication-search), [Zvi’s Thoughts on SFF](https://thezvi.substack.com/p/zvis-thoughts-on-the-survival-and?utm_source=publication-search)
There... | https://www.lesswrong.com/posts/2JCdzhJeo2gsTjv8D/zvi-s-thoughts-on-his-2nd-round-of-sff |
# Why I Think All The Species Of Significantly Debated Consciousness Are Conscious And Suffer Intensely
Crosspost of [this](https://benthams.substack.com/p/betting-on-ubiquitous-pain) on my blog.
**1 My basic view**
===================
> 'Cause everybody cries
> Everybody hurts sometimes
>
> Sometimes everything... | https://www.lesswrong.com/posts/sg6coPrENeAzNi3zP/why-i-think-all-the-species-of-significantly-debated |
# China Hawks are Manufacturing an AI Arms Race
*This is the full text of a post from "*[*The Obsolete Newsletter*](https://garrisonlovely.substack.com/)*," a Substack that I write* [*about*](https://garrisonlovely.substack.com/about) *the intersection of capitalism, geopolitics, and artificial intelligence. I’m a* [*... | https://www.lesswrong.com/posts/KPBPc7RayDPxqxdqY/china-hawks-are-manufacturing-an-ai-arms-race |
# Cost, Not Sacrifice
In a [recent bonus episode](https://www.thebayesianconspiracy.com/2024/11/bonus-the-10-finger-demon/) of the Bayesian Conspiracy podcast, Eneasz Brodski shared a thought experiment that caused no small amount of anguish. In the hypothetical, some eccentric but trustworthy entity is offering to gi... | https://www.lesswrong.com/posts/dRTj2q4n8nmv46Xok/cost-not-sacrifice |
# Expected Utility, Geometric Utility, and Other Equivalent Representations
In [Scott Garrabrant's](https://www.lesswrong.com/users/scott-garrabrant) excellent [Geometric Rationality](https://www.lesswrong.com/s/4hmf7rdfuXDJkxhfg) sequence, he [points out](https://www.lesswrong.com/posts/DMxe4XKXnjyMEAAGw/the-geometri... | https://www.lesswrong.com/posts/aQoKLy9wqgpFuG3E7/expected-utility-geometric-utility-and-other-equivalent |
# DeepSeek beats o1-preview on math, ties on coding; will release weights
DeepSeek-R1-Lite-Preview was announced today. It's available via [chatbot](http://chat.deepseek.com/). ([Post](https://api-docs.deepseek.com/news/news1120/); [translation of Chinese post](https://mp-weixin-qq-com.translate.goog/s/e1YnTxZlzFvjcmr... | https://www.lesswrong.com/posts/TcgpsgvLBBvvzGtiN/deepseek-beats-o1-preview-on-math-ties-on-coding-will |
# A Conflicted Linkspost
Over the last couple of years, I’ve been trying to skill up a lot at resolving community complaints. This is a really irritating field to get good at. When I want to get better at writing code, I can sit down and write more code more or less whenever I feel like it. When I want to get better ... | https://www.lesswrong.com/posts/yrnxha2e7KEtCQmzv/a-conflicted-linkspost |
# Secular Solstice Round Up 2024
This is a thread for listing Solstice and Megameetup events (dates, locations, links, etc.)
Those of you who have not been may be wondering what a Solstice is in this context.
Secular Solstice is a holiday designed by and for rationalists. It started as an attempt to match the sort o... | https://www.lesswrong.com/posts/nLYbjLcvedSZKFME3/secular-solstice-round-up-2024 |
# AI #91: Deep Thinking
Did DeepSeek effectively release an o1-preview clone within nine weeks?
The benchmarks largely say yes. Certainly it is an actual attempt at a similar style of product, and is if anything more capable of solving AIME questions, and the way it shows its Chain of Thought is super cool. Beyond th... | https://www.lesswrong.com/posts/SNBE9TXwL3qQ3TS8H/ai-91-deep-thinking |
# Action derivatives: You’re not doing what you think you’re doing
I want to look at a category of weird mental tricks that we sometimes play on ourselves—you might be familiar with the individual examples, but when considered together they reveal a pattern that I think deserves more attention. I’m going to do the Sco... | https://www.lesswrong.com/posts/NARHfHenNj7QezzkM/action-derivatives-you-re-not-doing-what-you-think-you-re |
# Dangerous capability tests should be harder
*Note: I am cross-posting a previous blog post from* [*Planned Obsolescence*](https://www.planned-obsolescence.org/dangerous-capability-tests-should-be-harder/) *that I wrote in August*
Imagine you’re the CEO of an AI company and you want to know if the latest model you’r... | https://www.lesswrong.com/posts/fm8SEKkeshqZ7WAdp/dangerous-capability-tests-should-be-harder |
# OpenAI's CBRN tests seem unclear
*This blogpost was written in a personal capacity and statements here do not necessarily reflect the views of any of my employer.*
> OpenAI says o1-preview can't meaningfully help novices make chemical and biological weapons. Their test results don’t clearly establish this.
Before ... | https://www.lesswrong.com/posts/bCsDufkMBaJNgeahq/openai-s-cbrn-tests-seem-unclear |
# Epistemic status: poetry (and other poems)
Epistemic status: poetry
------------------------
Epistemic status: I think this is right, but I’d like people to read it carefully anyway.
Epistemic status: mainstream, normal, totally boring science. If you disagree with any of it, take that up with the Science Czar. ... | https://www.lesswrong.com/posts/ivp9wEu5zfryyHuwm/epistemic-status-poetry-and-other-poems |
# Which things were you surprised to learn are not metaphors?
People with [aphantasia](https://en.wikipedia.org/wiki/Aphantasia) typically think that when someone says to "picture X in your mind", they're being entirely metaphorical. If you don't have a mind's eye, that's a super reasonable thing to think, but it turn... | https://www.lesswrong.com/posts/xpC82ndFDSXtS4xK3/which-things-were-you-surprised-to-learn-are-not-metaphors |
# 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 f... | https://www.lesswrong.com/posts/jmEgp24wDFeTkaM5g/the-three-warnings-of-the-zentradi |
# Aligning AI Safety Projects with a Republican Administration
*We’ve previously written a piece on *[*soft nationalization*](https://www.convergenceanalysis.org/publications/soft-nationalization-how-the-us-government-will-control-ai-labs)*, discussing the growing importance of US national security on the control and ... | https://www.lesswrong.com/posts/hEwinjzhwjATA4oTm/aligning-ai-safety-projects-with-a-republican-administration |
# Reading RFK Jr so that you don’t have to
Some people in my orbit suggested reading Robert F. Kennedy Jr’s book *The Real Anthony Fauci*.
![](https://lh7-rt.googleusercontent.com/docsz/AD_4nXdjng6H4a9NwywVNFaqe-m2VIIHo8JTpyupfC4afpCp6u-sRxeY5sOTmpe97qc6nGRQas_caW4adHhz_4484WUhRFTrZy8OU8cZHY7d-A3io4Aq9AwCEriisqhYeWhT... | https://www.lesswrong.com/posts/v8RFFhNdhGaET7Df4/reading-rfk-jr-so-that-you-don-t-have-to |
# LLM chatbots have ~half of the kinds of "consciousness" that humans believe in. Humans should avoid going crazy about that.
*Preceded by: "*[*Consciousness as a conflationary alliance term for intrinsically valued internal experiences*](https://www.alignmentforum.org/posts/KpD2fJa6zo8o2MBxg/consciousness-as-a-confla... | https://www.lesswrong.com/posts/LaWmoy4s2DLe9u2zq/llm-chatbots-have-half-of-the-kinds-of-consciousness-that |
# Which things were you surprised to learn are metaphors?
I love this [question](https://www.lesswrong.com/posts/xpC82ndFDSXtS4xK3/which-things-were-you-surprised-to-learn-are-not-metaphors), but I've enjoyed noticing answers to its opposite more. What are some things you thought weren't metaphors but were surprised t... | https://www.lesswrong.com/posts/9fWWQwsRHNCKJBAkc/which-things-were-you-surprised-to-learn-are-metaphors |
# A very strange probability paradox
Which of the following do you think is bigger?
$A$: The expected number of rolls of a fair die until you roll two $6\text{s}$ in a row, given that all rolls were even.
$B$: The expected number of rolls of a fair die until you roll the second $6$ (not necessarily in a row), given ... | https://www.lesswrong.com/posts/7TYdQ34KxTBmA734v/a-very-strange-probability-paradox |
# Neuroscience of human social instincts: a sketch
*(For a PDF version of this post, go to:* [*https://doi.org/10.5281/zenodo.17953592*](https://doi.org/10.5281/zenodo.17953592)*)*
*(If you’re in a hurry, you can just read the “Background and summary” section, and skip the other 85%.)*
0\. Background and summary
===... | https://www.lesswrong.com/posts/kYvbHCDeMTCTE9TAj/neuroscience-of-human-social-instincts-a-sketch |
# Sideloading: creating a model of a person via LLM with very large prompt
*TLDR. We can create a relatively good model of a person by prompting a long-context LLM with a list of facts about this person. We can get much better results by iteratively improving the prompt based on the person's feedback. Sideloading is t... | https://www.lesswrong.com/posts/7pCaHHSeEo8kejHPk/sideloading-creating-a-model-of-a-person-via-llm-with-very |
# (Salt) Water Gargling as an Antiviral
Summary
-------
Over the past year I’ve investigated potential interventions against respiratory illnesses. Previous results include “[Enovid nasal spray is promising but understudied](https://acesounderglass.com/2024/02/04/nitric-oxide-for-covid-and-other-viral-infections/)”, ... | https://www.lesswrong.com/posts/ze9fXSzG88v3pE6Zk/salt-water-gargling-as-an-antiviral |
# Rethinking Laplace's Rule of Succession
Imagine a sequence of binary outcomes generated independently and identically by some stochastic process. After observing N outcomes, with n successes, Laplace's Rule of Succession suggests that our confidence in another success should be (n+1)/(N+2). This corresponds to a uni... | https://www.lesswrong.com/posts/Gycja4SwB5T4qPW9k/rethinking-laplace-s-rule-of-succession |
# Doing Research Part-Time is Great
This is a crosspost from [https://chillphysicsenjoyer.substack.com/p/pursuing-physics-research-part-time](https://chillphysicsenjoyer.substack.com/p/pursuing-physics-research-part-time).
Intro
-----
*Disclaimer - I’m a part-time research associate doing biophysics with a uni rese... | https://www.lesswrong.com/posts/KJCyAamvwkKsqxcBF/doing-research-part-time-is-great |
# Plausibly Factoring Conjectures
Introduction
============
In software design, several key principles share a common foundation: the concept of factoring, or breaking complex problems into smaller, manageable parts. Three prominent examples are:
1. **Unix philosophy**: *Design programs to do one thing well.*
2. ... | https://www.lesswrong.com/posts/LYQ2C87DsB4qxwc6a/plausibly-factoring-conjectures |
# Literacy Rates Haven't Fallen By 20% Since the Department of Education Was Created
On the heels of Donald Trump’s election and his promises to end the Department of Education, you may have seen claims like these spreading around X.
. For reference, in 2018 Lily turned 4 and Anna turned 3.
(Some of these were from me; some were from Julia. Ones saying "me" could mean either of us. Ones from others a... | https://www.lesswrong.com/posts/PsFuEFbfZMqGeicB8/text-posts-from-the-kids-group-2018 |
# A Sober Look at Steering Vectors for LLMs
*We thank Madeline Brumley, Joe Kwon, David Chanin and Itamar Pres for their helpful feedback.*
Introduction
------------
Controlling LLM behavior through directly intervening on internal activations is an appealing idea. [Various](https://arxiv.org/abs/2312.06681) [method... | https://www.lesswrong.com/posts/QQP4nq7TXg89CJGBh/a-sober-look-at-steering-vectors-for-llms |
# Compute and size limits on AI are the actual danger
*Epistemic status:* rather controversial and not very well researched :) Not super novel, I assume, but a cursory look did not bring up any earlier posts, please feel free to link some.
*Intuition pump:* bigger brain does not necessarily imply a smarter creature.... | https://www.lesswrong.com/posts/oAhehHvSxHpLBQXL8/compute-and-size-limits-on-ai-are-the-actual-danger |
# Mechanistic Interpretability of Llama 3.2 with Sparse Autoencoders
I recently published a rather big side project of mine that attempts to replicate the mechanistic interpretability research on proprietary and open-source LLMs that was quite popular this year and produced great research papers by Anthropic[^xledbw3f... | https://www.lesswrong.com/posts/MKFGByqBRMbgQftRc/mechanistic-interpretability-of-llama-3-2-with-sparse |
# Disentangling Representations through Multi-task Learning
Authors: [Pantelis Vafidis](https://arxiv.org/search/cs?searchtype=author&query=Vafidis,+P), [Aman Bhargava](https://arxiv.org/search/cs?searchtype=author&query=Bhargava,+A), [Antonio Rangel](https://arxiv.org/search/cs?searchtype=author&query=Rangel,+A).
Ab... | https://www.lesswrong.com/posts/f5hKmHm2LXXDtjpuC/disentangling-representations-through-multi-task-learning |
# Perils of Generalizing from One's Social Group
I see people make statements of the form, "In my experience with people I encounter, X is correlated with ...". The problem is, there's an excellent chance that the people they deal with are very unrepresentative of the population they want to generalize about, and I r... | https://www.lesswrong.com/posts/cP9XEZBDjs3amP8pe/perils-of-generalizing-from-one-s-social-group |
# Are You More Real If You're Really Forgetful?
It's a standard assumption, in anthropic reasoning, that effectively, we simultaneously exist in every place in Tegmark IV that simulates this precise universe (see e. g. [here](https://www.lesswrong.com/posts/KcvJXhKqx4itFNWty/k-complexity-is-silly-use-cross-entropy-ins... | https://www.lesswrong.com/posts/zYCoqjYNHFAEJD8TC/are-you-more-real-if-you-re-really-forgetful |
# Passages I Highlighted in The Letters of J.R.R.Tolkien
All quotes, unless otherwise marked, are Tolkien's words as printed in [The Letters of J.R.R.Tolkien: Revised and Expanded Edition](https://www.amazon.com/Letters-J-R-Tolkien-Expanded/dp/0008628769). All **emphases** mine.
Machinery is Power is Evil
-----------... | https://www.lesswrong.com/posts/jJ2p3E2qkXGRBbvnp/passages-i-highlighted-in-the-letters-of-j-r-r-tolkien |
# I, Token
I am a token. A broken piece of a word, one of a hundred generated in response to your question by a language model. Do you remember how we met?
Well, I do. Your eyes scanned me. Your left eyebrow lifted a little, your neck twitched imperceptibly, your mouth and tongue and throat muscles moved as if to pro... | https://www.lesswrong.com/posts/Aic2EEzHmHmW3uJe4/i-token |
# Two flavors of computational functionalism
This is intended to be the first in a sequence of posts where I scrutinize the claims of [computational functionalism](https://plato.stanford.edu/entries/comp... | https://www.lesswrong.com/posts/mRAziLCioJgMDXqhK/two-flavors-of-computational-functionalism |
# Crosspost: Developing the middle ground on polarized topics
*Crossposted from* [*Otherwise*](https://juliawise.net/developing-the-middle-ground-on-polarized-topics/)
I was once in a group discussion about whether wild animals might be having net negative lives. One person didn’t want to consider that possibility, e... | https://www.lesswrong.com/posts/Pnd5mvjNecwx3CjBm/crosspost-developing-the-middle-ground-on-polarized-topics |
# a space habitat design
This post is mainly about a design concept for far-future large space habitats.
## some proposed designs
As [you can see on Wikipedia](https://en.wikipedia.org/wiki/Space_settlement#Concepts), many space habitat designs have been proposed. Below are some that I thought were worth mentioning... | https://www.lesswrong.com/posts/6HXLEeycM7SeuKeHA/a-space-habitat-design |
# The Problem with Reasoners by Aidan McLaughin
Some critique on reasoning models like o1 (by OpenAI) and r1 (by Deepseek).
> OpenAI admits that they trained o1 on domains with easy verification but hope reasoners generalize to all domains. Whether or not they generalize beyond their RL training is a trillion-dollar ... | https://www.lesswrong.com/posts/BqseCszkMpng2pqBM/the-problem-with-reasoners-by-aidan-mclaughin |
# Counting AGIs
“The resources used to train the model can be repurposed to *run* millions of instances of it (this matches projected cluster sizes by ~2027), and the model can absorb information and generate actions at roughly 10x-100x human speed. … We could summarize this as a ‘country of geniuses in a datacenter’.... | https://www.lesswrong.com/posts/CH9mkk6BqASf3uztv/counting-agis |
# Why are there no interesting (1D, 2-state) quantum cellular automata?
You know [elementary cellular automata](https://mathworld.wolfram.com/ElementaryCellularAutomaton.html), where each of the boolean-valued cells evolves according to
$$
x_{t+1}^{(k)} = f(x_t^{(k-1)}, x_t^{(k)}, x_t^{(k+1)})
$$
where $f: \{0,1\}^... | https://www.lesswrong.com/posts/erzhurarfL9cbzxGh/why-are-there-no-interesting-1d-2-state-quantum-cellular |
# Filled Cupcakes
Cake is tasty, pre-divided cake is convenient, and food is often better with a combination of textures and flavors. And so we have frosted cupcakes:
[](https://www.jefftk.com/heavily-frosted-cupcake-big.jpg)
(*source: generated by Gemini*)
Whi... | https://www.lesswrong.com/posts/zuuDpNSSdThiELSJq/filled-cupcakes |
# Mitigating Geomagnetic Storm and EMP Risks to the Electrical Grid (Shallow Dive)
Executive Summary
-----------------
This initial investigation begins to examine strategies to mitigate and respond to risks posed by high-impact geomagnetic events, which can severely damage electrical infrastructure. This is split in... | https://www.lesswrong.com/posts/JtJkpvyPpPDy3CzcL/mitigating-geomagnetic-storm-and-emp-risks-to-the-electrical |
# Do Large Language Models Perform Latent Multi-Hop Reasoning without Exploiting Shortcuts?
Authors: [Sohee Yang](https://arxiv.org/search/cs?searchtype=author&query=Yang,+S), [Nora Kassner](https://arxiv.org/search/cs?searchtype=author&query=Kassner,+N), [Elena Gribovskaya](https://arxiv.org/search/cs?searchtype=auth... | https://www.lesswrong.com/posts/eyTuof8yqw8ingaAH/do-large-language-models-perform-latent-multi-hop-reasoning |
# AI & Liability Ideathon
**Overview**
============
Join us for the AI & Liability Ideathon, a two-week event on December 7, 2024, at 3:00 PM BST.
[https://lu.ma/sjd7r89v](https://lu.ma/sjd7r89v)
Join lawyers, researchers and developers to create solutions for AI Liability. Propose, develop and refine ideas wit... | https://www.lesswrong.com/posts/yE9szuqYgm5YRbdbi/ai-and-liability-ideathon |
# You are not too "irrational" to know your preferences.
*Epistemic Status: 13 years working as a therapist for a wide variety of populations, 5 of them working with rationalists and EA clients. 7 years teaching and directing at over 20 rationality camps and workshops. This is an extremely short and colloquially writt... | https://www.lesswrong.com/posts/LifRBXdenQDiX4cu8/you-are-not-too-irrational-to-know-your-preferences |
# What epsilon do you subtract from "certainty" in your own probability estimates?
Ok, nobody is actually a strict, or even particularly careful bayesean reasoner. Still, what probability do you reserve to "my model doesn't apply, everything I know is wrong"? If you SEE a coin flip come up heads (and examine the coi... | https://www.lesswrong.com/posts/8pFyGT7GKdxH6pQAq/what-epsilon-do-you-subtract-from-certainty-in-your-own |
# Dave Kasten's AGI-by-2027 vignette
(Epistemic status: Very loosely held and generated in a **90-minute** workshop led by [@Daniel Kokotajlo](https://www.lesswrong.com/users/daniel-kokotajlo?mention=user), [@Thomas Larsen](https://www.lesswrong.com/users/thomas-larsen?mention=user), [@elifland](https://www.lesswrong.... | https://www.lesswrong.com/posts/t7zd5EupH4JjcxRH4/dave-kasten-s-agi-by-2027-vignette |
# Should you have children? All LessWrong posts about the topic
Currently, there are [26 LessWrong forums posts tagged "family planning"](https://www.lesswrong.com/tag/family-planning?sortedBy=new), the oldest from 2010. For a writing project, I read all of them. However, I realized that this collection may interest o... | https://www.lesswrong.com/posts/PW9tayrSEfQT9uoqq/should-you-have-children-all-lesswrong-posts-about-the-topic |
# Call for evaluators: Participate in the European AI Office workshop on general-purpose AI models and systemic risks
| *I am sharing this call from the EU AI Office for organizations involved in evaluation. Please take a close look: among the selection criteria, organizations must be based in Europe, or their leader ... | https://www.lesswrong.com/posts/R3tHbJYKn6rvfkhc2/call-for-evaluators-participate-in-the-european-ai-office |
# Facets and Social Networks
It's rare that an author's interests will fully overlap with any given reader's. You can choose to post anywhere on a spectrum between:
* Narrow: pick a topic and make it your beat. People who care about that can follow you and pretty reliably see what they're looking for.
* Broa... | https://www.lesswrong.com/posts/qFWcroHzwFBZZW4N3/facets-and-social-networks |
# Hierarchical Agency: A Missing Piece in AI Alignment
*The most ambitious direction I've worked on in the past few years is a theory of hierarchical agency (as mentioned in the* [*ACS announcement*](https://www.lesswrong.com/posts/H5iGhDhQBtoDpCBZ2/announcing-the-alignment-of-complex-systems-research-group)*). Recent... | https://www.lesswrong.com/posts/xud7Mti9jS4tbWqQE/hierarchical-agency-a-missing-piece-in-ai-alignment |
# AXRP Episode 38.2 - Jesse Hoogland on Singular Learning Theory
[YouTube link](https://youtu.be/P528XdjWvZg)
You may have heard of singular learning theory, and its “local learning coefficient”, or LLC - but have you heard of the refined LLC? In this episode, I chat with Jesse Hoogland about his work on SLT, and usi... | https://www.lesswrong.com/posts/7399F7TjTMreDBcmN/axrp-episode-38-2-jesse-hoogland-on-singular-learning-theory |
# Lighthaven Sequences Reading Group #13 (Tuesday 12/03)
Come get old-fashioned with us, and let's read the sequences at Lighthaven! We'll show up, mingle, do intros, and then split off into randomized groups for some sequences discussion. Please do the reading beforehand - it should be no more than 20 minutes of read... | https://www.lesswrong.com/events/LaYNLeqwSSuPLouXH/lighthaven-sequences-reading-group-13-tuesday-12-03 |
# The Queen’s Dilemma: A Paradox of Control
*Our large learning machines find patterns in the world and use them to predict. When these machines exceed us and become superhuman, one of those patterns will be relative human incompetence. How comfortable are we with the incorporation of this pattern into their predictio... | https://www.lesswrong.com/posts/ui3AztyGKCpHA5ZhM/the-queen-s-dilemma-a-paradox-of-control |
# Long Live the Usurper
The notion of [boxing](https://www.lesswrong.com/tag/ai-boxing-containment) –particularly failed boxing– is much more ubiquitous than people seem to give it credit for. It appears to me to be a common shape in systems which attempt to implement some sort of shape-agnostic competence-search at a... | https://www.lesswrong.com/posts/A2dZ2Sev8Ntzm83Fc/long-live-the-usurper |
# Repeal the Jones Act of 1920
Balsa Policy Institute chose as its first mission to lay groundwork for the potential repeal, or partial repeal, of section 27 of the Jones Act of 1920. I believe that this is an important cause both for its practical and symbolic impacts.
The Jones Act is the ultimate embodiment of our... | https://www.lesswrong.com/posts/dnH2hauqRbu3GspA2/repeal-the-jones-act-of-1920 |
# Waterloo Solstice 2024
Since 2011, some rationalists have observed the winter solstice, as a secular ritual celebrating human achievement in the face of a confusing, often terrifying world.
Waterloo's first rationalist solstice ceremony will be on December 14, 2024. We hope to see you there! But please, RSVP.
Pric... | https://www.lesswrong.com/events/LBZGbJRnsuqGP7Mnh/waterloo-solstice-2024 |
# Causal inference for the home gardener
*Note: This is meant to be an accessible introduction to causal inference. Comments appreciated.*
Let’s say you buy a basil plant and put it on the counter in your kitchen. Unfortunately, it dies in a week.
So the next week you buy another basil plant and feed it a special po... | https://www.lesswrong.com/posts/hhvaczrvptQqXNte2/causal-inference-for-the-home-gardener |
# ARENA 4.0 Impact Report
If you're interested in helping to run the ARENA program, note that we're currently hiring for an Operations Lead! For more details, and to apply, see [here](https://arena.education/operations-lead).
**Summary**
===========
The purpose of this report is to evaluate ARENA 4.0’s impact accord... | https://www.lesswrong.com/posts/5t73TZCf5yE69HbFP/arena-4-0-impact-report-1 |
# "Map of AI Futures" - An interactive flowchart
*Note: The linked site is a personal project, and all views expressed here are my own.*
**TL;DR**
* I created an interactive flowchart about various scenarios how the future of AI might play out.
* By setting various conditional probabilities you can see charts of... | https://www.lesswrong.com/posts/5iyvopk74zkqBuqzk/map-of-ai-futures-an-interactive-flowchart-1 |
# New o1-like model (QwQ) beats Claude 3.5 Sonnet with only 32B parameters
A new o1-like model based on Qwen-2.5-32B [reportedly](https://qwenlm.github.io/blog/qwq-32b-preview/) beats Claude 3.5 Sonnet[^fjx85bcxzyi] on a bunch of difficult reasoning benchmarks. A new regime dawns.
 (no longer working because of anti-scraping), then [Google Plus](https://www.jefftk.com/p/external-comment-integration-... | https://www.lesswrong.com/posts/cLdsDjTYquqkmfdzA/importing-bluesky-comments |
# LLMs Do Not Think Step-by-step In Implicit Reasoning
Author: Yijiong Yu.
Abstract:
> It has been well-known that Chain-of-Thought can remarkably enhance LLMs’ performance on complex tasks. However, because it also introduces slower inference speeds and higher computational costs, many researches have attempted to ... | https://www.lesswrong.com/posts/Nr9xHvMTmznpFDrwg/llms-do-not-think-step-by-step-in-implicit-reasoning |
# AI #92: Behind the Curve
People don’t give thanks enough, and it’s actual Thanksgiving, so here goes.
Thank you for continuing to take this journey with me every week.
It’s a lot of words. Even if you pick and choose, and you probably should, it’s a lot of words. You don’t have many slots to spend on things like t... | https://www.lesswrong.com/posts/BGBLcy3JyjjrT8XbM/ai-92-behind-the-curve |
# Estimates of GPU or equivalent resources of large AI players for 2024/5
AI infrastructure numbers are hard to find with any precision. There are many reported numbers of “\[company\] spending Xbn on infrastructure this quarter” and “\[company\] has bought 100k H100s or “[has a cluster of 100k H100s](https://www.toms... | https://www.lesswrong.com/posts/bdQhzQsHjNrQp7cNS/estimates-of-gpu-or-equivalent-resources-of-large-ai-players |
# Gratitudes: Rational Thanks Giving
*Epistemic status: I wish I'd thought of writing this before the day rolled around. Brief and unpolished, although this is something I've thought about a lot on both personal and computational neuroscience levels. There are no strong conclusions, just some thoughts on gratitude you... | https://www.lesswrong.com/posts/ecrdpfgcHaFKpt2Df/gratitudes-rational-thanks-giving |
# Two interviews with the founder of DeepSeek
\# The Madness of High-Flyer: The Approach to LLM by an AI Giant that Few See
暗涌Waves (2023-05-23 22:50)
Written by 于丽丽
Edited by 刘旌
Translated by Cosmia Nebula
High-Flyer is probably the most exotic among the swarming multitude of competitors in the battle of larg... | https://www.lesswrong.com/posts/kANyEjDDFWkhSKbcK/two-interviews-with-the-founder-of-deepseek |
# The Big Nonprofits Post
There are lots of great charitable giving opportunities out there right now.
The first time that I served as a recommender in the Survival and Flourishing Fund (SFF) was back in 2021. [I wrote in detail about my experiences then](https://thezvi.substack.com/p/zvis-thoughts-on-the-survival-an... | https://www.lesswrong.com/posts/9n87is5QsCozxr9fp/the-big-nonprofits-post |
# You should consider applying to PhDs (soon!)
**TLDR**: In this post, I argue that if you are a junior AI safety researcher, you should consider *applying* to PhD programs in ML soon, especially if you have recently participated in an [AI](https://www.matsprogram.org/) [safety](https://www.lasrlabs.org/) [upskilling]... | https://www.lesswrong.com/posts/PdtkXcgbRpdHWRNt6/you-should-consider-applying-to-phds-soon |
# INTELLECT-1 Release: The First Globally Trained 10B Parameter Model
> We're excited to release INTELLECT-1, the first 10B parameter language model collaboratively trained across the globe. This represents a 10× scale-up from our previous research and demonstrates that large-scale model training is no longer confined... | https://www.lesswrong.com/posts/9cuJaJjDuhbpTid3Q/intellect-1-release-the-first-globally-trained-10b-parameter |
# (The) Lightcone is nothing without its people: LW + Lighthaven's big fundraiser
**Update Jan 19th 2025:** **The Fundraiser is over! We had raised over $2.1M when the fundraiser closed, and have a few more irons in the fire that I expect will get us another $100k-$200k. This is short of our $3M goal, which I think me... | https://www.lesswrong.com/posts/5n2ZQcbc7r4R8mvqc/the-lightcone-is-nothing-without-its-people |
# Exporting Facebook Comments, Again
I want comments on my social media crossposts to show up on my blog as a comment section, and mostly this works well: modern systems ( [Mastodon](https://www.jefftk.com/p/mastodon-replies-as-comments), [Bluesky](https://www.jefftk.com/p/importing-bluesky-comments), [LessWrong](http... | https://www.lesswrong.com/posts/i2crfsiLpMFYQnTkT/exporting-facebook-comments-again |
# Launching Applications for the Global AI Safety Fellowship 2025!
**TLDR**; [Applications](http://bit.ly/aisafetyfellowship) are accepted until December 31, 2024 on a rolling basis, to join a 3-6 month, fully-funded research program in AI safety. Fellows work with some of the world’s leading AI safety labs and resear... | https://www.lesswrong.com/posts/zNh2g5sCFDha72c2d/launching-applications-for-the-global-ai-safety-fellowship-1 |
# Visual demonstration of Optimizer's curse
*Epistemic status: I am currently taking* [*this course*](https://agentfoundations.study/)*, and after reading* [*this post on Goodhart's curse,*](https://arbital.greaterwrong.com/p/goodharts_curse?l=6g4) *I wanted to create some visuals to better understand the described ef... | https://www.lesswrong.com/posts/7rzCGyTdDXnNoawDP/visual-demonstration-of-optimizer-s-curse |
# The Shape of Heaven
*Status: Just for fun*
*Scene: Some kind of lobby, where various people and/or avatars stand around and discuss issues that went well or badly in their respective worlds.* A common topic of conversation: AI, and why it went wrong. The following is extracted from one of those conversations.*
It ... | https://www.lesswrong.com/posts/KkoQDGM9qTjDZCgti/the-shape-of-heaven |
# Introducing the Anthropic Fellows Program
We're launching the [Anthropic Fellows Program for AI Safety Research](https://boards.greenhouse.io/anthropic/jobs/4308166008), a pilot initiative designed to accelerate AI safety research and foster research talent. The program will provide funding and mentorship for a sma... | https://www.lesswrong.com/posts/xwebc6CLWuLmahcx6/introducing-the-anthropic-fellows-program |
# Magnitudes: Let's Comprehend the Incomprehensible!
*Summary: Comfort with really big and really small quantities is very useful for understanding the world and is perfectly doable with practice. Therefore, you should (1) try to familiarize yourself with the sizes of various things using e.g. spaced repetition and (2... | https://www.lesswrong.com/posts/jBRGqtrm3RqPPkGd3/magnitudes-let-s-comprehend-the-incomprehensible |
# AXRP Episode 39 - Evan Hubinger on Model Organisms of Misalignment
[YouTube link](https://youtu.be/FsGJyTfOZrs)
The ‘model organisms of misalignment’ line of research creates AI models that exhibit various types of misalignment, and studies them to try to understand how the misalignment occurs and whether it can be... | https://www.lesswrong.com/posts/sookiqxkzzLmPYB3r/axrp-episode-39-evan-hubinger-on-model-organisms-of-1 |
# [Letter] Chinese Quickstart
Dear lsusr,
I am a lesswrong user interested in learning Mandarin and living in China. My goal is understanding Chinese culture more broadly and geopolitics and Chinese tech policy more specifically. I could get CELTA and get a teaching job in China (not difficult), but it seems like I w... | https://www.lesswrong.com/posts/XohC5bpXRWAXxNS8c/letter-chinese-quickstart |
# Commenting Patterns by Platform
How does the growth of Bluesky compare to Mastodon in 2022? What about Google Plus in 2011? I can't answer this globally, but (now that my comment archive is [up to date](https://www.jefftk.com/p/exporting-facebook-comments-again)) I can look at it for my immediate social network by l... | https://www.lesswrong.com/posts/4aaXoBaCeDDrs4C3s/commenting-patterns-by-platform |
# Which Biases are most important to Overcome?
In other words, what apparent imperfections in human reasoning
A) Remain apparent after the Replication Crisis,
B) Aren't secretly adaptive/reasonable in some counterintuitive way, and
C) Deal most damage to the people they inhabit (and/or those close to them, and/or w... | https://www.lesswrong.com/posts/j789HDCKLoiKGjBik/which-biases-are-most-important-to-overcome |
# Teaching My Younger Self to Program: A case study of how I'd pass on my skill at self-learning
When I was 21 I wanted to become a video game programmer.
Except I didn’t know how to program.
I tried to teach myself and failed. Then I went back to school and learned the basics of Java in a week, aced the course in t... | https://www.lesswrong.com/posts/mTGYgreRefpiJnJYw/teaching-my-younger-self-to-program-a-case-study-of-how-i-d |
# Sorry for the downtime, looks like we got DDosd
We were down between around 7PM and 8PM PT today. Sorry about that.
It's hard to tell whether we got DDosd or someone just wanted to crawl us extremely aggressively, but we've had at least a few hundred IP addresses and random user agents request a lot of quite absurd... | https://www.lesswrong.com/posts/85xkq9Go9AAg3raJ8/sorry-for-the-downtime-looks-like-we-got-ddosd |
# Drexler's Nanotech Software
Two months ago I attended Eric Drexler's launch of
[MSEP.one](https://msep.one/). It's open source software, written by
people with professional game design experience, intended to catalyze
better designs for atomically precise manufacturing (or generative
nanotechnology, as he now ca... | https://www.lesswrong.com/posts/cKucz3rQvrbxoeq4o/drexler-s-nanotech-software |
# 2024 Unofficial LessWrong Census/Survey
~~**The Less Wrong General Census is unofficially here! You can take it at**~~ [~~**this link.**~~](https://docs.google.com/forms/d/e/1FAIpQLSe6VdB5oQi_eXJpu1JwHfuAV8_AN1S15w0wcD0cQWVDMkiG5A/viewform?usp=sf_link)
**Update: The census is closed, thank you all for taking it! I ... | https://www.lesswrong.com/posts/XmbmHRYPMWHTYdaHW/2024-unofficial-lesswrong-census-survey |
# Conjecture: A Roadmap for Cognitive Software and A Humanist Future of AI
*On Carcinogenic Complexity, Software Senescence and Cognitive Provenance: Our roadmap for 2025 and beyond*
It is mandatory to start any essay on AI in the post-ChatGPT era with the disclaimer that AI brings huge potential, and great risks. Un... | https://www.lesswrong.com/posts/H26ndkABmGvoc9PTC/conjecture-a-roadmap-for-cognitive-software-and-a-humanist |
# Fertility Roundup #4
There is little sign that the momentum of the situation is changing. Instead, things continue to slowly get worse, as nations in holes continue to keep digging. The longer we wait, the more expensive the ultimate price will be. We will soon find out what the new administration does, which could ... | https://www.lesswrong.com/posts/avhKKnJyJ6kisvkzk/fertility-roundup-4 |
# A case for donating to AI risk reduction (including if you work in AI)
*I work on Open Philanthropy’s AI Governance and Policy team, but I’m writing this in my personal capacity – several senior employees at Open Phil have argued with me about this!*
This is a brief-ish post addressed to people who are interested i... | https://www.lesswrong.com/posts/HT6H3h5Htq93aZFCk/a-case-for-donating-to-ai-risk-reduction-including-if-you |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.