text stringlengths 300 320k | source stringlengths 52 154 |
|---|---|
# In defense of flailing, with foreword by Bill Burr
> I don't give a \[expletive\] who you are, if the world is ending, and you're getting chased by zombies, you're not running around saying "oh golly gee. Oh *heck*... Aw jimmety cricket!"...You know?... It's the *end of the world with zombies*. From the beginning, ... | https://www.lesswrong.com/posts/L3aWaaikkdyw7KQhk/in-defense-of-flailing-with-foreword-by-bill-burr |
# Things That Make Me Enjoy Giving Career Advice
Introduction
============
I spend a bunch of my time having various kinds of mentoring conversations with people. I generally enjoy doing this, but some things make these conversations *much *more fun for me than others! This post is an attempt to compile some of these... | https://www.lesswrong.com/posts/syZAZfnG8d3m8YFgw/things-that-make-me-enjoy-giving-career-advice |
# Apply for Productivity Coaching and AI Alignment Mentorship
Hello. I am an AI alignment researcher who has a background in mental health and teaching. I originally trained to do therapy before deciding I could best help the world by working on AI alignment directly. I want to use my experience in mental health and p... | https://www.lesswrong.com/posts/hZtfbLQemcp6sm59c/apply-for-productivity-coaching-and-ai-alignment-mentorship |
# Pivotal outcomes and pivotal processes
***tl;dr:** If you think humanity is on a dangerous path, and needs to "pivot" toward a different future in order to achieve safety, consider how such a pivot could be achieved by multiple acts across multiple persons and institutions, rather than a single act. Engaging more a... | https://www.lesswrong.com/posts/etNJcXCsKC6izQQZj/pivotal-outcomes-and-pivotal-processes |
# Forecasting Fusion Power
This is a linkpost for ["The future of fusion" book review.](https://astralcodexten.substack.com/p/your-book-review-the-future-of-fusion)
IMO it's not really a book review, it's a well-written and data-driven argument that commercial fusion power is about 10-15 years away, no really for... | https://www.lesswrong.com/posts/NQwCvpwtecXj34uSf/forecasting-fusion-power |
# Do yourself a FAVAR: security mindset
(We're always in search of a concise and actionable way to communicate security mindset. Here's my attempt.)
Just as a sprocket you don't use never fails and code you don't write has no bugs, an input constraint you don't require is never violated.
Want a secure system? Then D... | https://www.lesswrong.com/posts/GBStFZinQYsyP8qHY/do-yourself-a-favar-security-mindset |
# Why don't we think we're in the simplest universe with intelligent life?
My gut instinct on [metacosmology](https://www.lesswrong.com/posts/dPmmuaz9szk26BkmD/shortform?commentId=N8oamtFAhWKEbyCBq) is that if were a simpler computation than our universe that produced intelligent life, we'd probably be there instead o... | https://www.lesswrong.com/posts/qHnGv4LvFHX6AjaJ3/why-don-t-we-think-we-re-in-the-simplest-universe-with |
# Scott Aaronson is joining OpenAI to work on AI safety
Scott Aaronson is a computer scientist at the University of Texas in Austin, whose research mainly focuses on quantum computing and complexity theory. He's at least very adjacent to the Rationalist/LessWrong community. After some comments on his blog and then cov... | https://www.lesswrong.com/posts/Tk5ovpucaqweCu4tu/scott-aaronson-is-joining-openai-to-work-on-ai-safety |
# Can DALL-E understand simple geometry?
I recently got access to DALL-E and wanted to experiment with what it could and couldn't understand. (I'm a little behind the times, I know. Us non-famous people have to spend a while longer on the waitlist.) Most of the other investigations into DALL-E seemed to be focusing on... | https://www.lesswrong.com/posts/JGftZ8CgBt3dZJNto/can-dall-e-understand-simple-geometry |
# To what extent have ideas and scientific discoveries gotten harder to find?
This post was funded by a grant from Ben Pace.
* * *
Ben Pace asks:
> To what extent have ideas and scientific discoveries gotten harder to find?
>
> Related: Why are there no gentlemen scientists any more — i.e. rich people who make nov... | https://www.lesswrong.com/posts/iXtz9wTsT28eqCDjC/to-what-extent-have-ideas-and-scientific-discoveries-gotten |
# "Brain enthusiasts" in AI Safety
*TL;DR: If you're a student of cognitive science or neuroscience and are wondering whether it can make sense to work in AI Safety, this guide is for you! (Spoiler alert: the answer is "mostly yes").*
Motivation
==========
AI Safety is a rapidly growing field of research that is sin... | https://www.lesswrong.com/posts/nfoYnASKHczH4G5pT/brain-enthusiasts-in-ai-safety |
# The best 'free solo' (rock climbing) video
I think this might be the 'best' 'free solo' free solo video (including "Free Solo" the movie, or any other climbing movie I've seen):
- [I tried free solo with Alex Honnold **Insane experience** - YouTube](https://www.youtube.com/watch?v=Cyya23MPoAI)
I was pretty disappo... | https://www.lesswrong.com/posts/8gixvhiXWpyEYzfR4/the-best-free-solo-rock-climbing-video |
# Juneberry Cake
The [Juneberries](https://en.wikipedia.org/wiki/Amelanchier#Uses_and_cultivation) are ripe, and the kids got pretty excited about picking some at the park:
[](https://www.jefftk.com/picking-juneberries-big.jpg)
This is not a berry I knew about growi... | https://www.lesswrong.com/posts/LfxMY8qwKzKLEKxmE/juneberry-cake |
# Parliaments without the Parties
*Epistemic status:* [*butterfly idea*](https://www.lesswrong.com/posts/R6M4vmShiowDn56of/butterfly-ideas)
Parliaments might be great, but they suffer from a number of problems. For example, in Israel the opposition is [refusing to vote](https://www.timesofisrael.com/lacking-votes-coa... | https://www.lesswrong.com/posts/hPZtizHzuxmJwgG4C/parliaments-without-the-parties |
# [Link-post] On Deference and Yudkowsky's AI Risk Estimates
This is a link-post to [a piece](https://forum.effectivealtruism.org/posts/NBgpPaz5vYe3tH4ga/on-deference-and-yudkowsky-s-ai-risk-estimates) I just posted to the EA Forum, discussing negative aspects of Eliezer Yudkowsky's forecasting track record. In case i... | https://www.lesswrong.com/posts/27EznPncmCtnpSojH/link-post-on-deference-and-yudkowsky-s-ai-risk-estimates |
# AI misalignment risk from GPT-like systems?
Right now, it seems that the most likely way we're gonna get an (intellectually) universal AI is by scaling models such as GPT. That is, models trained by self-supervised learning on massive piles of data, perhaps with a similar architecture to the transformer.
I do not ... | https://www.lesswrong.com/posts/LsNMRYLKnSphpFqdt/ai-misalignment-risk-from-gpt-like-systems |
# Where I agree and disagree with Eliezer
(*Partially in response to *[*AGI Ruin: A list of Lethalities*](https://www.lesswrong.com/posts/uMQ3cqWDPHhjtiesc/agi-ruin-a-list-of-lethalities). *Written in the same rambling style. Not exhaustive.*)
### Agreements
1. Powerful AI systems have a good chance of deliberately... | https://www.lesswrong.com/posts/CoZhXrhpQxpy9xw9y/where-i-agree-and-disagree-with-eliezer |
# Half-baked alignment idea: training to generalize
Note: I think there's a decent chance that the idea I describe is misguided, redundant, naive. My apologies if that is the case, and please feel free to point me towards existing, related writing or research.
Thanks to Peter Barnett and Justis Mills for feedback on ... | https://www.lesswrong.com/posts/a4f9EkDtsBHvfskdt/half-baked-alignment-idea-training-to-generalize |
# Let's See You Write That Corrigibility Tag
The [top-rated comment](https://www.lesswrong.com/posts/uMQ3cqWDPHhjtiesc/agi-ruin-a-list-of-lethalities?commentId=HRDoDnHv8bvoW7oPZ) on "[AGI Ruin: A List of Lethalities](https://www.lesswrong.com/posts/uMQ3cqWDPHhjtiesc/agi-ruin-a-list-of-lethalities)" claims that many ot... | https://www.lesswrong.com/posts/AqsjZwxHNqH64C2b6/let-s-see-you-write-that-corrigibility-tag |
# An AI defense-offense symmetry thesis
*Epistemic status: Not well argued for, haven’t spent much time on it, and it’s not very worked out. This thesis is not new at all, and is implicit in a lot (most?) of the discussion about x-risk from AI.* The intended contribution of this post is to state the thesis explicitly ... | https://www.lesswrong.com/posts/dPe87urYGQPA4gDEp/an-ai-defense-offense-symmetry-thesis |
# Causal confusion as an argument against the scaling hypothesis
Abstract
========
*We discuss the possibility that causal confusion will be a significant alignment and/or capabilities limitation for current approaches based on "*the scaling paradigm"*: unsupervised offline training of increasingly large neural nets ... | https://www.lesswrong.com/posts/FZL4ftXvcuKmmobmj/causal-confusion-as-an-argument-against-the-scaling |
# Relationship Advice Repository
Over the years, LessWrong has hosted some pretty great advice threads, including [Best Textbooks on Every Subject](https://www.lesswrong.com/posts/xg3hXCYQPJkwHyik2/the-best-textbooks-on-every-subject), [Boring Advice Repository](https://www.lesswrong.com/posts/HEn2qiMxk5BggN83J/boring... | https://www.lesswrong.com/posts/AvjbBjAAbKBk73v5F/relationship-advice-repository |
# Key Papers in Language Model Safety
Introduction
============
One of the fastest growing subfields of AI safety focuses on language models (LMs) such as OpenAI’s GPT-3. This article provides summaries of a few dozen key papers in several important subfields of language model safety, namely:
* Persuasion
* Trut... | https://www.lesswrong.com/posts/H4jH7J2BbeokH3Chu/key-papers-in-language-model-safety |
# Announcing the DWATV Discord
I have decided this blog should at least experiment with having a discord server, so I made one. I kept it simple. There’s a few channels for the things the blog tends to be interested in – games, rationality, Covid – along with places to ask about suggesting future post topics, hiring m... | https://www.lesswrong.com/posts/2YAeneeqGz6KpXWTD/announcing-the-dwatv-discord |
# On corrigibility and its basin
Post somewhat inspired by Eliezers "well why didn't anyone else write it then?"
Corrigibility has various slightly different definitions, but the general rough idea is of an AI that does what we want. An AI that doesn't trick us. An AI that doesn't try to break out of its box. An AI ... | https://www.lesswrong.com/posts/3Mwm7bpWgyvqwrMBT/on-corrigibility-and-its-basin |
# Parable: The Bomb that doesn't Explode
You're an engineer working for a military contractor. One day, your project manager comes to you and asks you to design a container for plastic explosive. It is to contain several kilograms of C4, enough to destroy a building. This is pretty dangerous, but you know that C4 is a... | https://www.lesswrong.com/posts/u536zbXxJwYjybeyM/parable-the-bomb-that-doesn-t-explode |
# Steam
"Steam" is one possible opposite of [Slack](https://www.lesswrong.com/posts/yLLkWMDbC9ZNKbjDG/slack). I sketch a speculative view of steam as a third 'cognitive currency' almost like probability and utility. This is a very informal and intuitive post, outlining concepts which could possibly have a nice formal ... | https://www.lesswrong.com/posts/inedT6KkbLSDwZvfd/steam |
# Evaluating a Corsi-Rosenthal Filter Cube
When I wrote about [testing](https://www.jefftk.com/p/ceiling-air-purifier) my ceiling air purifier prototype, two common questions were:
* Where is your control group?
* How does this compare to a [Corsi-Rosenthal](https://en.wikipedia.org/wiki/Corsi%E2%80%93Rosenthal_B... | https://www.lesswrong.com/posts/27wG75fAeQLhnF5EQ/evaluating-a-corsi-rosenthal-filter-cube |
# Dagger of Detect Evil
A philosopher walked into a magic shop.
"Ugh," said Phil. The philosopher leaned his spear against the hatstand and slumped onto a fluffy cushioned chair.
"Can I get you something to drink? How about Essence of Essence?" said Wiz, the shop owner. She handed Phil a bottle of perfectly invisibl... | https://www.lesswrong.com/posts/BhoMqZuofPNmhG7We/dagger-of-detect-evil |
# Common but neglected risk factors that may let you get Paxlovid
[The CDC lists the conditions](https://www.cdc.gov/coronavirus/2019-ncov/need-extra-precautions/people-with-medical-conditions.html) that make a person "more likely to get very sick from COVID-19." Currently, Paxlovid supply is abundant, but available o... | https://www.lesswrong.com/posts/zt4WTTQpWbBWLhaBs/common-but-neglected-risk-factors-that-may-let-you-get |
# Getting from an unaligned AGI to an aligned AGI?
| **Summary / Preamble** ---------------------- In [AGI Ruin: A List of Lethalities](https://www.lesswrong.com/posts/uMQ3cqWDPHhjtiesc/agi-ruin-a-list-of-lethalities), Eliezer writes *“A cognitive system with sufficiently high cognitive powers, **given any medium-ban... | https://www.lesswrong.com/posts/ZmZBataeY58anJRBb/getting-from-an-unaligned-agi-to-an-aligned-agi |
# The inordinately slow spread of good AGI conversations in ML
Spencer Greenberg wrote [on Twitter](https://twitter.com/SpencrGreenberg/status/1536760598486196225):
> Recently @KerryLVaughan has been critiquing groups trying to build AGI, saying that by being aware of risks but still trying to make it, they’re reckle... | https://www.lesswrong.com/posts/Rkxj7TFxhbm59AKJh/the-inordinately-slow-spread-of-good-agi-conversations-in-ml |
# Progress links and tweets, 2022-06-20
**Links**
---------
* [A confident nation builds so that future people may look back and marvel at their ingenuity and vision](https://marginalrevolution.com/marginalrevolution/2022/06/against-historic-preservation-ii.html) (Alex Tabarrok)
* [“Construction is a sign of life... | https://www.lesswrong.com/posts/F6vsSvN34Bxx63vHQ/progress-links-and-tweets-2022-06-20 |
# Hydrophobic Glasses Coating Review
Several years ago I wrote one of my most controversial posts, [Don't Clean Your Glasses](https://www.jefftk.com/p/dont-clean-your-glasses). If I keep wearing dusty glasses I find I stop noticing the dust, and then I save the time of washing and drying. I'm embarrassed to admit it, ... | https://www.lesswrong.com/posts/4P9LkaG3rSZLxsrc5/hydrophobic-glasses-coating-review |
# A Quick List of Some Problems in AI Alignment As A Field
## 1. MIRI as central point of failure for... a few things...
For the past decade or more, if you read an article saying "AI safety is important", and you thought, "I need to donate or apply to work somewhere", MIRI was the default option. If you looked at FL... | https://www.lesswrong.com/posts/yQJxi5mMZQopYXZk5/a-quick-list-of-some-problems-in-ai-alignment-as-a-field |
# Security Mindset: Lessons from 20+ years of Software Security Failures Relevant to AGI Alignment
Background
==========
I have been doing red team, blue team (offensive, defensive) computer security for a living since September 2000. The goal of this post is to compile a list of general principles I've learned durin... | https://www.lesswrong.com/posts/Ke2ogqSEhL2KCJCNx/security-mindset-lessons-from-20-years-of-software-security |
# Solstice Movie Review: Summer Wars
If a classic is a movie everyone praises and no one watches, then Summer Wars is not a classic.
https://en.m.wikipedia.org/wiki/Summer_Wars
I have met only one person who knew the movie without me directing them to it - the one who directed me to it. Maybe I run in the wrong circ... | https://www.lesswrong.com/posts/CrEJpjzvLubfLH5WZ/solstice-movie-review-summer-wars |
# House Phone
Until recently, houses had phones. Anyone could place a call out, and whoever was there could answer an incoming call. With cell phones there's much less reason to have house phones, and since moving out of my parents' house I haven't lived anywhere that had one. This seems to work fine for adults, but w... | https://www.lesswrong.com/posts/W28s6tmTHMFPL9iiF/house-phone |
# Reflection Mechanisms as an Alignment target: A survey
This is a product of the [2022 AI Safety Camp](https://aisafety.camp/). The project has been done by [Marius Hobbhahn](https://www.mariushobbhahn.com/aboutme/) and [Eric Landgrebe](https://www.linkedin.com/in/eric-landgrebe-959698169/) under the supervision of [... | https://www.lesswrong.com/posts/XyBWkoaqfnuEyNWXi/reflection-mechanisms-as-an-alignment-target-a-survey-1 |
# Make learning a reality
The following passage comes from *What Do You Care What Other People Think?* by Richard Feynman. It's a collection of stories from his life and curious adventures.
> *We had the Encyclopedia Britannica at home. When I was a small boy he \[Feynman's father\] used to sit me on his lap and read... | https://www.lesswrong.com/posts/wj3y3omMqJJSv4rsq/make-learning-a-reality |
# How do I use caffeine optimally?
I just started a new job, and I'm really excited for the potential of drinking coffee, but I'm worried about dependency or other negative effects.
I don't want to be upset or feel unwell if I don't have coffee, and I don't want any long-term negative health effects from drinking cof... | https://www.lesswrong.com/posts/HyAjoi8mtHrrozNhk/how-do-i-use-caffeine-optimally |
# Google's new text-to-image model - Parti, a demonstration of scaling benefits
Google has released their latest text-to-image generation model- [Parti](https://parti.research.google/). They provide a few prompts and showcase the differences between models trained on 350M, 750M, 3B and 20B parameters.
One difference ... | https://www.lesswrong.com/posts/tMr3HJwitJCbQ5HTc/google-s-new-text-to-image-model-parti-a-demonstration-of |
# Announcing the LessWrong Curated Podcast
You can now listen to LessWrong Curated posts in podcast form on [Spotify](https://open.spotify.com/show/7vqBzO0ejqiLiXyTECEeBY), [Apple Podcasts](https://podcasts.apple.com/us/podcast/lesswrong-curated-podcast/id1630783021), [Audible](https://www.audible.com/pd/LessWrong-Cur... | https://www.lesswrong.com/posts/kDjKF2yFhFEWe4hgC/announcing-the-lesswrong-curated-podcast |
# Air Conditioner Test Results & Discussion
Background is in the [preregistration post](https://www.lesswrong.com/posts/AMmqk74zWmvP8tXEJ/preregistration-air-conditioner-test). This post will assume you’ve read that one.
First, the headline result: my main prediction was wrong. Paul’s predictions were also wrong. By ... | https://www.lesswrong.com/posts/5re4KgMoNXHFyLq8N/air-conditioner-test-results-and-discussion |
# AI Training Should Allow Opt-Out
Last year, GitHub [announced](https://github.blog/2021-06-29-introducing-github-copilot-ai-pair-programmer/) their [Copilot](https://github.com/features/copilot/) system, an AI assistant for developers based on OpenAI's [Codex](https://en.wikipedia.org/wiki/OpenAI_Codex) model, as a ... | https://www.lesswrong.com/posts/jZo6mapBzvFxbBQwE/ai-training-should-allow-opt-out |
# What’s the contingency plan if we get AGI tomorrow?
We’ve got lots of theoretical plans for alignment and AGI risk reduction, but what’s our current best bet if we know superintelligence will be created tomorrow? This may be too vague a question, so here’s a fictional scenario to make it more concrete (feel free to ... | https://www.lesswrong.com/posts/9daTCJAGtxrJKmafm/what-s-the-contingency-plan-if-we-get-agi-tomorrow |
# Covid 6/23/22: Under Five Alive
America has now approved the Covid vaccine for children as young as six months old. Should you vaccinate your very young child?
Yes, I believe that you should. I don’t think you _absolutely have to_ do that. Children that young are at minimal risk from Covid. However, vaccination dec... | https://www.lesswrong.com/posts/PEWGmv5rqJDDNeea4/covid-6-23-22-under-five-alive |
# Nonprofit Boards are Weird
> Note: anything in this post that you think is me subtweeting your organization is actually about, like, at least 3 organizations. (I'm currently on 4 boards in addition to [Open Philanthropy](https://openphilanthropy.org/)'s; I've served on a bunch of other boards in the past; and more t... | https://www.lesswrong.com/posts/nSjavaKcBrtNktzGa/nonprofit-boards-are-weird |
# Half-baked AI Safety ideas thread
*\[*[*Cross-posted*](https://forum.effectivealtruism.org/posts/fuEHtmY9gvQQtdawL/half-baked-ideas-thread-ea-ai-safety) *from the EA Forum. The EA Forum version of this post is for both half-baked EA ideas and half-baked AI Safety ideas, whereas this version of the post is for half-b... | https://www.lesswrong.com/posts/5WPwNvfYnh78QSxCA/half-baked-ai-safety-ideas-thread |
# [Link] OpenAI: Learning to Play Minecraft with Video PreTraining (VPT)
[https://openai.com/blog/vpt/](https://openai.com/blog/vpt/)
> We trained a neural network to play Minecraft by Video PreTraining (VPT) on a massive unlabeled video dataset of human Minecraft play, while using only a small amount of labeled cont... | https://www.lesswrong.com/posts/NcQN8LmjbbnJfdbGt/link-openai-learning-to-play-minecraft-with-video |
# Is CIRL a promising agenda?
**Richard Ngo writes**
> Since Stuart Russell's proposed alignment solution in Human Compatible is the most publicly-prominent alignment agenda, I should be more explicit about my belief that it almost entirely fails to address the core problems I expect on realistic pathways to AGI.
>... | https://www.lesswrong.com/posts/qdcfTWbeSvcBRgCjs/is-cirl-a-promising-agenda |
# 20 Critiques of AI Safety That I Found on Twitter
In no particular order, here's a collection of Twitter screenshots of people attacking AI Safety. A lot of them are poorly reasoned, and some of them are simply ad-hominem. Still, these types of tweets are influential, and are widely circulated among AI capabilities ... | https://www.lesswrong.com/posts/57zLLpeReWcGWd6hE/20-critiques-of-ai-safety-that-i-found-on-twitter |
# Book review: The Passenger by Lisa Lutz
_Spoiler warning: spoilers for The Passenger by Lisa Lutz, and In the Cart, by Anton Chekhov._
I took up [this book](https://www.goodreads.com/book/show/26154406-the-passenger) looking for a page-turner. It was, but hours before the end I thought its main contribution to my m... | https://www.lesswrong.com/posts/eyw3KuNaYEr4veamj/book-review-the-passenger-by-lisa-lutz |
# LessWrong Has Agree/Disagree Voting On All New Comment Threads
Starting today we're activating two-factor voting on all new comment threads.
Now there are two axes on which you can vote on comments: the standard karma axis remains on the left, and the new axis on the right lets you show much you agree or disagree ... | https://www.lesswrong.com/posts/HALKHS4pMbfghxsjD/lesswrong-has-agree-disagree-voting-on-all-new-comment |
# "Science Cathedrals"
My favorite place in the world is Chicago's [Museum of Science and Industry](https://en.wikipedia.org/wiki/Museum_of_Science_and_Industry_%28Chicago%29). When I was younger, I would constantly beg my parents to take me there. It was pretty far, and it's not exactly enjoyable to drive through Chi... | https://www.lesswrong.com/posts/ThzqDiEqdBHYpn9CQ/science-cathedrals |
# Intelligence in Commitment Races
*A distillation of my understanding of the commitment races problem.*
Greaser Courage
===============
It's 1950 and you're a greaser in a greaser gang. Recreationally, you're driving your heavy American cars (without seatbelts) at each other at speed, and seeing who swerves first. ... | https://www.lesswrong.com/posts/hDtrJrovmsiWn4cJj/intelligence-in-commitment-races |
# Feature request: voting buttons at the bottom?
This is a brief question/request regarding LessWrong and AlignmentForum UI and voting buttons.
## Uncontroversial (I expect) factual statements
When I read a post or comment which fills more than one screen, I want to consume the whole content before providing any rei... | https://www.lesswrong.com/posts/yxZMbjFYmhgJrHvQG/feature-request-voting-buttons-at-the-bottom |
# Worked Examples of Shapley Values
Three times in the past month, I've run across occasions where [Shapley values](https://en.wikipedia.org/wiki/Shapley_value) were mentioned or would have been useful. There are a couple of good explainers of Shapley value already on the internet, but what most of them lack is a bunc... | https://www.lesswrong.com/posts/3XqoK4C7r4tyGZ893/worked-examples-of-shapley-values |
# Raphaël Millière on Generalization and Scaling Maximalism
I interviewed Raphaël Millière, a Presidential Scholar in Society and Neuroscience at Columbia University about his critic of “[scaling maximalism](https://twitter.com/raphaelmilliere/status/1526568402600480768)” and his takes on AGI. This is part of an on-go... | https://www.lesswrong.com/posts/HzCFRvrHYFsCixWRs/raphael-milliere-on-generalization-and-scaling-maximalism |
# Do alignment concerns extend to powerful non-AI agents?
Right now the main focus of alignment seems to be on how to align powerful AGI agents, a.k.a AI Safety. I think the field can benefit from a small reframing: we should think not about aligning AI, but about alignment of systems in general, if we are not already... | https://www.lesswrong.com/posts/vFFxJkAfP6hm9Ykyf/do-alignment-concerns-extend-to-powerful-non-ai-agents |
# Updated Deference is not a strong argument against the utility uncertainty approach to alignment
**Thesis:** The [problem of fully updated deference](https://arbital.com/p/updated_deference/) is not a strong argument against the viability of the assistance games / utility uncertainty approach to AI (outer) alignment... | https://www.lesswrong.com/posts/ikYKHkffKNJvBygXG/updated-deference-is-not-a-strong-argument-against-the |
# [Link] Adversarially trained neural representations may already be as robust as
corresponding biological neural representations
Abstract:
> Visual systems of primates are the gold standard of robust perception. There is thus a general belief that mimicking the neural representations that underlie those systems will... | https://www.lesswrong.com/posts/oFXhiNS3gGFeKftEa/link-adversarially-trained-neural-representations-may |
# [Link] Childcare : what the science says
This is a pretty good summary of the research on daycare effectiveness. It is a Much More Than You Wanted To Know-style post from a science blogger [criticalscience](https://criticalscience.medium.com/) who has some interesting posts on Medium.
TL;DR:
* Adverse effects on... | https://www.lesswrong.com/posts/kivCEEpfJXHnJJBAT/link-childcare-what-the-science-says |
# Do you consider your current, non-superhuman self aligned with “humanity” already?
Any actual “hard” superhumans who may be listening are excluded from the audience of this question. I am currently refraining from describing my own ideas to avoid biasing the results; I'm not sure whether I'll reveal something later ... | https://www.lesswrong.com/posts/y7dvSzcZhYyk7THn9/do-you-consider-your-current-non-superhuman-self-aligned |
# Identification of Natural Modularity
### *Produced As Part Of The SERI ML Alignment Theory Scholars Program 2022 Under Guidance of* [*John Wentworth*](https://www.lesswrong.com/users/johnswentworth)
Summary
=======
The Natural Abstraction Hypothesis (NAH) proposes that common high-level abstractions utilised by... | https://www.lesswrong.com/posts/BqZDnv66x7FxGHqCL/identification-of-natural-modularity |
# Some reflections on the LW community after several months of active engagement
There seems to be some folks who might derive useful insights from a third-party, and mostly neutral, perspective of how the community appears after an honest and sustained effort of engagement, someone who doesn't really place AI risk as... | https://www.lesswrong.com/posts/tfqnyDPhK7v2KtzoB/some-reflections-on-the-lw-community-after-several-months-of |
# AI-Written Critiques Help Humans Notice Flaws
This is a linkpost for a recent paper from the OpenAI alignment team (disclaimer: I used to work with this team). They summarize their results as:
> We trained “critique-writing” models to describe flaws in summaries. Human evaluators find flaws in summaries much more o... | https://www.lesswrong.com/posts/AHBejZBsaTR6dkRHs/ai-written-critiques-help-humans-notice-flaws |
# Conversation with Eliezer: What do you want the system to do?
*This is a write-up of a conversation I overheard between Eliezer and some junior alignment researchers. Eliezer reviewed this and gave me permission to post this, but he mentioned that "there's a lot of stuff that didn't get captured well or accurately."... | https://www.lesswrong.com/posts/d4YGxMpzmvxknHfbe/conversation-with-eliezer-what-do-you-want-the-system-to-do |
# Fundamental Uncertainty: Chapter 3 - Why don't we agree on what's right?
*N.B. This is a chapter in a book about truth and knowledge. It is the first draft. I have since revised it. You can find the most up-to-date info/version on the* [*book's website*](https://www.fundamentaluncertainty.com/)*.*
You're walking do... | https://www.lesswrong.com/posts/mGYSFbrHKqsSce3vu/fundamental-uncertainty-chapter-3-why-don-t-we-agree-on-what |
# Conditioning Generative Models
*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 con... | https://www.lesswrong.com/posts/nXeLPcT9uhfG3TMPS/conditioning-generative-models |
# Child Contracting
The kids came to me with a contract dispute: they disagreed about what the deal had been. After some talking, and hearing what they each thought they'd agreed to, it turned out that the contract wasn't just metaphorical:
[](https://www.jefftk.com... | https://www.lesswrong.com/posts/pbSPQZ3YtxAdPyjDt/child-contracting |
# Formalizing Deception
*An attempt at formalizing deception, seeing what goes right, what goes wrong, and what compromises have to be made.*
Introduction
============
In a police interrogation one of the key challenges is working out whether a suspect is being deceptive. In a game of poker, one of the main goals is... | https://www.lesswrong.com/posts/gLLDika2pyKX6Afpe/formalizing-deception-1 |
# Seven ways to become unstoppably agentic
*Edit in September 2022: I wrote a response to this post* [*here*](https://forum.effectivealtruism.org/posts/acyfmFTN3cNgwnYw6/agency-needs-nuance)*, in which I lay out some concerns, some unexpected costs of taking "agentic" actions for me (and mistakes I’ve made), and thing... | https://www.lesswrong.com/posts/Tnpp3cyEHMGthjGAf/seven-ways-to-become-unstoppably-agentic |
# My current take on Internal Family Systems “parts”
I was recently asked how literal/metaphorical I consider the [Internal Family Systems model](https://www.lesswrong.com/posts/5gfqG3Xcopscta3st/building-up-to-an-internal-family-systems-model) of your mind being divided into “parts” that are kinda like subpersonaliti... | https://www.lesswrong.com/posts/YXBpBCNC66daaofoY/my-current-take-on-internal-family-systems-parts |
# Training Trace Priors and Speed Priors
*Thanks to Evan Hubinger for suggesting this idea.*
[Training Trace Priors](https://www.alignmentforum.org/posts/hjrqXjEpaw9ogScPh/training-trace-priors) are priors over boolean circuits which examine the outputs of gates on samples from the training distribution, typically fo... | https://www.lesswrong.com/posts/9o2mjdo7eb7677EJS/training-trace-priors-and-speed-priors |
# Robin Hanson asks "Why Not Wait On AI Risk?"
**UPDATE: This is an accidental duplicate. Please go to the first posting** [**here**](https://www.lesswrong.com/posts/H2N6eWw8JgxwKHPM3/linkpost-robin-hanson-why-not-wait-on-ai-risk.)**.**
Since the [AI-Foom Debate](https://www.lesswrong.com/tag/the-hanson-yudkowsky-ai-... | https://www.lesswrong.com/posts/7QKD4ypxAXwcwp634/robin-hanson-asks-why-not-wait-on-ai-risk |
# Contest: An Alien Message
### Scenario
You're not certain that aliens are real. Even after receiving that email from your friend who works at NASA, with the subject line "What do you make of this?" and a **[single binary file attached](https://github.com/DaemonicSigil/an-alien-message)**, you're still not certain. ... | https://www.lesswrong.com/posts/dFFdAdwnoKmHGGksW/contest-an-alien-message |
# Air Conditioner Repair
A thing happened to me, seemed worth sharing.
What Happened
-------------
A few weeks ago, the in-wall air conditioner in our bedroom ceased to function.
Troy Barnes was unavailable, so instead we talked to the super, he recommended a company called Amhac, we ignored normative determinism (... | https://www.lesswrong.com/posts/xtH8sNHTkKNXdQ4wP/air-conditioner-repair |
# Announcing Epoch: A research organization investigating the road to Transformative AI
Summary
=======
* We are a new research organization working on **investigating trends in Machine Learning** and **forecasting the development of Transformative Artificial Intelligence**
* This work is done in close collaborat... | https://www.lesswrong.com/posts/AJ6GHm5n6fBRJbMhq/announcing-epoch-a-research-organization-investigating-the |
# Announcing the Inverse Scaling Prize ($250k Prize Pool)
*TL;DR*: We’re launching the [Inverse Scaling Prize](https://twitter.com/EthanJPerez/status/1541454949397041154): a contest with $250k in prizes for finding zero/few-shot text tasks where larger language models show increasingly undesirable behavior (“inverse s... | https://www.lesswrong.com/posts/eqxqgFxymP8hXDTt5/announcing-the-inverse-scaling-prize-usd250k-prize-pool |
# Custom iPhone Widget to Encourage Less Wrong Use
**TLDR**: I made a little custom iPhone widget for the EA Forum recently and have been getting a lot of value out of it. It takes an extra 10s to make this a LessWrong or Alignment forum widget. At the moment I don't feel bothered to make this into an app which is eas... | https://www.lesswrong.com/posts/gmaG9EFJxSCf2XdRb/custom-iphone-widget-to-encourage-less-wrong-use |
# Are long-form dating profiles productive?
There's been a noticeable trend over the last few years of in this community of writing long-form dating profiles. They have the benefit of being slightly costly signals, requiring significantly more time and attention and a much greater degree of disclosure than can be devo... | https://www.lesswrong.com/posts/axmxy7kiqwcRPSGXH/are-long-form-dating-profiles-productive |
# Deliberation, Reactions, and Control: Tentative Definitions and a Restatement of Instrumental Convergence
_This analysis is speculative. The framing has been refined in conversation and private reflection and research. To some extent it feels vacuous, but at least valuable for further research and communication._
A... | https://www.lesswrong.com/posts/u4BaLRK6mJJcvycEk/deliberation-reactions-and-control-tentative-definitions-and |
# Deliberation Everywhere: Simple Examples
_The analysis and definitions used here are tentative. My familiarity with the concrete systems discussed ranges from rough understanding (markets and parliaments), through abiding amateur interest (biology), to meaningful professional expertise (AI/ML things). The abstractio... | https://www.lesswrong.com/posts/5GzqD7fgtxjepPERp/deliberation-everywhere-simple-examples |
# Epistemic modesty and how I think about AI risk
*\[This is a direct cross-post from a* [*Facebook post of mine*](https://www.facebook.com/permalink.php?story_fbid=pfbid027vXyKvMsYJ3N5fkzN2iHyPKJWnD14d5A9iwQtihbjNgtYUkVNfXSW4LnnFEkhG3kl&id=100010608396052) *and was originally intended for the people who read what I w... | https://www.lesswrong.com/posts/wx2hQskeDuKvM2Yd8/epistemic-modesty-and-how-i-think-about-ai-risk |
# Exploring Mild Behaviour in Embedded Agents
*Thanks to Tristan Cook, Emery Cooper, Nicolas Macé and Will Payne for discussions, advice, and comments. All views are my own.*
***I think this post is best viewed as an effort to point at a potentially interesting research direction. The specific arguments I make center... | https://www.lesswrong.com/posts/vNS4vGKGD6ygyz5ok/exploring-mild-behaviour-in-embedded-agents |
# [Yann Lecun] A Path Towards Autonomous Machine Intelligence
# Abstract
> How could machines learn as efficiently as humans and animals? How could machines learn to reason and plan? How could machines learn representations of percepts and action plans at multiple levels of abstraction, enabling them to reason, pre... | https://www.lesswrong.com/posts/Y7XkGQXwHWkHHZvbm/yann-lecun-a-path-towards-autonomous-machine-intelligence-1 |
# Limits of Bodily Autonomy
Bodily autonomy is a strong consideration in favor of both abortion and optional vaccination. Because it is not the only consideration, however, many of us are in favor of abortion for anyone who wants one while also seeing some cases in which we favor mandatory vaccination. I've recently s... | https://www.lesswrong.com/posts/5nEcTW3zHDqdKxWEd/limits-of-bodily-autonomy |
# Is there any way someone could post about public policy relating to abortion access (or another sensitive subject) on LessWrong without getting super downvoted?
I'm aware there are some sensitive subjects that have been debated to death among rationalists, if not on LessWrong, than elsewhere online, such that it wou... | https://www.lesswrong.com/posts/HhWXDoTFjuCtxfibv/is-there-any-way-someone-could-post-about-public-policy |
# Assessing AlephAlphas Multimodal Model
The currently probably most impressive large language model imho is [Flamingo](https://www.deepmind.com/blog/tackling-multiple-tasks-with-a-single-visual-language-model), which has the ability to answer questions about images. Flamingo is not accessible to the general public, h... | https://www.lesswrong.com/posts/EzuBSASuui5qekhLA/assessing-alephalphas-multimodal-model |
# What Diet Books Don't Teach: A book review and a request for more reading
I’ve read a fair number of self-help books in my life. I’m long past the point where I expect a book to change my life, but they often are fun to pleasure-read. [As Scott Alexander notes,](https://slatestarcodex.com/2019/11/20/book-review-all-... | https://www.lesswrong.com/posts/ZPTw9cMhYFNLuHpA3/what-diet-books-don-t-teach-a-book-review-and-a-request-for |
# Low-Friction MBTA Predictions
This has not been a good week for the MBTA:
* The Red, Orange, and Blue lines are running [Saturday schedules on weekdays](https://www.mbta.com/news/2022-06-17/subway-service-changes-take-effect-monday-june-20) ([news coverage](https://www.wbur.org/news/2022/06/17/mbta-to-cut-weekday... | https://www.lesswrong.com/posts/rEvBzk8ZpZdYDtTEx/low-friction-mbta-predictions |
# Four reasons I find AI safety emotionally compelling
‘I don’t feel emotionally motivated to work on AI safety, even though I’m intellectually convinced that it’s important.’
It always surprises me when people say this because I find my work at [Nonlinear](https://www.nonlinear.org/) on AI safety *incredibly* motiva... | https://www.lesswrong.com/posts/mffMfysE4SntoARkj/four-reasons-i-find-ai-safety-emotionally-compelling |
# What success looks like
**TL;DR:** We wrote a post on possible success stories of a transition to TAI to better understand which factors causally reduce the risk of AI risk. Furthermore, we separately explain these catalysts for success in more detail and this post can thus be thought of as a high-level overview of ... | https://www.lesswrong.com/posts/aKpqwtZN6ifAhqJYK/what-success-looks-like |
# A physicist's approach to Origins of Life
*Note: this is cross-posted from the* [*post here*](https://www.pchvykov.com/post/a-physicist-approach-to-origins-of-life)*, where I had a few videos that I could not add in here - see the original post for the full experience*
I spent the 6 years in my Ph.D. thinking of th... | https://www.lesswrong.com/posts/GvmGmfzDcqLYzEE9C/a-physicist-s-approach-to-origins-of-life |
# Scott Aaronson and Steven Pinker Debate AI Scaling
> Before June 2022 was the month of the possible start of the Second American Civil War, it was the month of [a](https://astralcodexten.substack.com/p/my-bet-ai-size-solves-flubs?s=w) [lively](https://garymarcus.substack.com/p/what-does-it-mean-when-an-ai-fails) [de... | https://www.lesswrong.com/posts/qyeHe7aQa4HkJmHqE/scott-aaronson-and-steven-pinker-debate-ai-scaling |
# Units of Exchange
**Epistemic status:** Established and confirmed
*The lessons taught in Units of Exchange are straightforward applications of extremely well-established principles from economics and sociology, such as supply and demand, Pareto curves, value of information, the sunk cost fallacy, and arbitrage. The... | https://www.lesswrong.com/posts/cvDtmPNCyrkpg4d4F/units-of-exchange |
# CFAR Handbook: Introduction
The [Center for Applied Rationality](rationality.org) is a Bay Area non-profit that, among other things, ran lots of workshops to offer people tools and techniques for solving problems and improving their thinking. Those workshops were accompanied by a reference handbook, which has been a... | https://www.lesswrong.com/posts/dbDHEQyKqnMDDqq2G/cfar-handbook-introduction |
# What Are You Tracking In Your Head?
A large chunk - plausibly the majority - of real-world expertise seems to be in the form of *illegible skills*: skills/knowledge which are hard to transmit by direct explanation. They’re not necessarily things which a teacher would even notice enough to consider important - just ... | https://www.lesswrong.com/posts/bhLxWTkRc8GXunFcB/what-are-you-tracking-in-your-head |
# Reflections on Living in "Guess Culture"
*This is cross-posted from* [*my blog*](https://www.daltonmabery.com/essays/inconvenienced-guesser)*.*
* * *
There's been two articles I've seen recently on the difference between people who ask for what they want (ask culture) vs. people who drop subtle hints at what they ... | https://www.lesswrong.com/posts/MKfxwEccGAGgWoWXp/reflections-on-living-in-guess-culture |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.