id stringlengths 36 36 | source stringclasses 15
values | formatted_source stringclasses 13
values | text stringlengths 2 7.55M |
|---|---|---|---|
20e0ae76-52b3-4cfe-98d5-b8af29fefd13 | trentmkelly/LessWrong-43k | LessWrong | Need help searching for a quote
Hey,
I'm trying to find a quote that I recall seeing in the sequences. It was a post by Eliezer, and near the end he made a statement to the effect that 'We don't always know what is right, but you should never physically harm someone in the course of a dispute. Argumentation is an acc... |
f577f0bd-0a6c-4cee-99f6-8600c64d733a | trentmkelly/LessWrong-43k | LessWrong | Why are young, healthy people eager to take the Covid-19 vaccine?
Clearly a lot of people on LW want to take it ASAP. I strongly don't want that - to the point where I will most likely emigrate if it becomes obligatory in my country. Please help me understand what I'm missing. Here is my understanding:
* As a young,... |
9e865c2f-4b9b-465c-b6b1-24c1231424b6 | StampyAI/alignment-research-dataset/special_docs | Other | The vulnerable world hypothesis
The Vulnerable World Hypothesis
Nick Bostrom
Future of Humanity Institute, University of Oxford
Abstract
Scienti fic and technological progress might change people ’s capabilities or incentives in ways that would destabilize civiliza-
tion. For example, advances in DIY biohacking tools m... |
5adc752a-4f9c-419d-8d97-bc1a03ef785b | StampyAI/alignment-research-dataset/arxiv | Arxiv | Learning to Resolve Conflicts for Multi-Agent Path Finding with Conflict-Based Search
1 Introduction
---------------
Multi-Agent Path Finding (MAPF) is the problem of finding a set of conflict-free paths for a given number of agents on a given graph that minimizes the sum of costs or the makespan. Although MAPF is... |
484f7e03-84cb-4c72-aaa6-6ae8426c84a8 | trentmkelly/LessWrong-43k | LessWrong | Slightly known unknowns
Inspired by What visionary project would you fund?, I'm wondering about whether there are known blank areas in our knowledge which might turn up surprising knowledge.
Once upon a time, the sun was a mystery. People had a pretty good idea of its mass, and how much chemical energy would be neede... |
9a7f0df0-9ad3-48a4-b138-b30e347bf3d9 | trentmkelly/LessWrong-43k | LessWrong | When training AI, we should escalate the frequency of capability tests
Inspired by ideas from Lucius Bushnaq, David Manheim, Gavin Leech, Davidad. Any errors are mine.
Cross-posted from my Substack.
Compute, algorithms, and data form the AI triad—the main inputs for better AI.[1] AI uses compute to run algorithms th... |
b4ef79e7-4afd-47ae-b93c-67e0f83acfa5 | trentmkelly/LessWrong-43k | LessWrong | Guarded learning
A putative new idea for AI control; index here.
"Guarded learning" is a model for unbiased learning, the kind of learning where the AI has an incentive to learn its values, but not to bias the direction in which it learns.
----------------------------------------
The value-uncertain U
Assume the A... |
cb6f0463-b8c0-4623-9eb7-60adea888045 | trentmkelly/LessWrong-43k | LessWrong | Beware of Stephen J. Gould
Followup to: Natural Selection's Speed Limit and Complexity Bound
If you've read anything Stephen J. Gould has ever said about evolutionary biology, I have some bad news for you. In the field of evolutionary biology at large, Gould's reputation is mud. Not because he was wrong. Many hon... |
d2a54edd-0fd4-4df8-b66b-731fc9a95a39 | trentmkelly/LessWrong-43k | LessWrong | About Rationality Quotes
...so do these go to the main LW now, or do we keep them in Discussion?
My vote would be "to the main LW", but since I both want to discuss this and would like to play with the "feature" that allows moving a thread from one to the other at some point:
Discuss. |
1abdcfe4-bd38-4ee0-9b50-7135045fd7ca | trentmkelly/LessWrong-43k | LessWrong | The case for becoming a black-box investigator of language models
Interpretability research is sometimes described as neuroscience for ML models. Neuroscience is one approach to understanding how human brains work. But empirical psychology research is another approach. I think more people should engage in the analogou... |
bfdc88a4-4f07-4b5a-ad25-4cd423075ef3 | trentmkelly/LessWrong-43k | LessWrong | Use the Nato Alphabet
I plan to extend this into a longer post, but for now I want to tell a story.
I have a name that is pretty difficult for English speakers to just "know" the spelling of from the pronounciation.
Recently I started using the Nato alphabet. It really greatly reduces the friction of e.g. communicat... |
32838cca-82e2-427c-aa2e-a92993eb2b70 | trentmkelly/LessWrong-43k | LessWrong | What could be done with RNA and DNA sequencing that's 1000x cheaper than it's now?
In How to Invent the Future Alan Kay proposes that a good way of inventing the feature is to take an exponential and project it into the feature. Afterwards it's useful to ask what can be done with the new capabilities.
On important ex... |
c3cf4ce9-60bd-4ef9-88ea-9f1fb5cea143 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | 14 Ways ML Could Improve Informative Video
*~30-minute brainstorm. I haven’t done ML engineering myself, but am an enthusiast.*
ML systems are getting scarily good. I’m a fan of [online video for sharing information](https://quri.substack.com/p/why-does-academiaea-produce-so-few). Video is tougher to automate than te... |
ba26c18c-b20c-4f6e-9b50-e1679a99fcd5 | StampyAI/alignment-research-dataset/arbital | Arbital | Bayes' rule: Vector form
%todo: This page conflates two concepts: (1) You can perform a Bayesian update on multiple hypotheses at once, by representing hypotheses via vectors; and (2) you can perform multiple Bayesian updates by multiplying by all the likelihood functions (and only normalizing once at the end). We sho... |
3a14fd1e-1a3f-4b6d-a9e1-46551cc9adbb | trentmkelly/LessWrong-43k | LessWrong | Tips/tricks/notes on optimizing investments
I've been optimizing various aspects of my investment setup recently, and will write up some tips and tricks that I've found in the form of "answers" here. Others are welcome to share their own here if they'd like. (Disclaimer: I’m not a lawyer, accountant, or investment adv... |
c0455c0c-97a6-4e92-b47a-9668eb5a679e | StampyAI/alignment-research-dataset/blogs | Blogs | Announcing a new colloquium series and fellows program
The Machine Intelligence Research Institute is accepting applicants to two summer programs: a three-week AI robustness and reliability colloquium series (co-run with the Oxford [Future of Humanity Institute](https://www.fhi.ox.ac.uk/)), and a two-week fellows prog... |
6dc2f152-b72d-48d5-ad48-9440701c45f0 | trentmkelly/LessWrong-43k | LessWrong | 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 laun... |
eff55943-6837-45c9-a235-12c4606bc22e | trentmkelly/LessWrong-43k | LessWrong | How much does consumption affect production?
A ewe for a ewe
In a discussion with Benquo over his recent suffering-per-calorie estimates I learned that there have been a few different proponents of incorporating short term elasticities into such estimates. But do empirical short term elasticities really improve our es... |
7737dc62-673d-43e0-96a0-01d630bf1cb8 | trentmkelly/LessWrong-43k | LessWrong | Some quick thoughts on "AI is easy to control"
There are many things I feel like the post authors miss, and I want to share a few thoughts that seem good to communicate.
I'm going to focus on controlling superintelligent AI systems: systems powerful enough to solve alignment (in the CEV sense) completely, or to kill ... |
ed7d6d73-09ac-4240-95a4-6fa65e421c77 | trentmkelly/LessWrong-43k | LessWrong | Yet Another Covid Risk Calculator
https://thomaswc.com/crisk.html
There are many covid risk calculators online. Some express your risk in arbitrary units, some give you your probability of getting infected, and some tell you how likely you are to die if you get infected. Mine calculates both of those last two thing... |
31130b24-ae36-4ad6-aeda-6c092dd2dd64 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Washington, D.C.: Value of Information
Discussion article for the meetup : Washington, D.C.: Value of Information
WHEN: 24 May 2015 03:00:00PM (-0400)
WHERE: National Portrait Gallery
We'll be congregating at the courtyard of the National Portrait Gallery from 3:00 - 3:30 until closing, to talk about value... |
9f53a571-a9d6-406b-8955-a01a39a7eb87 | trentmkelly/LessWrong-43k | LessWrong | Unsolved Problems in Philosophy Part 1: The Liar's Paradox
Graham Priest discusses The Liar's Paradox for a NY Times blog. It seems that one way of solving the Liar's Paradox is defining dialethei, a true contradiction. Less Wrong, can you do what modern philosophers have failed to do and solve or successfully dissolv... |
613f54e1-5a99-46e7-801f-afd020255bd1 | trentmkelly/LessWrong-43k | LessWrong | Simplified Anthropic Doomsday
Here is a simplified version of the Doomsday argument in Anthropic decision theory, to get easier intuitions.
Assume a single agent A exists, an average utilitarian, with utility linear in money. Their species survives with 50% probability; denote this event by S. If the species survives... |
84dce84c-3f91-4f80-bf4a-fff54a03bc85 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | On how various plans miss the hard bits of the alignment challenge
*This post has been recorded as part of the LessWrong Curated Podcast, and can be listened to on* [*Spotify*](https://open.spotify.com/episode/6cDwctrmpWW6UNEKTT3Vvf?si=7T7SaHNCQ9axzL4D-crPQw)*,* [*Apple Podcasts*](https://podcasts.apple.com/us/podcast... |
1efe020a-9061-43c1-9193-da00c5ce8dfa | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | AGISF adaptation for in-person groups
This past semester,[HAIST](https://haist.ai/) and[MAIA](https://www.mitalignment.org/) (the Harvard and MIT AI safety student groups) ran an adapted version of Richard Ngo's [AGI Safety Fundamentals alignment](https://www.agisafetyfundamentals.com/ai-alignment-curriculum) curricul... |
c175c3b8-2b6d-40ac-b03a-b0c03cfecbfc | trentmkelly/LessWrong-43k | LessWrong | [Cross-post] Is the Fermi Paradox due to the Flaw of Averages?
[This article is copy-pasted from the Lumina blog, very lightly edited for LessWrong.]
Where is everybody?
— Enrico Fermi
The omnipresence of uncertainty is part of why making predictions and decisions is so hard. We at Lumina advocate treating uncertai... |
ef427aab-9f8d-439c-a329-e60b380c69cb | trentmkelly/LessWrong-43k | LessWrong | What are intuitive ways for presenting certainty/confidence in continuous variable inferences (i.e. numerical predictions)?
I enjoy the idea of thinking in probabilities, and also, as part of my job, I have to deal with the hellish task of figuring out how to present probabilities to people in a simple way.
It seems ... |
63f7fb35-623a-4966-b487-dd518a73a2f3 | StampyAI/alignment-research-dataset/arbital | Arbital | Relevant limited AI
It is an open problem to propose a [limited AI](https://arbital.com/p/5b3) that would be [relevant](https://arbital.com/p/2s) to the [value achievement dilemma](https://arbital.com/p/2z) - an agent cognitively constrained along some dimensions that render it much safer, but still able to perform so... |
67e429e5-b301-4252-bb16-8e6a9343ec2b | trentmkelly/LessWrong-43k | LessWrong | Would a Misaligned SSI Really Kill Us All?
Preamble
Honest question.
* Some clarification on terms:
SSI: Strongly Superhuman Intelligence
* Misaligned: for the purposes of this question an AI is misaligned if it is indifferent to human welfare and wellbeing.
* An SSI that was actively malicious would also be... |
be699509-c4e8-4cbb-9b90-ed0ff77c534d | trentmkelly/LessWrong-43k | LessWrong | Total Nano Domination
Followup to: Engelbart: Insufficiently Recursive
The computer revolution had cascades and insights aplenty. Computer tools are routinely used to create tools, from using a C compiler to write a Python interpreter, to using theorem-proving software to help design computer chips. I would not ye... |
725878ec-306b-4642-9d7d-9c31f8352faf | StampyAI/alignment-research-dataset/lesswrong | LessWrong | How is reinforcement learning possible in non-sentient agents?
*(Probably a stupid nooby question that won't help solve alignment)*
**Suppose you implement a goal in an AI through a reinforcement learning system. Why does the AI really "care" about this goal? Why does it obey? It does because it is punished and/o... |
95edc5b1-443a-40f4-ade4-a4320782fa54 | trentmkelly/LessWrong-43k | LessWrong | Meetup : (Improv?) games meetup
Discussion article for the meetup : (Improv?) games meetup
WHEN: 28 July 2013 03:00:00PM (-0400)
WHERE: National Portrait Gallery courtyard, Washington, DC 20001, USA
We'll be meeting to do improv games! However, because I will not actually be there and I'm not sure anyone (including... |
5affcf7d-c714-4f4e-b442-af3ea229889e | trentmkelly/LessWrong-43k | LessWrong | An Anchoring Experiment
The experiment is closed, for the results look here.
In recent discussion I have expressed an opinion that anchoring may, for some quantitative questions, cause the answer to lie further away from the correct value than the anchor itself. For concreteness, let's suppose that the correct value ... |
329b509e-ca52-4f28-8f1a-3f792896fdcf | trentmkelly/LessWrong-43k | LessWrong | [POLL] Do You Feel Oppressed?
At Reason Rally a couple of months ago, we noticed that a lot of atheists there seemed to be there for mutual support - because their own communities rejected atheists, because they felt outnumbered and threatened by their peers; the rally was a way for them to feel part of an in-group. ... |
67471ddd-f410-4abb-9eb3-bb0bf916f2f3 | trentmkelly/LessWrong-43k | LessWrong | Entrepreneurship ETG Might Be Better Than 80k Thought
Summary
1. In 2014, Ryan Carey at 80k estimated that YCombinator-backed startup founders averaged $2.5M/year
2. I repeat his analysis, and find that this number is now substantially higher: $3.8-9.9M/year
3. When this amount is discounted by 12%/year (average S&... |
84da061f-fd0b-44c2-bc59-903e71686043 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Less Wrong Israel Meetup: How Modern Math Is Different
Discussion article for the meetup : Less Wrong Israel Meetup: How Modern Math Is Different
WHEN: 28 April 2015 07:00:00PM (+0300)
WHERE: Google, Electra Tower, 98 Yigal Alon Street, Tel Aviv, Israel
On Tuesday, April 28, Anatoly Vorobey will be speakin... |
4b36102b-0772-4600-aef0-3c9ec3c033ee | trentmkelly/LessWrong-43k | LessWrong | Common failure modes in habit formation
> In one project, 256 members of a health-insurance plan were invited to classes stressing the importance of exercise. Half the participants received an extra lesson on the theories of habit formation (the structure of the habit loop) and were asked to identify cues and rewards ... |
4746a78d-9132-4a99-a18e-107956a506ad | trentmkelly/LessWrong-43k | LessWrong | The Devil You Know
A short story set in the EVE Universe.
The air was thick with smoke and shit. It smelled like oil and rot, rust and decay and desperation. Between the market and the river lay the slums. Between the tank farms and the market stalls rose a shantytown of ill fortunes and shattered dreams. Young peop... |
f948686e-db7a-4eb2-b36a-f5f41e604b9d | StampyAI/alignment-research-dataset/lesswrong | LessWrong | AI Summer Harvest
Trending metaphor
=================
I've noticed that AI Notkilleveryonists have begun appealing to a new(?) metaphor of an **AI Summer Harvest.**
Here's the FLI open letter *Pause Giant AI Experiments:*
> Humanity can enjoy a flourishing future with AI. Having succeeded in creating powerful AI s... |
36e0dc2c-c945-4919-892d-c60f4c802752 | trentmkelly/LessWrong-43k | LessWrong | Forecasting Newsletter: February 2022
Highlights
* The FTX foundation will potentially give out millions in forecasting
* Insight Predictions and Futuur have real-money prediction markets on the invasion of Ukraine
* At least $444k paid out in over the counter bets on the invasion of Ukraine
Index
* Thoughts on ... |
552a8662-cc95-45e8-9f79-e6585730d052 | trentmkelly/LessWrong-43k | LessWrong | ChatGPT and Bing Chat can't play Botticelli
Botticelli is a guessing game similar to Twenty Questions. It's a simple game that I sometimes play with my 7 year-old daughter. I tried to get ChatGPT and Bing Chat to play Botticelli with me, but neither tool could do it.
How to play Botticelli
Description of the game
... |
00008357-f694-4923-8641-971011f67b54 | trentmkelly/LessWrong-43k | LessWrong | What should be reified?
I've said elsewhere, I think the sticking point of a lot of important questions is just this question: what should be reified? Another way to say this is: what are our ontological and axiological commitments? This is basically a cold take in many of my circles, but certainly not elsewhere, and ... |
df3fcc0a-112a-4367-9b1f-9e91a3f1bcc0 | trentmkelly/LessWrong-43k | LessWrong | link to a Feynman video (and a 2nd (distantly related) video)
Video presentation from 1963 at Cornell. Character of Physical Law #2 The Relation of Mathematics to Physics.
This may be the best one of all the Feynman videos. He explains how physicists do mathematics and distinguishes how physicists do math from how ma... |
7ebaa73c-ac3c-42ca-9a02-8ddcbf3b5015 | trentmkelly/LessWrong-43k | LessWrong | Assume long serving politicians are rationally maximizing their careers
Imagine you had no idea how to ride a bull, and you encountered a bull rider. The rider is swinging his free arm and body wildly, at apparent random. Surely hugging the bulls back and moving less is better. You would immediately think the man is c... |
eeef106b-d73f-4576-933a-306e3d411b4a | trentmkelly/LessWrong-43k | LessWrong | A way of specifying utility functions for UDT
The original UDT post defined an agent's preferences as a list of programs it cares about, and a utility function over their possible execution histories. That's a good and general formulation, but how do we make a UDT agent care about our real world, seeing as we don't kn... |
0d22ff2a-ff85-4d46-84b4-b2cf7e4c62dd | trentmkelly/LessWrong-43k | LessWrong | Is Cryonics Possible for Theists?
I expect to have a conversation soon with my parents about cryonics. My parents are both professional scientists, but are vaguely religious. They certainly aren't fundamentalists; they are pro-choice, pro-gay rights, think "intelligent design" is nuts, etc. My father is less religious... |
f8f754d8-b0fd-4977-b287-9d1e7c35c6b0 | trentmkelly/LessWrong-43k | LessWrong | The lost millennium
Note: This post is speculative, and you should take the claims in it with a grain of salt.
Throughout human history, we know that economic growth, for which population growth was a good proxy under Malthusian conditions, has accelerated many times. Indeed, we can see this just by extrapolating cur... |
9375093e-fe02-424b-b74c-e40f4bef1d5e | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Miriam Yevick on why both symbols and networks are necessary for artificial minds
Miriam Yevick was a mathematician who [corresponded with physicist David Bohm in the 1950s](https://www.google.com/books/edition/David_Bohm_Causality_and_Chance_Letters/wWCwDgAAQBAJ?hl=en&gbpv=1&dq=Miriam+Yevick&pg=PA2&printsec=frontcove... |
f61e9ea0-3de8-48c7-86a3-4d3e1f16d235 | StampyAI/alignment-research-dataset/arbital | Arbital | Function: Physical metaphor
Many functions can be visualized as physical mechanisms of wheels and gears, that take their inputs on conveyor belts, manipulate them (using mechanical sensors and tools), and produce an output which is placed on an outgoing conveyor belt.
![A binary function](http://i.imgur.com/lKxAXcK.p... |
d632be75-5e55-4e3d-91c3-0ccd1f65245a | trentmkelly/LessWrong-43k | LessWrong | Subjective Anticipation and Death
tldr; It is incoherent to talk about a "you" which stretches through time. Instead, we should think of a series of similar mind-moments.
Once upon a time, there was a little boy, who answered to the name Lucas Sloan and was scared of dying. I too answer to the name Lucas Sloan, and ... |
5a3c3514-690f-449d-afaa-b118480cd5ef | trentmkelly/LessWrong-43k | LessWrong | A Master-Slave Model of Human Preferences
[This post is an expansion of my previous open thread comment, and largely inspired by Robin Hanson's writings.]
In this post, I'll describe a simple agent, a toy model, whose preferences have some human-like features, as a test for those who propose to "extract" or "extrapol... |
320065de-2cac-418f-ab9e-804d4e92088c | trentmkelly/LessWrong-43k | LessWrong | Test of the Bene Gesserit
Jessica didn't say what she felt, only what her son needed to hear.
"Paul…," Jessica said it with love, "You are going to die. But remember you are a duke's son. Do not dishonor Leto with your passing." Jessica whirled and strode from the room with a swish of her skirt. The door closed with ... |
c9a8175b-7ac7-4f97-bc5c-ad696e63f514 | trentmkelly/LessWrong-43k | LessWrong | The Pyramid And The Garden
I.
A recent breakthrough in pseudoscience: the location of the Great Pyramid of Giza encodes the speed of light to seven decimal places.
This is actually true. The speed of light in a vacuum is 299,792,458 meters per second. The coordinates of of the Great Pyramid are 29.9792458° N, 31.134... |
c08a56f6-b012-4d42-bba8-72a129605e67 | trentmkelly/LessWrong-43k | LessWrong | Detecting Genetically Engineered Viruses With Metagenomic Sequencing
This represents work from several people at the NAO. Thanks especially to Dan Rice for implementing the duplicate junction detection, and to @Will Bradshaw and @mike_mclaren for editorial feedback.
Summary
If someone were to intentionally cause a ... |
c4fef993-32c6-4192-a86e-0395df1e6bd0 | trentmkelly/LessWrong-43k | LessWrong | The Peril of the Great Leaks (written with ChatGPT)
Context: I had an idea of a thought I wanted to spread but decided to make it via chatgpt. The full prompt can be found in the comments below.
In less than three years, the world will be a very different place. Thanks to the rise of large language models (LLMs), hac... |
668e64a2-01fc-41f5-9d0c-56b6604b212b | trentmkelly/LessWrong-43k | LessWrong | Current safety training techniques do not fully transfer to the agent setting
TL;DR: We are presenting three recent papers which all share a similar finding, i.e. the safety training techniques for chat models don’t transfer well from chat models to the agents built from them. In other words, models won’t tell you how... |
214d4ebd-c9c2-4d6e-8555-03088966c9c4 | trentmkelly/LessWrong-43k | LessWrong | [LINK] Meteorologists are Epistemically Rational
Nate Silver of the New York Times's political prediction blog fivethirtyeight has posted an excerpt of his upcoming book on predictions in various disciplines, The Signal and the Noise. The excerpt describes how meteorologists, in contrast to prognosticators in other d... |
5cc1c7b3-ceef-4c8b-b3b7-75706c3283df | trentmkelly/LessWrong-43k | LessWrong | Meetup : Zagreb Meetup
Discussion article for the meetup : Zagreb Meetup
WHEN: 07 September 2013 05:00:00PM (+0200)
WHERE: Stupnička 14, Zagreb, Croatia
Meetup will start with open discussion, but we have prepared some exercises and games to play :)
Discussion article for the meetup : Zagreb Meetup |
56e81166-7ee1-4538-be14-5966f43ea7bf | trentmkelly/LessWrong-43k | LessWrong | Is valuing life undervaluing it?
People often object to human life having a value placed on it, explicitly or implicitly. (I’m told there are good reasons, apparently to do with dignity, compassion, holism, souls and me being sick and inhuman, but I must admit they seem incoherent to me – if anyone would like to expla... |
45f0c6e1-40ee-4ae9-99f3-caf4325034e9 | trentmkelly/LessWrong-43k | LessWrong | Still Not in Charge
Epistemic Status: Speed premium, will hopefully flesh out more carefully in future
Previously: Why I Am Not in Charge
A brief update about my exchange with Scott from this past week.
After Scott Alexander wrote a post about why WebMD is terrible and it would be good but impossible to put me in c... |
84039186-d872-4edc-bc7e-4d4c4cf7a159 | StampyAI/alignment-research-dataset/blogs | Blogs | AI alignment and wolfram physics
AI alignment and wolfram physics
--------------------------------
[wolfram physics](https://writings.stephenwolfram.com/2020/04/finally-we-may-have-a-path-to-the-fundamental-theory-of-physics-and-its-beautiful/) is a project by [stephen wolfram](https://www.youtube.com/watch?v=0bMYtE... |
1a7f19ad-d239-4b3a-bd9e-6ca1418130bb | trentmkelly/LessWrong-43k | LessWrong | Brief Break
I've been feeling burned on Overcoming Bias lately, meaning that I take too long to write my posts, which decreases the amount of recovery time, making me feel more burned, etc.
So I'm taking at most a one-week break. I'll post small units of rationality quotes each day, so as to not quite abandon you. ... |
60cba42a-bf3c-416e-bf39-41779b06a4ef | trentmkelly/LessWrong-43k | LessWrong | Source of Karma
I was thinking the other day it would be interesting to know something about the karma I was getting. Particularly the quality, to the extent that is possible to assess.
I suspect everyone would prefer to have those they respect up voting posts and comments but that type of transparency might not be d... |
0a7d82cd-988b-4628-bbc0-217805278530 | trentmkelly/LessWrong-43k | LessWrong | Rational Feed: Last Week's Community Articles and Some Recommended Posts
===Highly Recommended Articles:
Slack by Zvi Moshowitz - You need slack in your life. Slack lets you explore and invest. If you don't have slack you can't relax or uphold your morals. Fight hard to maintain your slack and don't let people or thi... |
e48c2b31-a790-4d37-823e-8ef5ba6ddbee | trentmkelly/LessWrong-43k | LessWrong | Refine: An Incubator for Conceptual Alignment Research Bets
I’m opening an incubator called Refine for conceptual alignment research in London, which will be hosted by Conjecture. The program is a three-month fully-paid fellowship for helping aspiring independent researchers find, formulate, and get funding for new co... |
f6e73b90-7ae2-419a-bb61-a6ad7c56abde | trentmkelly/LessWrong-43k | LessWrong | Weekly LW Meetups
This summary was posted to LW Main on July 3rd. The following week's summary is here.
Irregularly scheduled Less Wrong meetups are taking place in:
* Australian Less Wrong Mega Meetup #2: 17 July 2015 07:00PM
* Australia-wide Mega-Camp!: 17 July 2015 07:00PM
* Hamburg : 17 July 2015 06:30PM
* O... |
0a04b81f-a4b5-43b3-8e9c-38d02f97573e | trentmkelly/LessWrong-43k | LessWrong | D&D.Sci Scenario Index
There have been a lot of D&D.Sci scenarios, but there's a lot of variance between them in complexity and quality. Some are more difficult, and might not be a good place to start, while others are much simpler - some were very good, while others on reflection didn't flow quite right.
Unfortun... |
57b74e7d-a057-4ca0-b7cd-c68cf4a38db1 | trentmkelly/LessWrong-43k | LessWrong | Contact with reality
(Cross-posted from Hands and Cities)
In thought experiments descended from Nozick’s classic “experience machine,” you consider how being plugged into a machine that generates the experience of a certain kind of life (generally, a very pleasant one) compares with some alternative. Such comparisons... |
3a649894-a7eb-4028-8565-9192e82c781b | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | AXRP Episode 6 - Debate and Imitative Generalization with Beth Barnes
[Google Podcasts link](https://podcasts.google.com/feed/aHR0cHM6Ly9heHJwb2RjYXN0LmxpYnN5bi5jb20vcnNz/episode/NGY3ZTBiMzgtYWM5Ny00NDZmLTgxZmQtOTc1M2VhMGVkMzQz)
This podcast is called AXRP, pronounced axe-urp and short for the AI X-risk Research Pod... |
94fad763-eda4-4e8c-a3cd-c0ee4e1d545e | trentmkelly/LessWrong-43k | LessWrong | Worldly Positions archive, briefly with private drafts
I realized it was hard to peruse past Worldly Positions posts without logging in to Tumblr, which seemed pretty bad. So I followed Substack’s instructions to import the archives into world spirit sock stack. And it worked pretty well, except that SUBSTACK ALSO PUB... |
2cec6ee4-a125-4ff3-946e-ba0586fd770f | trentmkelly/LessWrong-43k | LessWrong | TSR #5 The Nature of Operations
**This is part of a series of posts where I call out some ideas from the latest edition of The Strategic Review (written by Sebastian Marshall), and give some prompts and questions that I think people might find useful to answer. I include a summary of the most recent edition, but it's ... |
a71e0a08-a001-462c-a938-1ea519af57a0 | trentmkelly/LessWrong-43k | LessWrong | SSC Meetups Everywhere 2018
Scott over at SlateStarCodex is running another round of "SSC Meetups Everywhere" and since currently a record-high of 80+ meetups are registered for the coming month, I figured we should help people find their nearest meetup, and make them aware of the opportunity. As part of that, we are ... |
605a029d-6813-4fe9-9471-7e619688e2cc | StampyAI/alignment-research-dataset/lesswrong | LessWrong | The Hidden Complexity of Wishes
>
> "I wish to live in the locations of my choice, in a physically healthy, uninjured, and apparently normal version of my current body containing my current mental state, a body which will heal from all injuries at a rate three sigmas faster than the average given the medical technolo... |
5d5fa9ec-891a-42ae-84a6-7bd415d289d7 | trentmkelly/LessWrong-43k | LessWrong | Is Claude a mystic?
There's an odd tendency for large language models such as Claude to output spiritual meta content if they run long enough. See, for example, some LLM outputs selected by Repligate (LLM content starts in 2024), and Worldspider outputs selected by John Pressman.
One thing that is unclear is how cons... |
ae1fb6be-23c6-4d19-bd7a-33e1a0b095d6 | trentmkelly/LessWrong-43k | LessWrong | Foot-Chording Chords
As part of my rhythm stage setup, I have four pedals which I primarily use for heel-toe drumming. When playing mandolin, or holding a baby, however, I can instead use these pedals to play bass notes or chords. I tell it what key and mode I'm in, and then can choose one of the four chords I'm most ... |
ee73c9b5-cfb1-448c-a5c5-1cdb552a98d7 | trentmkelly/LessWrong-43k | LessWrong | Meetup : HPMOR/Other media discussion
Discussion article for the meetup : HPMOR/Other media discussion
WHEN: 02 March 2015 06:15:00PM (-0800)
WHERE: 1061 Market St #4, San Francisco, CA 94103
This is the first meetup of the month and is therefore the Schelling meetup: If you want to come once a month, now's the tim... |
e1a3f6e6-5c65-45d4-8e9e-26f2e7577952 | trentmkelly/LessWrong-43k | LessWrong | Paper: Discovering novel algorithms with AlphaTensor [Deepmind]
The authors apply an AlphaZero-like algorithm to discover new matrix multiplication algorithms. They do this by turning matrix multiplication into a one-player game, where the state represents how far from correct the current output is, moves are algorith... |
109c104b-7d9f-4bf9-a3c2-e90f4d05a8d6 | trentmkelly/LessWrong-43k | LessWrong | Amoeba roles in tech
In the past couple of years, I’ve found myself roles that are defined by their lack of definition. It’s the kind of role that lets you add value in ways that don’t fit neatly into tech’s boxes, where there’s room for more undirectedness and using your curiosity as a battering ram. Depending on whe... |
5e4b6fc2-d55c-41d8-918f-c7b4901a96a9 | trentmkelly/LessWrong-43k | LessWrong | Writing Russian and Ukrainian words in Latin script
I wanted to figure out how to properly write Russian and Ukrainian words using the Latin alphabet, but it turns out there are dozen different standards for Russian and half a dozen different standards for Ukrainian.
My original plan was to post two short algorithms ... |
9709a282-8596-4480-bbb9-e608a3e7cb26 | trentmkelly/LessWrong-43k | LessWrong | Tetherware #1: The case for humanlike AI with free will
In this post, I argue that more humanlike AI with greater autonomy and freedom isn’t just easier to align with our values; it could also help reduce economic inequality, foster mutual collaboration and accountability, and simply make living with AI more enjoyable... |
9b957ba9-ce52-4ffe-a861-5356a3d7ed3a | StampyAI/alignment-research-dataset/blogs | Blogs | John Ridgway on safety-critical systems
John Ridgway studied physics at the University of Newcastle Upon Tyne and Sussex University before embarking upon a career in software engineering. As part of that career he worked ... |
da4f9206-2018-4919-835b-56431fb8529b | trentmkelly/LessWrong-43k | LessWrong | Resolving von Neumann-Morgenstern Inconsistent Preferences
I consider the problem of resolving preferences that are inconsistent under the von Neumann-Morgenstern axioms into consistent preferences. For preferences over deterministic options, I model inconsistent preferences as directed graphs, and the resolution as s... |
766eae02-ddb9-4847-941f-dde20b7c6c57 | trentmkelly/LessWrong-43k | LessWrong | Results from the language model hackathon
We ran a black box investigation hackathon (original post) with 7 projects submitted by 15 participants. Here we share some of the results (with permission). In summary:
* GPT-3 is less truthful when it is prompted to also be friendly.
* Soliciting harmful advice is a very ... |
660d84e9-e941-465e-84de-74a8a951980a | StampyAI/alignment-research-dataset/blogs | Blogs | clarifying formal alignment implementation
clarifying formal alignment implementation
------------------------------------------
one thing i think bears clarification for the purpose of how i intend to save the world using [AI aligned to formal goals](formal-alignment.html) is that i think giving an AI a model of th... |
70b2cdf7-a315-4a26-8de4-1793969aa168 | trentmkelly/LessWrong-43k | LessWrong | Defensiveness as hysterical invisibility
Sometimes people get defensive. That is, Alice tells Bob, "You have property X", and then Bob tenses his muscles, moves rigidly, pretends to not be bothered, and denies that he has property X, or changes the subject, or asks Alice why she thinks he's X (perhaps in a tenser / h... |
2a5066c3-4bd6-442d-960e-0aceb881d85d | StampyAI/alignment-research-dataset/arbital | Arbital | Category (mathematics)
A **category** consists of a collection of objects with morphisms between them. A morphism $f$ goes from one object, say $X$, to another, say $Y$, and is drawn as an arrow from $X$ to $Y$. Note that $X$ may equal $Y$ (in which case $f$ is referred to as an [https://arbital.com/p/-endomorphism](h... |
12821e02-55b9-4a01-92bc-c0d7659eb0b7 | trentmkelly/LessWrong-43k | LessWrong | A Frontier AI Risk Management Framework: Bridging the Gap Between Current AI Practices and Established Risk Management
We (SaferAI) propose a risk management framework which we think should improve substantially upon existing Frontier Safety Frameworks if followed. It introduces and borrows a range of practice and con... |
2fbf4bcc-e87f-4c05-9ebf-18749530bd60 | trentmkelly/LessWrong-43k | LessWrong | To Be Born in a Bag
Written by Tom Ireland & published by Asimov Press.
The womb is a remarkable organ — a muscular, pear-shaped chamber that supports the transformation of a tiny cluster of dividing cells into an entirely new person. All humans begin their lives in this sturdy chamber.
At least for now.
Several re... |
96e91de1-7d87-4ae4-ac80-7cfd42cfcf51 | trentmkelly/LessWrong-43k | LessWrong | Powering Through vs Working Around
Lately, I’ve been musing on the nature of self-improvement in general. When I notice that something I’ve been doing-- be it mental or physical, the next immediate chain of thought is “Okay, how do I improve my life now, knowing this phenomena exists?” In doing so, I’ve recently rea... |
0dc2075a-a2f6-4f33-8070-1a168504cb78 | trentmkelly/LessWrong-43k | LessWrong | Covid 10/21: Rogan vs. Gupta
I finally got my booster shot yesterday. I intended to get it three weeks ago, but there was so much going on continuously that I ended up waiting until I could afford to be knocked out for a day in case that happened, and because it’s always easy to give excuses for not interacting with s... |
bfe6339a-a632-48fa-9582-6e9938966358 | trentmkelly/LessWrong-43k | LessWrong | D&D.Sci Dungeonbuilding: the Dungeon Tournament
This is an entry in the 'Dungeons & Data Science' series, a set of puzzles where players are given a dataset to analyze and an objective to pursue using information from that dataset.
Complexity Rating: 3/5
STORY
The Dungeon Tournament is held regularly, calling dunge... |
3e9b0269-c7a4-418b-8f4c-86929ba31e6f | trentmkelly/LessWrong-43k | LessWrong | Intransitive Preferences You Can't Pump
In the usual argument of money-pumping we take an agent with preferences A>B, B>C and C>A. Then we offer it to exchange C+$1 for B, then B+$1 for A, and finally A+$1 for C. Now the agent paid $3 and ended up where it started.
The assumption here is that not only does this agent... |
822a6052-5aef-4474-b0aa-db033cb3d7c6 | trentmkelly/LessWrong-43k | LessWrong | The commercial incentive to intentionally train AI to deceive us
We have all experienced application programs telling us something we did not want to hear, e.g., poor financial status, or results of design calculations outside practical bounds. While we may feel like shooting the messenger, applications are treated as... |
5a7f54f3-8158-4734-9cca-0748cd5d6e8e | trentmkelly/LessWrong-43k | LessWrong | It's time to worry about online privacy again
As we all know, if you have nothing to hide, you have nothing to fear. Nobody cares about your private life. You are not an important geopolitical target. Nobody's going to spy on you to know what weird pornography you watch.
And so, around 2015, people gave up on online... |
b8c69c7e-5512-4f8c-a9ca-da9973797e07 | LDJnr/LessWrong-Amplify-Instruct | LessWrong | "A comment I made on social media last year about why MIRI cares about making progress on decision theory:We aren't working on decision theory in order to make sure that AGI systems are decision-theoretic, whatever that would involve. We're working on decision theory because there's a cluster of confusing issues here (... |
388d633c-ba48-4676-ad6c-d7b2204dc4c8 | trentmkelly/LessWrong-43k | LessWrong | Disincentives for participating on LW/AF
I was in a research retreat recently with many AI alignment researchers, and found that the vast majority of them do not participate (post or comment) on LW/AF or participate to a much lesser extent than I would prefer. It seems important to bring this up and talk about whether... |
de854312-598a-48c4-ae44-df97a9b21f8f | trentmkelly/LessWrong-43k | LessWrong | Meetup : Less Wrong NH Meetup
Discussion article for the meetup : Less Wrong NH Meetup
WHEN: 10 May 2016 07:00:00PM (-0400)
WHERE: 269 Pearl St Manchester NH 03104
The eighteenth NH meet-up is Tuesday, 5/10, in Manchester, NH at 7 pm at a private residence. Light refreshments will be provided.
Have you read Ration... |
cf97614f-12ed-4055-ad28-1b1bacc3ba44 | trentmkelly/LessWrong-43k | LessWrong | What are beliefs you wouldn't want (or would feel apprehensive about being) public if you had (or have) them?
Update: Beliefs that are about the world in general, and not about yourself in particular (ie. things you don't want to say about yourself) |
d5c094f6-e9c0-431d-a3f6-00e7ad2e53de | trentmkelly/LessWrong-43k | LessWrong | Ways of being with you
Suppose you want to get in touch with your care for someone, or to feel empathy for them. One way is to imagine what it is like to be them, looking out of their eyes and feeling what they are feeling. For instance, if your friend did something frustrating, you might imagine the experience of cho... |
d8d34efb-41e8-4e49-a6c7-9171cc9502d6 | trentmkelly/LessWrong-43k | LessWrong | Gemini 2.5 Pro: From 0506 to 0605
Google recently came out with Gemini-2.5-0605, to replace Gemini-2.5-0506, because I mean at this point it has to be the companies intentionally fucking with us, right?
> Google: Our updated Gemini 2.5 Pro Preview continues to excel at coding, helping you build more complex web apps.... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.