text stringlengths 300 320k | source stringlengths 52 154 |
|---|---|
# Quantitative cruxes in Alignment
***Summary:** Sometimes researchers talk past each other because their intuitions disagree on underlying quantitative variables. Getting better estimates of these variables is action-guiding. This is an epistemological study to bring them to the surface, as well as discover which sou... | https://www.lesswrong.com/posts/ryCfHod3eFhkYipW9/quantitative-cruxes-in-alignment |
# Sources of evidence in Alignment
***Summary:** A short epistemological study to discover which sources of evidence can inform our predictions of action-relevant quantities in alignment.*
*This post follows* [*Quantitative cruxes*](https://www.lesswrong.com/posts/ryCfHod3eFhkYipW9/quantitative-cruxes-in-alignment)*,... | https://www.lesswrong.com/posts/9sjEFuGbb8WFZE9ER/sources-of-evidence-in-alignment |
# VC Theory Overview
In this post, I will give a brief overview of VC theory and computational learning theory. The reason for this is that I will later write a post about singular learning theory, in which I will refer back to this post. However, this post is self-contained, and may be interesting for anyone interest... | https://www.lesswrong.com/posts/uEwECj53prjKLcBC5/vc-theory-overview |
# Meta-rationality and frames
How should we think about thinking? In this sequence I outline an epistemology called *meta-rationality*, which in my opinion improves on existing epistemologies in a number of important ways. This first post introduces some of the key concepts in meta-rationality (in particular the conce... | https://www.lesswrong.com/posts/BEB7jjxgArztnzmse/meta-rationality-and-frames |
# Frames in context
In [my previous post](https://www.lesswrong.com/posts/BEB7jjxgArztnzmse/meta-rationality-and-frames), I introduced meta-rationality and frames, and described some examples of frames and some of their properties. In this post I’ll outline some of the limitations of existing ways of thinking about co... | https://www.lesswrong.com/posts/NqDrjFonQnKbLM3XJ/frames-in-context |
# Douglas Hofstadter changes his mind on Deep Learning & AI risk (June 2023)?
[A podcast interview](https://www.youtube.com/watch?v=lfXxzAVtdpU&t=1763s) (posted 2023-06-29) with noted AI researcher Douglas Hofstadter discusses his career and current views on AI (via [Edward Kmett](https://twitter.com/kmett/status/1675... | https://www.lesswrong.com/posts/kAmgdEjq2eYQkB5PP/douglas-hofstadter-changes-his-mind-on-deep-learning-and-ai |
# Monthly Roundup #8: July 2023
As always, plenty of non-AI things are happening, most of which do not justify their own posts. This is where all of those go. Skimming for things relevant to your interests is encouraged.
Before I begin, a Point of Order: I plan to have a post later regarding the rate limitation and d... | https://www.lesswrong.com/posts/83EmLzstzcM7X4isf/monthly-roundup-8-july-2023 |
# A Subtle Selection Effect in Overconfidence Studies
***TLDR:** The standard methods for studying overconfidence might be overestimating it. In asking people for confidence intervals—the* **narrowest** *intervals that they’re confident contain the true value—we’re selecting for the intervals that are most likely to b... | https://www.lesswrong.com/posts/e92rNCrN7Rtp8jx8T/a-subtle-selection-effect-in-overconfidence-studies |
# Ten Levels of AI Alignment Difficulty
Image from https://threadreaderapp.com/thread/1666482929772666880.html
Chris Olah rece... | https://www.lesswrong.com/posts/EjgfreeibTXRx9Ham/ten-levels-of-ai-alignment-difficulty |
# Should you announce your bets publicly?
Over time, I've noticed many examples of people making bets that don't make much sense from a subjective profit-maximizing point of view. Without naming the people involved, here are some examples:
1. A lends B some amount of USD to be paid back in seven years, with an effect... | https://www.lesswrong.com/posts/53EoY9rxTdmnDEnvf/should-you-announce-your-bets-publicly |
# Mechanistic Interpretability is Being Pursued for the Wrong Reasons
The success of prosiac A.I. seems to have generally surprised the rationalist community, including myself. Rationalists have correctly observed that deep learning is a process of mesa-optimization; stochastic gradient descent by backpropogation find... | https://www.lesswrong.com/posts/RTmFpgEvDdZMLsFev/mechanistic-interpretability-is-being-pursued-for-the-wrong |
# Two Percolation Puzzles
Take a very large chessboard (NxN, where N is huge).
Remove some fraction **1-p** of the squares at random, leaving a fraction **p** of them.
Can you place a queen on the firs... | https://www.lesswrong.com/posts/w3bJtJ5aD8ikQawZK/two-percolation-puzzles |
# The literature on aluminum adjuvants is very suspicious. Small IQ tax is plausible - can any experts help me estimate it?
I am a PhD biostatistician who has worked on a project in this area. I am hoping to crowdsource opinions on this issue, especially from readers with knowledge of neurology, nephrology and/or toxi... | https://www.lesswrong.com/posts/oFXXfP88nJrseeoMD/the-literature-on-aluminum-adjuvants-is-very-suspicious |
# Rational Unilateralists Aren't So Cursed
Much informal discussion of the Unilateralist’s Curse from Bostrom et al ([2016](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4959137/)) presents it as a sort of collective action problem: the chance of purely altruistic agents causing harm rises with the number of agents tha... | https://www.lesswrong.com/posts/TkwJupKzDpGHLxoG6/rational-unilateralists-aren-t-so-cursed |
# Twitter Twitches
The situation is evolving rapidly. Here’s where we stand as of the morning of July 4th.
#### Well You See What Happened Was…
Oh no! To be clear, by twitches, I mean ‘Elon refused to pay the cloud bill.’
As a result, Twitter has been forced to rate limit users.
1. This started out as 600 posts p... | https://www.lesswrong.com/posts/GFagfc5aYsRGQu7DL/twitter-twitches |
# My Time As A Goddess
[Epistemic status: Retrospective on 6-year-old actions]
Back in 2017, there was Rationalist Tumblr, which I participated in; then Discord came along. A number of regulars from Rattumb created a server and started to hang out.
It was a lawless hive of scum and villainy at the time. It has sin... | https://www.lesswrong.com/posts/jnPeBFx3sZBmJfeXA/my-time-as-a-goddess |
# AI labs' statements on governance
*June 6, 2024: **THIS POST HAS BEEN SUCCEEDED BY*** [**Companies' policy advocacy**](https://ailabwatch.org/resources/company-advocacy/)***. READ THAT INSTEAD OF THIS. THIS POST WILL NOT BE MAINTAINED.***
This is a collection of statements on government policy, regulation, and stan... | https://www.lesswrong.com/posts/iFrefmWAct3wYG7vQ/ai-labs-statements-on-governance |
# Ways I Expect AI Regulation To Increase Extinction Risk
The following are some very short stories about some of the **ways that I expect AI regulation to make x-risk worse.**
I don't think that each of these stories will happen. Some of them are mutually exclusive, by design. But, conditional upon AI heavy regulati... | https://www.lesswrong.com/posts/6untaSPpsocmkS7Z3/ways-i-expect-ai-regulation-to-increase-extinction-risk |
# Six (and a half) intuitions for SVD
The long-awaited[^ev3hm7mjvjb] sequel to my "[Six (and a half) intuitions for KL divergence](https://www.lesswrong.com/posts/no5jDTut5Byjqb4j5/six-and-a-half-intuitions-for-kl-divergence)" is finally here!
Thanks to the following people for feedback: Denizhan Akar, Rudolf Laine... | https://www.lesswrong.com/posts/iupCxk3ddiJBAJkts/six-and-a-half-intuitions-for-svd |
# Three camps in AI x-risk discussions: My personal very oversimplified overview
*\[I originally wrote this as a Facebook post, but I'm cross-posting here in case anybody finds it useful.\]*
Here's my current overview of the AI x-risk debate, along with a very short further reading list:
At a \*very\* overly simplif... | https://www.lesswrong.com/posts/5g2t4R9LJPB7BznE2/three-camps-in-ai-x-risk-discussions-my-personal-very |
# Dominant Assurance Contract Experiment #2: Berkeley House Dinners
As an afterthought at the end of my [previous post](https://www.lesswrong.com/posts/9cvK8QX9ZRBxmEfcz/consider-paying-for-literature-or-book-reviews-using) on dominant assurance contracts, I decided on a whim as I was typing to append a small dominant... | https://www.lesswrong.com/posts/nwjTPtbvcJeA6xDuu/dominant-assurance-contract-experiment-2-berkeley-house |
# Final Lightspeed Grants coworking/office hours before the application deadline
This is the last day to apply to [Lightspeed Grants](https://www.lesswrong.com/posts/xQ4ajnzavSgbYiko2/launching-lightspeed-grants-apply-by-july-6th)!
It's also the last office hours during which you can co-work with others on your appl... | https://www.lesswrong.com/events/5FkXquTewqPqY5ruA/final-lightspeed-grants-coworking-office-hours-before-the |
# Puffer-pope reality check
*This post is humor. It's meant to be entertaining and nothing more. **tl;dr:** nothing quite summarizes the weirdness of the 21rst century as well as this picture.*
I was sorting through my mind dump recently and found a "first-principles rant" on ... | https://www.lesswrong.com/posts/MicN7trMryDMm4B4u/puffer-pope-reality-check |
# Optimized for Something other than Winning or: How Cricket Resists Moloch and Goodhart's Law
Recently, there has been some controversy in the world of cricket. The full details are described well in [this piece](https://www.theguardian.com/sport/2023/jul/05/a-crease-a-stumping-and-spirit-of-the-game-the-ashes-contro... | https://www.lesswrong.com/posts/AZ4Hx7br9v5m5KbNm/optimized-for-something-other-than-winning-or-how-cricket |
# Exploring Functional Decision Theory (FDT) and a modified version (ModFDT)
### **TL;DR**
This post explores the role of Functional Decision Theory (FDT) in AI alignment, showcasing its efficacy in complex decision-making scenarios. The results obtained highlight the promising potential of FDT as a subject for furt... | https://www.lesswrong.com/posts/DMtzwPuFQtDmPEppF/exploring-functional-decision-theory-fdt-and-a-modified |
# AISN #13: An interdisciplinary perspective on AI proxy failures, new competitors to ChatGPT, and prompting language models to misbehave
Welcome to the AI Safety Newsletter by the [Center for AI Safety](https://www.safe.ai/). We discuss developments in AI and AI safety. No technical background required.
Subscribe [h... | https://www.lesswrong.com/posts/vu5gNmBKft4vDyNCZ/aisn-13-an-interdisciplinary-perspective-on-ai-proxy |
# (tentatively) Found 600+ Monosemantic Features in a Small LM Using Sparse Autoencoders
Using a sparse autoencoder, I present evidence that the resulting decoder (aka "dictionary") learned 600+ features for Pythia-70M layer_2's mid-MLP (after the GeLU), although I expect around 8k-16k features to be learnable.
Dicti... | https://www.lesswrong.com/posts/wqRqb7h6ZC48iDgfK/tentatively-found-600-monosemantic-features-in-a-small-lm |
# The risk-reward tradeoff of interpretability research
Interpretability research is conducted to improve our understanding of AI. Many see interpretability as essential for AI safety, but recently some have argued that it can also increase the risk posed by AI by facilitating improved AI capabilities. We agree, and i... | https://www.lesswrong.com/posts/HdqdqNC3MyABHzSqf/the-risk-reward-tradeoff-of-interpretability-research |
# If you wish to make an apple pie, you must first become dictator of the universe
*\[Cross-posted from The Roots of Progress. Written for a general audience. LW readers may find the first couple of sections to be familiar intro material.\]*
The word “robot” is derived from the Czech [*robota*](https://en.wiktionary.... | https://www.lesswrong.com/posts/5hhjY4QQk59DEGWuc/if-you-wish-to-make-an-apple-pie-you-must-first-become |
# [Linkpost] Introducing Superalignment
Some quotes:
> **Our approach**
> ----------------
>
> Our goal is to build a roughly human-level [automated alignment researcher](https://openai.com/blog/our-approach-to-alignment-research). We can then use vast amounts of compute to scale our efforts, and iteratively align s... | https://www.lesswrong.com/posts/Hna4aoMwr6Qx9rHBs/linkpost-introducing-superalignment |
# Infra-Bayesian Logic
*This research was produced during SERI MATS. Thanks to Vanessa Kosoy for providing ideas and reading drafts.*
Introduction
============
In this article, we investigate "infra-Bayesian logic" (IBL for short), which is an idea introduced by Vanessa Kosoy. The original idea is outlined in [Vanes... | https://www.lesswrong.com/posts/nEFAno6PsCKnNgkd5/infra-bayesian-logic |
# Announcing Manifund Regrants
Manifund is launching a new regranting program! We will allocate ~$2 million over the next six months based on the recommendations of our regrantors. Grantees can [apply for funding through our site](https://manifund.org/create); we’re also looking for additional regrantors and donors to... | https://www.lesswrong.com/posts/uq4BABSNyH6E8aArA/announcing-manifund-regrants |
# AI Intermediation
Why is Google such a valuable company? More than half of its money comes from search ads: when people are looking to spend money it's very common that they use Google. Even slightly influencing spending decisions, with the scale of their userbase, lets them bring in massive amounts of money. They a... | https://www.lesswrong.com/posts/ycxocYbFuwthps8zT/ai-intermediation |
# Open Thread - July 2023
If it’s worth saying, but not worth its own post, here's a place to put it.
If you are new to LessWrong, here's the place to introduce yourself. Personal stories, anecdotes, or just general comments on how you found us and what you hope to get from the site and community are invited. This is... | https://www.lesswrong.com/posts/mc8r2kNKLidAWJeaD/open-thread-july-2023 |
# AI #19: Hofstadter, Sutskever, Leike
The big news of the week is [OpenAI introducing the Superalignment Taskforce](https://openai.com/blog/introducing-superalignment). OpenAI is pledging 20% of compute secured to date towards a four year effort to solve the core technical challenges of superintelligence alignment. T... | https://www.lesswrong.com/posts/45D7rZXiG5gCoCveM/ai-19-hofstadter-sutskever-leike |
# Agency begets agency
The second way of cultivating excitement-based motivation I'll discuss in this sequence is cultivating agency, by which I mean the ability to try to shape the world. The key idea I want to convey here is that the feeling of agency is in many cases a self-fulfilling prophecy, because the hardest ... | https://www.lesswrong.com/posts/vL8A62CNK6hLMRp74/agency-begets-agency |
# Announcing the EA Archive
About the EA Archive
====================
The EA Archive is a project to preserve resources related to effective altruism in case of a sub-existential catastrophe such as nuclear war.
Its more specific, downstream motivating aim is to increase the likelihood that a movement akin to EA (i... | https://www.lesswrong.com/posts/kxTAaJJX4AftRSyc2/announcing-the-ea-archive |
# Understanding the two most common mental health problems in the world
*Co-authored with Amanda Metskas*
This is a linkpost from [ClearerThinking.org](https://www.clearerthinking.org/post/understand-how-other-people-think-a-theory-of-worldviews). We've included some excerpts of the article below, but [you can read t... | https://www.lesswrong.com/posts/7GpvjBDuzFf47hwom/understanding-the-two-most-common-mental-health-problems-in |
# A Defense of Work on Mathematical AI Safety
AI Safety was, a decade ago, nearly synonymous with obscure mathematical investigations of hypothetical agentic systems. Fortunately or unfortunately, this has largely been overtaken by events; the successes of machine learning and the promise, or threat, of large language... | https://www.lesswrong.com/posts/9KBcFxLLHpPv6CEhb/a-defense-of-work-on-mathematical-ai-safety |
# Progress links and tweets, 2023-07-06:
Terraformer Mark One, Israeli water management, & more
**Opportunities**
-----------------
* [Free virtual course for people in tech who want to work on biology](https://frequency.pillar.vc/bio/) (via [@kulesatony](https://twitter.com/kulesatony/status/1674078896134709249))
... | https://www.lesswrong.com/posts/LPBRvWi2i4JSfTeY3/progress-links-and-tweets-2023-07-06-terraformer-mark-one |
# Jesse Hoogland on Developmental Interpretability and Singular Learning Theory
[Jesse Hoogland](https://www.lesswrong.com/users/jesse-hoogland) is a research assistant at [David Krueger's lab](https://www.davidscottkrueger.com/) in Cambridge studying AI Safety who has recently been [publishing](https://www.lesswrong.... | https://www.lesswrong.com/posts/LmHJPWRMEvmCWe382/jesse-hoogland-on-developmental-interpretability-and |
# Localizing goal misgeneralization in a maze-solving policy network
***TLDR:** I am trying to understand how goal misgeneralization happens in the same maze-solving network* [*TurnTrout et al. work on.*](https://www.lesswrong.com/posts/cAC4AXiNC5ig6jQnc/understanding-and-controlling-a-maze-solving-policy-network) *No... | https://www.lesswrong.com/posts/vY9oE39tBupZLAyoC/localizing-goal-misgeneralization-in-a-maze-solving-policy |
# BOUNTY AVAILABLE: AI ethicists, what are your object-level arguments against AI notkilleveryoneism?
I am prepared to pay out anywhere between $20 and $100 to AI ethicists of the DAIR/"Stochastic Parrots" school of thought if they provide their object-level arguments against the idea that preventing AI from killing e... | https://www.lesswrong.com/posts/uTRafHCcjNfbAByyo/bounty-available-ai-ethicists-what-are-your-object-level |
# Empirical Evidence Against "The Longest Training Run"
Summary
=======
In this post, we empirically test Epoch AI’s theoretical model of an upper bound on AI training run lengths. According to this model, an upper bound for training run time can be estimated by assuming that the length of a training run is optimized... | https://www.lesswrong.com/posts/DaeHpWxvht43zaaje/empirical-evidence-against-the-longest-training-run |
# Two paths to win the AGI transition
Path 1: Solve Alignment
=======================
If alignment is possible, i think we are vastly more likely to win if we set up a scenario where we can safely iterate. Humans are really good at collectively collaborating on iterative engineering improvements! This is not an align... | https://www.lesswrong.com/posts/4ydJvoXLkXzGWDZnh/two-paths-to-win-the-agi-transition |
# What are the best non-LW places to read on alignment progress?
I've lately been thinking I should prioritize a bit more keeping up with alignment-relevant progress outside of LessWrong/Alignment Forum.
I'm curious if people have recommendations that stand out as reliably valuable, and/or have tips for finding "the... | https://www.lesswrong.com/posts/e2FQB5v5C7yMDPE8A/what-are-the-best-non-lw-places-to-read-on-alignment |
# Apparently, of the 195 Million the DoD allocated in University Research Funding Awards in 2022, more than half of them concerned AI or compute hardware research
Suggests interesting things about what the Department of Defense is focusing on, so I thought I should mention it.
The awards concerned "pursuing basic res... | https://www.lesswrong.com/posts/rEDjo94iPvXWkkt4L/apparently-of-the-195-million-the-dod-allocated-in |
# ask me about technology
I have a lot of opinions about technology - what's feasible, what isn't, what the best ways to do things are, and so on. Sometimes I write them down on my [blog](https://www.bhauth.com/blog/), but usually I don't. I also know some people who are good at designing technology.
If you want my p... | https://www.lesswrong.com/posts/3fZD6tK6emzdbvaB2/ask-me-about-technology |
# Can LessWrong provide me with something I find obviously highly useful to my own practical life?
When I meet new people, I often like to figure out how I can be useful in a really short amount of time, just by talking.
Example:
Often this will involve evangelizing some items from a short list of technologies th... | https://www.lesswrong.com/posts/u4JZDyAs5nau7xQMQ/can-lesswrong-provide-me-with-something-i-find-obviously |
# Internal independent review for language model agent alignment
Abstract:
---------
Language model agents (LMAs) expanding on AutoGPT are a highly plausible route to AGI. This route has large potential timeline and proliferation downsides, but large alignment advantages relative to other realistic paths to AGI. LMAs... | https://www.lesswrong.com/posts/Q7XWGqL4HjjRmhEyG/internal-independent-review-for-language-model-agent |
# Interpreting Modular Addition in MLPs
Summary
=======
In this post, we investigate one of Neel Nanda's [200 Concrete Open Problems in Mechanistic Interpretability](https://www.lesswrong.com/posts/LbrPTJ4fmABEdEnLf/200-concrete-open-problems-in-mechanistic-interpretability), namely problem A3.3: Interpret a 2L MLP (... | https://www.lesswrong.com/posts/cbDEjnRheYn38Dpc5/interpreting-modular-addition-in-mlps |
# Meetup Tip: Ask Attendees To Explain It
Summary
-------
A question I like asking when running meetups is, "do you think you could explain how to do this to someone else?" I usually ask it just after explaining the rules or steps for whatever we're doing this meetup. This timing only works for a meetup with an expli... | https://www.lesswrong.com/posts/xZxcqtFizatPwc4S7/meetup-tip-ask-attendees-to-explain-it |
# Introducing bayescalc.io
I made a simple online calculator for doing elementary hypothesis testing!
[Link to example shown.](https://bayescalc.io/#KCdoeXBvdGhlc2VzLSdIZWFsdGh5J35IYXMgZGlzZWFzZUFwKjkuM... | https://www.lesswrong.com/posts/QHjLgK59aHbJEbakG/introducing-bayescalc-io |
# Views on when AGI comes and on strategy to reduce existential risk
Summary: AGI isn't super likely to come super soon. People should be working on stuff that saves humanity in worlds where AGI comes in 20 or 50 years, in addition to stuff that saves humanity in worlds where AGI comes in the next 10 years.
T... | https://www.lesswrong.com/posts/sTDfraZab47KiRMmT/views-on-when-agi-comes-and-on-strategy-to-reduce |
# Request for feedback - infohazards in testing LLMs for causal reasoning?
**Edit:** I have no idea why this is getting silently downvoted, and would appreciate a comment or PM to explain.
Judea Pearl has tweeted a couple of times the claim that ChatGPT can't perform causal reasoning ([1](https://twitter.com/yudapear... | https://www.lesswrong.com/posts/cx9f8tYWtjMJRkGfQ/request-for-feedback-infohazards-in-testing-llms-for-causal |
# What Does LessWrong Think of Human Intelligence Augmentation in 2023?
[Zvi Recently asked on Twitter](https://twitter.com/TheZvi/status/1676916762594402304):
> If someone was founding a new AI notkilleveryoneism research organization, what is the best research agenda they should look into pursuing right now?
To wh... | https://www.lesswrong.com/posts/ewitKJEwvttzk6zMi/what-does-lesswrong-think-of-human-intelligence-augmentation |
# Writing this post as rationality case study
\[Before I begin: If you don't like this post, please let me know why. Even just few words like boring/off-topic/poorly-written may give me something to work with.\]
I wish to share a late struggle I have with rationality, because I think that it touches some interesting ... | https://www.lesswrong.com/posts/JzPxE4pnpxtPwAzWb/writing-this-post-as-rationality-case-study |
# Announcing AI Alignment workshop at the ALIFE 2023 conference
The upcoming [ALIFE 2023 conference](https://2023.alife.org/) is hosting a workshop on AI Alignment and Artificial Life. This will complement the [special session](https://humanvaluesandartificialagency.com/) on AI Alignment at the conference. The worksho... | https://www.lesswrong.com/posts/LQAQe9Bkgb3oqFCaJ/announcing-ai-alignment-workshop-at-the-alife-2023 |
# Fixed Point: a love story
> *Ring the bells that still can ring*
> *Forget your perfect offering*
> *There is a crack, a crack in everything*
> *That's how the light gets in.*
God is making excuses again, which is a real drag.
“I want to be authentic with you, I really do. But there’s a part of my mind that ... | https://www.lesswrong.com/posts/zDSZWsLEp7Pw3zePL/fixed-point-a-love-story |
# You must not fool yourself, and you are the easiest person to fool
The title of this post comes from [Richard Feynman's famous talk about cargo cult science](https://calteches.library.caltech.edu/51/2/CargoCult.htm). The principle is very broadly applicable, though. Applying the steps I’ve described in this sequence... | https://www.lesswrong.com/posts/rzK9YvNor3ZCccFFf/you-must-not-fool-yourself-and-you-are-the-easiest-person-to |
# What is everyone doing in AI governance
What is this post about, and how to use it
==========================================
This is a list of AI governance organizations with a brief description of their directions of work, so a reader can get a basic understanding of what they are doing and explore them in-depth... | https://www.lesswrong.com/posts/6gJMwXuWd3oskaLCo/what-is-everyone-doing-in-ai-governance |
# Why does anxiety (?) make me dumb?
Something about anxiety and my brain shutting off.
Disclaimer: I’m grappling with something I hardly understand at all, and I’m not all that used to the exercise.
One of our goals here is to improve our thinking. That’s fairly obvious. And I have another obvious statement: The pr... | https://www.lesswrong.com/posts/e7DuokQbLNxedbcru/why-does-anxiety-make-me-dumb |
# Blanchard's Dangerous Idea and the Plight of the Lucid Crossdreamer
> I'm beginning to wonder if he's constructed an entire system of moral philosophy around the effects of the loyalty mod—a prospect that makes me distinctly uneasy. It would hardly be the first time a victim of mental illness has responded to their ... | https://www.lesswrong.com/posts/RxxqPH3WffQv6ESxj/blanchard-s-dangerous-idea-and-the-plight-of-the-lucid |
# "Concepts of Agency in Biology" (Okasha, 2023) - Brief Paper Summary
This is a brief summary post of a recent paper titled "[The Concept of Agent in Biology: Motivations and Meanings](https://link.springer.com/article/10.1007/s13752-023-00439-z)" by Samir Okasha (2023).
In both Biology and AI Alignment, the concep... | https://www.lesswrong.com/posts/c27yRmcBxC6txibWW/concepts-of-agency-in-biology-okasha-2023-brief-paper |
# Eight Strategies for Tackling the Hard Part of the Alignment Problem
Thanks to Michael Ripa for feedback.
TL;DR
=====
* There are two types of problems that AI systems can have: problems that we can observe during development and problems that we can’t. The hardest part of the alignment problem involves problem... | https://www.lesswrong.com/posts/amBsmfFK4NFDtkHiT/eight-strategies-for-tackling-the-hard-part-of-the-alignment |
# Really Strong Features Found in Residual Stream
\[I'm writing this quickly because the results are really strong. I still need to do due diligence & compare to baselines, but it's really exciting!\]
[Last post](https://www.lesswrong.com/posts/wqRqb7h6ZC48iDgfK/tentatively-found-600-monosemantic-features-in-a-small-... | https://www.lesswrong.com/posts/Q76CpqHeEMykKpFdB/really-strong-features-found-in-residual-stream |
# A Two-Part System for Practical Self-Care
*Category: Life optimization*
*Epistemic status: Seems reasonable, not sure how this could possibly backfire.*[^6c66p8q2fum]
* * *
Sometimes I feel blue. The world turns dull and grey, and I feel like I don’t have the time or energy to accomplish anything. This is occasio... | https://www.lesswrong.com/posts/QHPJXfgdwnnfB5yNz/a-two-part-system-for-practical-self-care |
# H5N1. Just how bad is the situation?
How bad do you think the [H5N1](https://forum.effectivealtruism.org/topics/h5n1-virus) situation is? I tried to search on LessWrong and EffectiveAltruism, but haven't found that much discussion.
When I was reading about **H5N1**, a lot of my assumptions were broken. I was shocke... | https://www.lesswrong.com/posts/2wyz8yQEevJF8GDGo/h5n1-just-how-bad-is-the-situation |
# Whether LLMs "understand" anything is mostly a terminological dispute
*Epistemic status: speculating on what other people mean when they make certain claims, without having done any checking to validate. Probably (mostly) true anyways.*
Thought prompted by [Tsvi's latest post](https://www.lesswrong.com/posts/sTDfr... | https://www.lesswrong.com/posts/gEiWA3yp8Xvhi8yn9/whether-llms-understand-anything-is-mostly-a-terminological |
# What to read on the "informal multi-world model"?
People on LessWrong often talk in terms of an "informal many-world model". For example they talk about the worlds where the alignment problem is relatively easy vs the worlds where the alignment problem is really hard.
I wonder what are good references for this "in... | https://www.lesswrong.com/posts/gxoohskLWWhmZTXAh/what-to-read-on-the-informal-multi-world-model |
# Where are the people building AGI in the non-dumb way?
I am somewhat baffled by the fact that I have never ran into somebody who is actively working on developing a paradigm of AGI which is targeted at creating a system that the system is just inherently transparent to the operators.
If you're having a list sorting... | https://www.lesswrong.com/posts/x7qozbXNMvFFXr8Tm/where-are-the-people-building-agi-in-the-non-dumb-way |
# Quick proposal: Decision market regrantor using manifund (please improve)
*Epistemic status: Some fundamental errors but worth leaving up*
*I think the error with this piece is that it is a hammer looking for a nail, rather than the reverse. Austin makes the point well* [*here*](https://forum.effectivealtruism.org/... | https://www.lesswrong.com/posts/of3unwQLhdiaCXNey/quick-proposal-decision-market-regrantor-using-manifund |
# Why have exposure notification apps been (mostly) discontinued?
I recently noticed that exposure notifications for COVID-19 in the app that I had been using, called CA Notify, had stopped working (I had attended an event a few days prior, and was feeling ever-so-slightly unwell, so I thought to check to see if there... | https://www.lesswrong.com/posts/EMhYiNrSFSvz26JTX/why-have-exposure-notification-apps-been-mostly-discontinued |
# The Seeker’s Game – Vignettes from the Bay
Introduction
============
Last year, one conversation left a lasting impression on me. A friend remarked on the challenges of navigating "corrupting forces" in the Bay Area. Intrigued by this statement, I decided to investigate the state of affairs in the Bay if I had the ... | https://www.lesswrong.com/posts/yXLEcd9eixWucKGHg/the-seeker-s-game-vignettes-from-the-bay |
# Some reasons to not say "Doomer"
"Doomer" has become a common term to refer to people with pessimistic views about outcomes from AI. I claim this is not a helpful term on net, and generally will cause people to think less clearly.
Reification of identity + making things tribal
--------------------------------------... | https://www.lesswrong.com/posts/aB4o34HyqBtRJwYGG/some-reasons-to-not-say-doomer |
# The virtue of determination
At the end of Replacing Guilt, Nate talks about the virtues of [desperation](https://mindingourway.com/desperation/), [recklessness](https://mindingourway.com/recklessness/) and [defiance](https://mindingourway.com/defiance/). None of these really hook neatly into my motivation system, bu... | https://www.lesswrong.com/posts/wJyKcFYMq4DLFtHvN/the-virtue-of-determination |
# The world where LLMs are possible
In [Artificial Addition](https://www.lesswrong.com/posts/YhgjmCxcQXixStWMC/artificial-addition), Eliezer used the ability to do arithmetic as a metaphor for intelligence. I really like this essay. It's witty and enlightening. And yet I have to admit it aged not so well. Among severa... | https://www.lesswrong.com/posts/fduAAMAWfp4bw9GQo/the-world-where-llms-are-possible |
# Consciousness as a conflationary alliance term for intrinsically valued internal experiences
*Tl;dr: In this post, I argue from many anecdotes that the concept of 'consciousness' is more conflated than people realize, in that there's a lot of divergence in what people mean by "consciousness", and people are unaware ... | https://www.lesswrong.com/posts/KpD2fJa6zo8o2MBxg/consciousness-as-a-conflationary-alliance-term-for |
# Open-minded updatelessness
Summary
-------
Bounded agents might be unaware of possibilities relevant to their decision-making. That is, they may not just be uncertain, but fail to conceive of some relevant hypotheses entirely. What's more, commitment races might pressure early AGIs into adopting an updateless polic... | https://www.lesswrong.com/posts/uPWDwFJnxLaDiyv4M/open-minded-updatelessness |
# “Reframing Superintelligence” + LLMs + 4 years
**Background**
--------------
In January 2019, FHI published *Reframing Superintelligence,*[^85plkbsx4qr] a book-length technical report on prospects for advanced AI. OpenAI published the first paper on GPT-2 a month later. Advances since then have been strange and rap... | https://www.lesswrong.com/posts/LxNwBNxXktvzAko65/reframing-superintelligence-llms-4-years |
# Consider Joining the UK Foundation Model Taskforce
A few months ago, Ian Hogarth wrote the Financial Times Op-Ed headlined “[We must slow down the race to God-like AI](https://www.ft.com/content/03895dc4-a3b7-481e-95cc-336a524f2ac2).”
A few weeks ago, he was appointed head of the UK Foundation Model Taskforce, and ... | https://www.lesswrong.com/posts/xgXcZQd5eqMqpAw3i/consider-joining-the-uk-foundation-model-taskforce |
# Frontier AI Regulation
This paper is about (1) "government intervention" to protect "against the risks from frontier AI models" and (2) some particular proposed safety standards. It's by Markus Anderljung, Joslyn Barnhart (Google DeepMind), Jade Leung (OpenAI governance lead), Anton Korinek, Cullen O'Keefe (OpenAI),... | https://www.lesswrong.com/posts/gDgxnNSDeFQicaHtJ/frontier-ai-regulation |
# Is the Endowment Effect Due to Incomparability?
Thanksgiving break, early 2000s. A hectic time in the Dorst household. Grandparents are in town. The oversized puppy threatens books, lamps, and (he thinks) squirrels. There are meals to make, grandparents to impress, children to manage.
But most important—at least, a... | https://www.lesswrong.com/posts/MrcTzbYeZ3xnh9mGj/is-the-endowment-effect-due-to-incomparability |
# Incentives from a causal perspective
*Post 4 of *[*Towards Causal Foundations of Safe AGI*](https://www.alignmentforum.org/s/pcdHisDEGLbxrbSHD)*, preceded by *[*Post 1: Introduction*](https://www.alignmentforum.org/posts/3oQCY4he4zRzF6vQb/introduction-to-towards-causal-foundations-of-safe-agi)*,* [*Post 2: Causality... | https://www.lesswrong.com/posts/Xm4vSHaKAmfRvgBgi/incentives-from-a-causal-perspective |
# I think Michael Bailey's dismissal of my autogynephilia questions for Scott Alexander and Aella makes very little sense
I am autogynephilic, and [there's been a lot of autogynephilia talk lately.](https://www.lesswrong.com/posts/RxxqPH3WffQv6ESxj/) One subject that sometimes comes up, but hasn't been discussed much ... | https://www.lesswrong.com/posts/oHn8yvzn5uGvPYmsb/i-think-michael-bailey-s-dismissal-of-my-autogynephilia |
# Modeling the impact of AI safety field-building programs
Summary
=======
AI safety field-building is a large space, with many possible programs. The [Center for AI Safety](https://www.safe.ai/) (CAIS) has commissioned cost-effectiveness models to support decisions regarding how to prioritize between these programs.... | https://www.lesswrong.com/posts/RriSsowEA5dgqbTyM/modeling-the-impact-of-ai-safety-field-building-programs-1 |
# Cost-effectiveness of student programs for AI safety research
Summary
=======
This post explores the cost-effectiveness of field-building programs for students, specifically the Atlas Fellowship (a rationality program, with some AI safety programming), MLSS (an ML safety course for undergraduates), a top-tier unive... | https://www.lesswrong.com/posts/GcpvkusNtmMhwfChS/cost-effectiveness-of-student-programs-for-ai-safety-1 |
# Drawn Out: a story
**Note: this story shouldn’t be interpreted as anywhere near realistic. It’s way too anthropomorphic, and intended to be a work of fiction not prediction—although hopefully it will spark some interesting ideas, like good fiction often does.**
> *Will he make many supplications unto thee? will he ... | https://www.lesswrong.com/posts/jzcsugm382vp4h36H/drawn-out-a-story |
# ACX Montreal Meetup July 22nd 2023
**Description:**
Come on out to the ACX (Astral Codex Ten) Montreal Meetup! This week, we're keeping it open-ended for a freewheeling discussion.
**Venue**: L'Esplanade Tranquille, 1442 Clark, indoors. When you have reached the second floor, go to the right.
**Date & Time:** Sat... | https://www.lesswrong.com/events/vWocMp9vNGwjPmM8z/acx-montreal-meetup-july-22nd-2023 |
# Two Hot Takes about Quine
I read Quine's *Word and Object* on vacation last week. Overall it was fine, but there were two things that might be worth quick mentions.
Quine, Supervised Learning Supremacist
--------------------------------------
One important facet of the book is Quine's picture of how humans learn l... | https://www.lesswrong.com/posts/csCqXMuYYFuTgtCwo/two-hot-takes-about-quine |
# [UPDATE: deadline extended to July 24!] New wind in rationality’s sails: Applications for Epistea Residency 2023 are now open
UPDATE 2: Do you have any questions about the Epistea Residency? Join us for a live Q&A call with the organizers on July 20, Thursday 9 am CET / Wednesday midnight 0 am PDT at [https://us06we... | https://www.lesswrong.com/posts/qHxW26uouWxk6SK4e/update-deadline-extended-to-july-24-new-wind-in-rationality |
# OpenAI Launches Superalignment Taskforce
In their announcement [Introducing Superalignment](https://openai.com/blog/introducing-superalignment), OpenAI committed 20% of secured compute and a new taskforce to solving the technical problem of aligning a superintelligence within four years. Cofounder and Chief Scientis... | https://www.lesswrong.com/posts/NSZhadmoYdjRKNq6X/openai-launches-superalignment-taskforce |
# Announcing The Roots of Progress Blog-Building Intensive
Today we’re launching [The Roots of Progress Blog-Building Intensive](https://fellowship.rootsofprogress.org/), an 8-week program for aspiring progress writers to start or grow a blog.
**What:** Through this program you will launch (or re-launch) a blog/Subst... | https://www.lesswrong.com/posts/iJDmyht7iGw53E2ub/announcing-the-roots-of-progress-blog-building-intensive |
# My Weirdest Experience
I wrote about this a long time ago in some comments here- I'm not sure where or when- but there seemed to be a lot of interest in it. Because of this, I thought I would share the post I've written about it here.
\*Note- this is a true story. It actually happened to me. I have done my best to... | https://www.lesswrong.com/posts/FxjJn8ZyWPbm75qpP/my-weirdest-experience |
# Introducing Fatebook: the fastest way to make and track predictions
[Fatebook](https://fatebook.io/) is a website that makes it extremely low friction to make and track predictions.
It's designed... | https://www.lesswrong.com/posts/yS3d46m23wRKDQobt/introducing-fatebook-the-fastest-way-to-make-and-track |
# Do we automatically accept propositions?
*Epistemic Status: Exploratory (I'm searching for an expert or someone who has studied and rigorously investigated these claims/essays, as I intend to, but time doesn't permit me now)*
16 years ago, Eliezer asked "[Do we Believe Everything we're told?](https://www.lessw... | https://www.lesswrong.com/posts/XZCRK69pqvzDqgeBi/do-we-automatically-accept-propositions |
# How do low level hypotheses constrain high level ones? The mystery of the disappearing diamond.
Consider two friends, Alice and Bob, trying to figure out what happened to a diamond that disappeared from a museum. They do so in the form of a game that is kind of an approximation to [Solomonoff induction](https://www.... | https://www.lesswrong.com/posts/CsyJCenESEsAESzWg/how-do-low-level-hypotheses-constrain-high-level-ones-the |
# Why it's necessary to shoot yourself in the foot
Without shooting yourself in the foot, learning lacks motivation. Complexity without reason is really confusing.
---
At the Recurse Center Feb 2020: I watch a talk during presentations about someone optimizing a database engine. It was really complicated, and I reme... | https://www.lesswrong.com/posts/JLcHnXioJ74Pjn4jo/why-it-s-necessary-to-shoot-yourself-in-the-foot |
# Announcing the AI Fables Writing Contest!
Winners have [been announced](https://forum.effectivealtruism.org/posts/EAmfYSBaJsMzHY2cW/ai-fables-writing-contest-winners)!
> TL;DR: Writing Contest for AI Fables
>
> Deadline: Sept 1/Oct 1
>
> Prizes: $1500/1000/500, consideration for writing retreat
>
> Word length:... | https://www.lesswrong.com/posts/CquHeXpbCjaLBffnr/announcing-the-ai-fables-writing-contest |
# Popularizing vibes vs. models
I'm [working on](https://www.lesswrong.com/posts/RNqguj2PdcodYLnkr/aging-and-the-geroscience-hypothesis) a series of blog posts to popularize The Handbook of the Biology of Aging.
One of the dilemmas I face is between popularizing vibes vs. popularizing models. What do I mean by that?
... | https://www.lesswrong.com/posts/nvx7Xe2shwP2thBRL/popularizing-vibes-vs-models |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.