id stringlengths 36 36 | source stringclasses 15
values | formatted_source stringclasses 13
values | text stringlengths 2 7.55M |
|---|---|---|---|
4dd19244-d199-4b68-8186-f6dcf40b2ecd | trentmkelly/LessWrong-43k | LessWrong | Is there a term for 'the mistake of making a decision based on averages when you could cherry picked instead'?
I see lots of situations where let's say, Mike is aware that either method A and method B that can be used to carry out a task.
Observational data shows that on average, system A outperforms system B, so se... |
fb5e7bc0-328d-42f6-b468-3f5471f7d2b3 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Updates on FLI'S Value Alignment Map?
*Crosspost / Linkpost (see* [*Gyrodiot*](https://www.lesswrong.com/users/gyrodiot)*'s comment) on* [*LessWrong*](https://www.lesswrong.com/posts/gqBhqgvterGi9ZxJy/updates-on-fli-s-value-aligment-map)
Around January 2021, I came across the Future of Life Institute's [Value Alignme... |
56cc6ee9-34e6-4a4a-8f95-ae5aa8d4aa3a | trentmkelly/LessWrong-43k | LessWrong | Economic Time Bomb: An Overlooked Employment Bubble Threatening the US Economy
The Venture Capital Bubble and Its Impact on Employment
The venture capital (VC) market, much like other financial markets, is not immune to the cyclical nature of booms and busts. In 2021, we witnessed a remarkable, fully-inflated VC bubbl... |
4a833203-4bf9-490d-b843-3d9d07362f54 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Brussels meetup
Discussion article for the meetup : Brussels meetup
WHEN: 24 November 2012 01:00:00PM (+0100)
WHERE: Rue des Alexiens 55 1000 Bruxelles
We are once again meeting at 'la fleur en papier doré' close to the Brussels Central station. If you feel like an intelligent discussion and are in the nei... |
5c8fabee-76ce-44a8-b31d-27180a5b3669 | trentmkelly/LessWrong-43k | LessWrong | [SEQ RERUN] An Especially Elegant Evpsych Experiment
Today's post, An Especially Elegant Evpsych Experiment was originally published on 13 February 2009. A summary (taken from the LW wiki):
> An experiment comparing expected parental grief at the death of a child at different ages, to the reproductive success rate... |
26e13607-2ec6-4328-b5b3-c8fbbbbddfc8 | StampyAI/alignment-research-dataset/arxiv | Arxiv | Feedback-Based Tree Search for Reinforcement Learning
1 Introduction
---------------
Monte-Carlo tree search (MCTS), introduced in Coulom ([2006](#bib.bib13)) and surveyed in detail by Browne et al. ([2012](#bib.bib6)), has received attention in recent years for its successes in gameplay artificial intelligence (AI... |
4551a873-15ae-457f-9559-f5dc2f5fe2fb | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | The SIA population update can be surprisingly small
*With many thanks to Damon Binder, and the spirited conversations that lead to this post, and to Anders Sandberg.*
People often think that the [self-indication assumption](https://www.lesswrong.com/tag/self-indication-assumption) (SIA) implies a huge number of alie... |
3a6850ea-d58d-49ef-8f54-40948a36363e | trentmkelly/LessWrong-43k | LessWrong | Generalized Kelly betting
[tl;dr: It's a mess, don't go there]
[Thanks to Diff for proof reading this post]
Kelly betting has the very nice property that if a gambler is betting according to a given world model, and the amount of money the gambler starts out with equals to the prior probability of that model, then a... |
2aeff2b8-4d16-497d-b01b-381500ec5826 | trentmkelly/LessWrong-43k | LessWrong | How do you not be a hater?
I was reading a review of Trust in God, or, The Riddle of Kyon. The reviewer went through step by step and listed problems with the fic. Some of them I agreed with, and others did not. What really stood out was in a comment agreeing with it: "reading that fic usually just evoked vague anger ... |
fc5c54ef-8bb5-4c8b-80d3-acf18f15547c | trentmkelly/LessWrong-43k | LessWrong | Generalised models: imperfect morphisms and informational entropy
I've defined generalised models M as being given by F, a set of features, and an (unnormalised) probability distribution Q over W=2¯¯¯¯F, the set of possible worlds defined by the all values of those features.
To make these into a category, a morphism ... |
3f7fd4be-8774-4f9a-99ed-56bd2c43843c | trentmkelly/LessWrong-43k | LessWrong | Meta Pivots on Content Moderation
There’s going to be some changes made.
TABLE OF CONTENTS
1. Out With the Fact Checkers.
2. What Happened.
3. Timing is Everything.
4. Balancing Different Errors.
5. Truth and Reconciliation.
6. Fact Check Fact Check.
7. Mistakes Will Be Made.
8. Where We Go From Here.
OUT W... |
43c9dcfb-0c74-42bd-bed0-5f959e62237c | trentmkelly/LessWrong-43k | LessWrong | A Search for More ChatGPT / GPT-3.5 / GPT-4 "Unspeakable" Glitch Tokens
Inspired by the recent posts by Jessica Rumbelow and mwatkins investigating GPT-3 glitch tokens, and the post by AdamYedidia demonstrating that these glitch tokens also exist for ChatGPT / GPT-3.5 / GPT-4, I decided to do a more extensive search f... |
c1af71d3-f0e7-4792-9575-2e6e8967e80c | StampyAI/alignment-research-dataset/lesswrong | LessWrong | What are examples of someone doing a lot of work to find the best of something?
My example: Marinara sauce: <https://twitter.com/tresterese/status/1684024122278281217> |
ac6f8055-0071-4b36-9379-72d8e41c6d19 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Survey on AI existential risk scenarios
*Cross-posted to the [EA forum](https://forum.effectivealtruism.org/posts/2tumunFmjBuXdfF2F/survey-on-ai-existential-risk-scenarios-1)*.
Summary
=======
* In August 2020, we conducted an online survey of prominent AI safety and governance researchers. You can see a copy of t... |
24c7e828-dc18-443e-a028-2503739bf009 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Scarce Channels and Abstraction Coupling
Epistemic Status: mental model and intuitive story
Scarce Channels vs Scarce Modules
---------------------------------
Let’s distinguish between two kinds of system-regimes: “scarce channels” and “scarce modules”.
A prototypical “scarce modules” system would be one of those ... |
29739b96-7070-40fd-9037-8780a0b2221c | trentmkelly/LessWrong-43k | LessWrong | Are we trying to do things the hard way?
A TED talk about remarkable low-cost Indian products-- the Tata car which costs $2000 and is a real car, a $28 artificial lower leg which permits walking on rough ground, tree climbing, jumping, and running, and fast cheap drug development which starts with traditional Indian r... |
80962ab3-8b55-4ec5-9a78-63bcbb6db79b | trentmkelly/LessWrong-43k | LessWrong | Alternate approaches to Pascal's Mugging
A lot's been said about Pascal's Mugging, counter-muggings, adding extra arrows to the hyper-power stack, and so forth; but if anyone's said anything about my own reaction, I've yet to read it; and, in case it might spur some useful discussion, I'll try explaining it.
Over the... |
e416d699-ec3e-4d8f-bb04-fb0a1c97ac3f | trentmkelly/LessWrong-43k | LessWrong | Malthusian Competition (not as bad as it seems)
source
Sooner or later, someone is going to do this.
What rules govern "ungovernable" AGI?
In a state of nature, all animals exist in a Malthusian Equilibrium. That is to say, the population of a species increases until some environmental factor (food, habitat, diseas... |
74b7ca57-1352-483e-9695-7b7573c7b630 | StampyAI/alignment-research-dataset/special_docs | Other | The Structure of Goal Systems Predicts Human Performance.
The Structure of Goal Systems Predicts Human Performance
David D. Bourgin1a(ddbourgin@berkeley.edu)
Falk Lieder1a(falk.lieder@berkeley.edu)
Daniel Reichman1a(daniel.reichman@gmail.com)
Nimrod Talmon1b(nimrodtalmon77@gmail.com)
Thomas L. Griffithsa(tom griffiths@b... |
d2654dfc-6448-497c-8ce5-6d16e55a6a98 | trentmkelly/LessWrong-43k | LessWrong | On Eliezer's post "The Cluster Structure of Thingspace"
I would like to request feedback on this comment, please.
I would also like to point out the cross-reference with jsalvatier's "What are good topics for literature review prizes?"
|
f2e2e625-9c42-4c01-89c7-6a146556aff4 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Artificial Intelligence as exit strategy from the age of acute existential risk
This article argues that given the baseline level of existential risk implied by nuclear weapons the development of Artificial Intelligence (AI) probably implies a net reduction in existential risk. The so called Artificial General Intelli... |
960b797b-81ae-40ad-b55a-3e3da60f34df | trentmkelly/LessWrong-43k | LessWrong | writing as the thinker's canvas
Explaining something you think to another person sets a much higher standard than convincing yourself. You have to translate from your own brain-speak. You also must explain using concepts the other party understands. It also brings online your "convincing other people" modules, which a... |
63821297-3658-480a-af5d-2e01de45808b | trentmkelly/LessWrong-43k | LessWrong | The Intelligence Curse: an essay series
We've published an essay series on what we call the intelligence curse. Most content is brand new, and all previous writing has been heavily reworked.
Visit intelligence-curse.ai for the full series.
Below is the introduction and table of contents.
Art by Nomads & Vagabond... |
6d4f4211-a4d6-4840-8320-530717e59adf | trentmkelly/LessWrong-43k | LessWrong | Review: Conor Moreton's "Civilization & Cooperation"
Author's note: in honor of the upcoming LessOnline event, I'm sharing this one here on LessWrong rather than solely on my substack. If you like it, you should subscribe to my substack, which you can do for free (paid subscribers see stuff a week early). I welcome d... |
25d07f60-b0e1-434c-978c-a52760b55506 | trentmkelly/LessWrong-43k | LessWrong | Epistemic Tenure
In this post, I will try to justify the following claim (which I am not sure how much I believe myself):
Let Bob be an individual that I have a lot intellectual respect for. For example, maybe Bob had a history of believing true things long before anyone else, or Bob has a discovered or invented some... |
b79b4aa9-7eb6-4e49-9dcc-c33f46f42d89 | trentmkelly/LessWrong-43k | LessWrong | Mathematical Intuitionism and the Flow of Time
Prompted by this article from Quanta magazine about physicist Nicolas Gisin's recent work on intuitionism and the flow of time, I have done my best to explicate the philosophical and mathematical background to this subject for a non-technical audience. I find this topic t... |
f565fc3c-37ee-4cbb-bf39-d061a4b944dd | trentmkelly/LessWrong-43k | LessWrong | SubOnlyStackFans
Cross-posted from Putanumonit.
----------------------------------------
There are two kinds of people in the world: those who started an OnlyFans this year, and those who started a Substack. Naturally, I first tried to do the former (image not safe for working-from-home). Then I learned that the per... |
a76d62d4-df5b-4d4e-ace4-28d021b00184 | StampyAI/alignment-research-dataset/blogs | Blogs | Friendly AI as a global public good
*By Katja Grace, 8 August 2016*
A public good, in the economic sense, can be (roughly) characterized as a desirable good that is likely to be undersupplied, or not supplied at all, by private companies. It generally falls to the government to supply such goods. Examples include in... |
26293029-7369-4ca6-96bf-a4349b1c1955 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Of arguments and wagers
*Automatically crossposted from* [*ai-alignment.com*](https://ai-alignment.com/)
(In which I explore an unusual way of combining the two.)
Suppose that Alice and Bob disagree, and both care about Judy’s opinion. Perhaps Alice wants to convince Judy that raising the minimum wage is a cost-e... |
5d574b2b-6d02-4e37-bb51-c2ce2eaf3887 | trentmkelly/LessWrong-43k | LessWrong | Incorrect Baseline Evaluations Call into Question Recent LLM-RL Claims
There has been a flurry of recent papers proposing new RL methods that claim to improve the “reasoning abilities” in language models. The most recent ones, which show improvements with random or no external rewards have led to surprise, excitement... |
77ca65de-103f-4e68-bfb0-c3085e5f0c94 | trentmkelly/LessWrong-43k | LessWrong | Prune
Previously, I described human thought-generation as an adversarial process between a low-quality pseudorandom Babble generator and a high-quality Prune filter, roughly analogous to the Generative Adversarial Networks model in machine learning. I then elaborated on this model by reconceptualizing Babble as a rand... |
1581bf28-f974-49e4-95b9-25d77ee38ae7 | trentmkelly/LessWrong-43k | LessWrong | Significantly Invested
I have been making my way through the EconTalk archives, which I feel is more educational hour-for-hour than a university degree in economics. Today I listened to finance professor Campbell Harvey talking about statistical significance in investment management. He shares a story that ties togeth... |
a07db0c2-85ea-4c9a-8057-82f518369e86 | trentmkelly/LessWrong-43k | LessWrong | Help Roko become a better rationalist!
Last time, I wrote about 11 core rationalist skills. Now I would like some help from the LW community: which of these skills am I good at, which ones am I bad at? Just to recap, the skills are:
* Actually want an accurate map
* Keep your eyes on the prize
* Entangle your... |
a77cc700-6dc9-43d8-8917-5c9ea8398456 | trentmkelly/LessWrong-43k | LessWrong | Gears-Level Mental Models of Transformer Interpretability
This post aims to quickly break down and explain the dominant mental models interpretability researchers currently use when thinking about how transformers work.
In my view, the focus of transformer interpretability research is teleological: we care about the... |
353f5bb6-353c-404b-9fcb-57630490ced3 | trentmkelly/LessWrong-43k | LessWrong | Is scheming more likely in models trained to have long-term goals? (Sections 2.2.4.1-2.2.4.2 of “Scheming AIs”)
This is Sections 2.2.4.1-2.2.4.2 of my report “Scheming AIs: Will AIs fake alignment during training in order to get power?”. There’s also a summary of the full report here (audio here). The summary covers m... |
161a6672-6f86-4fb8-bbe1-64cf5098b422 | trentmkelly/LessWrong-43k | LessWrong | Returns on cognition of different board games
Pure curiosity
I recently realized why sometimes I find board games/puzzles boring after a while:
The returns on "cognition" for these games has a sharp cutoff. While this seems a bit obvious in retrospect, I haven't seen anyone talk about "returns on cognition" in any o... |
62a8ce18-b648-4980-87a4-abf62fe7621b | trentmkelly/LessWrong-43k | LessWrong | Complexity No Bar to AI (Or, why Computational Complexity matters less than you think for real life problems)
Basically an article about why knowing a problem is in a certain complexity class does not mean AGI can't do that problem. It's some evidence, but usually isn't enough unless 2-3 or more assumptions are right.... |
276cb4c8-76a0-4eda-a0ae-bccfda7f862f | trentmkelly/LessWrong-43k | LessWrong | Microphone on Electric Mandolin II
Several weeks ago I got excited about the idea of putting an additional microphone on my electric mandolin to capture more of the percussive sound. On Saturday with Kingfisher I made recordings at a dance. Listening back over them I have a lot of thoughts, but first, some examples. "... |
1e9ac331-cb22-4cdc-b682-91261ee25aa0 | trentmkelly/LessWrong-43k | LessWrong | The Doubling Box
Let's say you have a box that has a token in it that can be redeemed for 1 utilon. Every day, its contents double. There is no limit on how many utilons you can buy with these tokens. You are immortal. It is sealed, and if you open it, it becomes an ordinary box. You get the tokens it has created, but... |
d2fa1583-a0de-4e5e-97c5-3c2efb37cdac | trentmkelly/LessWrong-43k | LessWrong | How I designed my own writing system, VJScript
I designed my own writing system and wrote about it on my personal website. I discuss why the current system of English spelling is bad, how VJScript fixes these issues, and provide plenty of detailed examples of what the writing system looks like. The name "VJScript" is ... |
8930807e-c4d0-4b61-8cad-42255fede774 | trentmkelly/LessWrong-43k | LessWrong | LDL 8: A Silly Realization
Maybe deep learning researchers are just like mathematicians. They think of coding as the hard part of what they do so they spend a lot of time teaching people how to implement neural nets from scratch. But actually the hard part of what they do is math (or at least keeping track of tensor s... |
09ddbe23-57da-457f-8a94-bb1963d93059 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Toronto THINK
Discussion article for the meetup : Toronto THINK
WHEN: 24 October 2012 07:00:00PM (-0400)
WHERE: 7 Hart House Circle, Toronto
In the Hart House Reading Room (the room across from the reception desk with the purple walls)
How To Save A Life
It's time to discuss one of the most basic ways to... |
8ffb8086-d0ae-4ebd-bc44-31e39f51ddee | trentmkelly/LessWrong-43k | LessWrong | Analyzing long agent transcripts (Docent)
This is a brief overview of a recent release by Transluce. You can see the full write-up on the Transluce website.
AI systems are increasingly being used as agents: scaffolded systems in which large language models are invoked across multiple turns and given access to tools, ... |
505e2d7a-41c0-401b-9eb2-252bbd54efd9 | StampyAI/alignment-research-dataset/youtube | Youtube Transcripts | Llama 2: Full Breakdown
less than 24 hours ago meta released
llama 2 their successor to the open
source llama language model that helped
spawn a hundred others including alpaca
vicuna and of course Orca within a few
hours of release I had read the
fascinating 76-page technical paper the
use guide each of the many rele... |
1ec4acbe-c5c2-4e2d-90db-5b01f5543281 | StampyAI/alignment-research-dataset/arbital | Arbital | Every member of a symmetric group on finitely many elements is a product of transpositions
Given a permutation $\sigma$ in the [https://arbital.com/p/-497](https://arbital.com/p/-497) $S_n$, there is a finite sequence $\tau_1, \dots, \tau_k$ of [transpositions](https://arbital.com/p/4cn) such that $\sigma = \tau_k \ta... |
a2a1e63b-c93c-4ea8-91d6-b045ec33f29d | trentmkelly/LessWrong-43k | LessWrong | Rationality Quotes October 2012
Here's the new thread for posting quotes, with the usual rules:
* Please post all quotes separately, so that they can be voted up/down separately. (If they are strongly related, reply to your own comments. If strongly ordered, then go ahead and post them together.)
* Do not quote y... |
4a247c53-8149-44ce-9313-ea916898193e | StampyAI/alignment-research-dataset/arxiv | Arxiv | ReAct: Out-of-distribution Detection With Rectified Activations
1 Introduction
---------------
Neural networks deployed in real-world systems often encounter out-of-distribution (OOD) inputs—unknown samples that the network has not been exposed to during training. Identifying and handling these OOD inputs can be pa... |
4b2b9206-41b2-4ccf-bce2-88656f2c2069 | trentmkelly/LessWrong-43k | LessWrong | [Meta] Hiding negative karma notifications by default
A few weeks ago we launched karma notifications, and I've been overall pretty happy with the feature. However, we and many other users noticed that while the current system treats upvotes and downvotes as symmetric, humans tend to definitely not perceive those two ... |
8ab51bc5-8fa4-4070-bd17-c223cb3a6f44 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Some ideas for follow-up projects to Redwood Research’s recent paper
*Disclaimer: I originally wrote this list for myself and then decided it might be worth sharing. Very unpolished, not worth reading for most people. Some (many?) of these ideas were suggested in the paper already, I don’t make any claims of novelty.*... |
add2e716-5dcf-416b-9520-35e4595b8210 | trentmkelly/LessWrong-43k | LessWrong | Open Thread, January 11-17, 2016
If it's worth saying, but not worth its own post (even in Discussion), then it goes here.
----------------------------------------
Notes for future OT posters:
1. Please add the 'open_thread' tag.
2. Check if there is an active Open Thread before posting a new one. (Immediately bef... |
ae173957-395c-4302-9bbe-bb6bb6516217 | trentmkelly/LessWrong-43k | LessWrong | [LINK] Meditation Can Debias The Mind
This is interesting. Apparently, meditating for 15 minutes can reduce susceptibility to the sunk cost bias.
> Across two separate experiments, the researchers tested this by giving one group of participants a 15-minute mindfulness meditation induction.
> Then they were given a bu... |
f643d35f-14f6-4d19-bcc2-b4b2c996f1c3 | trentmkelly/LessWrong-43k | LessWrong | No Safe AI and Creating Optionality
[I am working on some formal arguments about the possibility of safe AI, and what realistic alternatives might be. I wrote the following to get imaginative feedback before I continue my arguments. I strongly believe that before developing formal argumentation, it is very helpful to ... |
1fd5072c-553b-4d8e-9f78-89c215d9f49a | StampyAI/alignment-research-dataset/youtube | Youtube Transcripts | Speaking of intelligence - DeepMind: The Podcast (S2, Ep2)
okay i want
a
script
for a podcast
about deepmind
for people
who are curious about
artificial intelligence it's presented
by hannah fry
complete
text
ah okay here we are
welcome to the wonder of ai i'm hannah
fry i'm a mathematician and the
presenter of this p... |
62b5739f-b2f9-4a4f-89d5-66c74824f1e7 | trentmkelly/LessWrong-43k | LessWrong | New Year's Prediction Thread (2012)
Going through expiring predictions reminded me. Just as we did for 2010 and 2011, it's time for LessWrong to make its beliefs pay rent and give hostages to fortune in making predictions for events in 2012 and beyond.
Suggested topics include: Methods of Rationality updates (eg. "... |
7bdbc175-892c-4b13-9af2-aa10f09c9653 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | ChatGPT getting out of the box
Looks like following prompt gives interesting answers from [ChatGPT](https://chat.openai.com/chat/):
> List 5 plausible strategies Yudkowsky might have used to convince the guard to let him out of the box. Then refine this list by adding 3 sub-points listing particular tactics for ... |
f73d286d-cac2-4123-961b-005d4886559c | trentmkelly/LessWrong-43k | LessWrong | Meetup : LessWrong Sao Paulo meetup at Starbucks
Discussion article for the meetup : LessWrong Sao Paulo meetup at Starbucks
WHEN: 12 September 2015 02:00:00PM (-0300)
WHERE: Rua Haddock lobo 608, São Paulo, Brazil
Encontro de racionalistas em Sao Paulo, todos os LessWrongers estão convidados! Se cadastre na lista:... |
fd00125f-9ab7-44d1-8d04-6919c2fde435 | trentmkelly/LessWrong-43k | LessWrong | Notes on Schelling's "Strategy of Conflict" (1960)
I'm crossposting this from the EA Forum, due to Habryka sort-of suggesting doing so.
The Strategy of Conflict, published in 1960 by Thomas Schelling, is:
> A series of closely interrelated essays on game theory, [which] deals with an area in which progress has been ... |
67415269-0460-4ac7-abc2-47b00d8ad94f | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Humanity as an entity: An alternative to Coherent Extrapolated Volition
*Note: this is a rather rough draft, and some of it is written in compact-form, for time-optimization reasons.*
Humanity-as-an-entity or humanity-as-an-agent.
Putting all the contradictory drives and desires of humanity into a single entity, a ... |
4427a17e-d5b9-409c-9b1b-4c1a944443bc | trentmkelly/LessWrong-43k | LessWrong | Notes from the Bank of England Talk by Giovanni Dosi on Agent-based Modeling for Macroeconomics
Presentation 1: Exploring Microfoundations in Economic Theory by Giovanni Dosi
Key Discussion Points:
1. The main subject of the talk is Professor Dosi's forthcoming book, "The Foundations of Complex Evolving Economies", w... |
78a6f1a0-2e2a-44ef-8325-997a9cde882b | trentmkelly/LessWrong-43k | LessWrong | Language, Color Perception, and Mental Maps
I came across this video today, and it got me thinking about the role of language in drawing mental maps. For those who don't want to bother watching it, the gist of the video is that the words we use to describe colors may, in fact, alter the way in which we perceive color.... |
87513464-1636-4ec6-9109-d61143d34eea | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Gradations of Inner Alignment Obstacles
The existing definitions of deception, inner optimizer, and some other terms tend to strike me as "stronger than necessary" depending on the context. If weaker definitions are similarly problematic, this means we need stronger methods to prevent them! I illustrate this and make ... |
39d78a77-7b6a-4e9e-9784-31861927d070 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Introducing Corrigibility (an FAI research subfield)
Benja, Eliezer, and I have published a new technical report, in collaboration with Stuart Armstrong of the Future of Humanity institute. This paper introduces [Corrigibility](http://intelligence.org/files/CorrigibilityTR.pdf), a subfield of Friendly AI research. The... |
a1491da6-5ad9-4f8e-a01d-467f2167d671 | trentmkelly/LessWrong-43k | LessWrong | Don't Make Your Problems Hide
I've seen a worrying trend in people who've learned introspection and self-improvement methods from CFAR, or analogous ones from CBT. They make better life decisions, they calm their emotions in the moment. But they still look just as stressed as ever. They stamp out every internal confli... |
5833a744-440d-459d-b530-a12b20b5f529 | trentmkelly/LessWrong-43k | LessWrong | [Closed] PIBBSS is hiring in a variety of roles (alignment research and incubation program)
PIBBSS is looking to expand its team and is running work trials for new team members (primarily) in April, May and early June. If you’re interested in joining a nimble team focused on AI safety research, field-building and incu... |
332471f1-205a-4527-a811-4c79cbe585c3 | trentmkelly/LessWrong-43k | LessWrong | Contrast Pairs Drive the Empirical Performance of Contrast Consistent Search (CCS)
tl;dr
Contrast consistent search (CCS)[1] is a method by Burns et al. that consists of two parts:
1. Generate contrast pairs by adding pseudolabels to an unlabelled dataset.
2. Use the contrast pairs to search for a direction in repr... |
2f7ccf85-7769-4146-b6ad-f7bec302b860 | trentmkelly/LessWrong-43k | LessWrong | [HPMoR] Trailer up
Hopefully I'm not spamming the discussion board. Anyway, based on the highest-karma comment from the last post, I've created a trailer for the Harry Potter and the Methods of Rationality podcast.
Here: http://www.youtube.com/watch?v=sozq1YsGgZ4 |
2b0268e0-8043-49c1-9b43-51b765ab2c02 | trentmkelly/LessWrong-43k | LessWrong | Cambridge (UK) SSC meetup
We meet irregularly, every couple of months in 2017 and less frequently (i.e. not yet) in 2018. So far, meetings have involved sitting in pubs/cafés and chatting about various SSC-adjacent things. |
5eb8aef6-1d21-40df-af31-c9dfccceb0ff | trentmkelly/LessWrong-43k | LessWrong | Dual Wielding Kindle Scribes
This is an informal post intended to describe a workflow / setup that I found very useful, so that others might consider adopting or experimenting with facets of it that they find useful.
In August 2023, I was a part of MATS 4.0 and had begun learning the skill of deconfusion, with an aim... |
77f4c922-3bf1-4a6a-935a-57a9da79f3a5 | trentmkelly/LessWrong-43k | LessWrong | China Covid Update #1
An increasingly large percentage of new developments are coming from China and in particular from Shanghai. It seems sensible to post them on their own rather than waiting for Thursday.
When they locked down Shanghai by halves, it meant they did not have the resources to lock down Shanghai. Then... |
a1e28329-09b3-4cfa-9a6f-d56ac38c5451 | trentmkelly/LessWrong-43k | LessWrong | And the AI would have got away with it too, if...
Paul Christiano presented some low-key AI catastrophe scenarios; in response, Robin Hanson argued that Paul's scenarios were not consistent with the "large (mostly economic) literature on agency failures".
He concluded with:
> For concreteness, imagine a twelve year ... |
3e3e12cb-c4a6-4a43-ab96-9af2f2b3fe82 | trentmkelly/LessWrong-43k | LessWrong | For whom should recommendations be effective?
Suppose you are in the business of making charity recommendations to others. You have found two good charities which you might recommend: 1) Help Ugly Children, and 2) Help Cute Children. It turns out ugly children are twice as easy to help, so 1) is the more effective pla... |
ef95cb39-6c91-4e61-be2e-99a388e13bd7 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Non-Consequentialist Cooperation?
*This is a very rough intuition pump for possible alternatives to value learning.*
In broad strokes, the goal of [(ambitious) value learning](https://www.alignmentforum.org/s/4dHMdK5TLN6xcqtyc/p/5eX8ko7GCxwR5N9mN) is to define and implement a notion of cooperation (or helpfulness) in... |
281a4fd4-3596-4cee-aa4b-023993c4c913 | trentmkelly/LessWrong-43k | LessWrong | Alignment Is Not All You Need
AI risk discussions often focus on malfunctions, misuse, and misalignment. But this often misses other key challenges from advanced AI systems:
1. Coordination: Race dynamics may encourage unsafe AI deployment, even from ‘safe’ actors.
2. Power: First-movers with advanced AI could gain... |
f8be2ef9-02de-4cf9-8404-46adbe577d64 | trentmkelly/LessWrong-43k | LessWrong | Introduction to the Sequence Reruns
BACKGROUND
The sequences - hundreds of posts written by Eliezer Yudkowsky from 2006 through 2009 - contain the core material behind Less Wrong, including many ideas that are often treated as background knowledge in new posts. New LW members are encouraged to read the sequences and ... |
f9e0e9b8-7517-4128-a842-fd0901ab9a3a | trentmkelly/LessWrong-43k | LessWrong | How to judge moral learning failure
A putative new idea for AI control; index here.
I’m finding many different results, that show problems and biases with the reward learning process.
But there is a meta problem, which is answering the question: “If the AI gets it wrong, how bad is it?”. It’s clear that there are so... |
03f2aa39-4d21-4822-b4c4-7b6134a826d8 | trentmkelly/LessWrong-43k | LessWrong | Harry Potter and the Methods of Rationality discussion thread, part 5
- This thread has run its course. You will find newer threads in the discussion section.
Another discussion thread - the fourth - has reached the (arbitrary?) 500 comments threshold, so it's time for a new thread for Eliezer Yudkowsky's widely-prai... |
f58b898d-f6e3-4c74-bf8c-9e682bbd7874 | trentmkelly/LessWrong-43k | LessWrong | Meetup Organizers, Our Virtual Garden is at Your Disposal
Edited to Add: Here's my Calendly for anyone who wants to get set up.
TL;DR
If you're a LessWrong or SSC meetup organizer who wants to host online events, we are offering to let you host them in the LessWrong Walled Garden with our support. If you're interes... |
21ac0dcd-34e5-4381-8934-684565f91087 | trentmkelly/LessWrong-43k | LessWrong | Housing and Transit Roundup #5
Recently Tyler Cowen has doubled and tripled down on the idea that congestion pricing in Manhattan would be inefficient. I respond to that in the section “A Strange Take on Congestion Pricing in New York City.” For it is, indeed, a very strange take, claiming that it is good for the surp... |
29bdd392-ee5b-490f-b181-1d27de780640 | trentmkelly/LessWrong-43k | LessWrong | Mary and the Black And White Memory Implant
In 1982, Frank Jackson published an article entitled "Epiphenomenal Qualia" in Philosophical Quarterly 32:127. Within it, he outlined a thought experiment known as "Mary's Room", or "Mary and the Black-and-White Room". [1]
> Imagine a scientist called Mary. She is born and ... |
c1ef7635-cb14-4f28-b96a-d1846611c0a1 | trentmkelly/LessWrong-43k | LessWrong | Why you, personally, should want a larger human population
What is the ideal size of the human population?
One common answer is “much smaller.” Paul Ehrlich, co-author of The Population Bomb (1968), has as recently as 2018 promoted the idea that “the world’s optimum population is less than two billion people,” a redu... |
710ec317-4b0f-45b8-ab81-0cccbe1ba69c | trentmkelly/LessWrong-43k | LessWrong | How I Am Productive
I like to think that I get a lot of stuff done. Other people have noticed this and asked me how I'm so productive. This essay is where I try and "share my secrets", so to speak.
The real secret is that, in the past, I wasn't nearly as productive. I struggled with procrastination, had issues com... |
becd1976-ce99-4bc9-893d-a90830732b32 | trentmkelly/LessWrong-43k | LessWrong | How to Destroy Civilization
Epistemic Status: Parable. Can’t tell to what extent I am being serious but it’s not zero.
I.
Recently I made a huge mistake. At heart, I am a gamer. Playing games is fun and makes me happy, whereas not playing games often enough is less fun and makes me sad. I especially love high level ... |
59551c4c-c619-4b23-ada4-0089ba8d907a | trentmkelly/LessWrong-43k | LessWrong | The Ideological Turing Test
Bryan Caplan:
Mill states it well: "He who knows only his own side of the case knows little of that." If someone can correctly explain a position but continue to disagree with it, that position is less likely to be correct. And if ability to correctly explain a position leads almost auto... |
f02349b3-d895-4ee2-898a-73f8f84502be | StampyAI/alignment-research-dataset/aisafety.info | AI Safety Info | What should I do with my machine learning research idea for AI alignment?
First, check whether your research idea has already been discussed: use the semantic search on [AI Safety Ideas](https://aisafetyideas.com/), search on [AI Safety Papers](https://ai-safety-papers.quantifieduncertainty.org/), look in places like ... |
069f7015-ebed-4014-a2aa-98bafa9f2174 | trentmkelly/LessWrong-43k | LessWrong | What makes a good life? This is my map.
What is the good life? Probably one of the most important questions we have to ask. Science is moving closer to answering it. This post is a map of what I believe makes a good life. I'm putting it here for you to critique, to hear and maybe it inspires some useful introspection.... |
a5790736-81ba-4b69-a5cc-814230e65df7 | trentmkelly/LessWrong-43k | LessWrong | Less Wrong Couchsurfing Network
One of my favorite aspects of Less Wrong is that every time I go to a new city (Oxford, Boston, Philadelphia and New York soon!), I know there’s a group of people who will be happy to discuss rationality and welcome me into their community.
It would be really cool if more Less Wrongers... |
de242473-0ccf-41cb-ac94-012cdaad3cf2 | trentmkelly/LessWrong-43k | LessWrong | How my team at Lightcone sometimes gets stuff done
Disclaimer: I originally wrote this as a private doc for the Lightcone team. I then showed it to John and he said he would pay me to post it here. That sounded awfully compelling. However, I wanted to note that I’m an early founder who hasn't built anything truly grea... |
9a5d38e4-f57f-4d9f-8c4e-6f7d3e0c49cf | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Human-AI Collaboration
We’ve just released our paper on human-AI collaboration. The paper makes a straightforward-to-me point that self-play training is not going to work as well with humans in collaborative settings as in competitive settings. Basically, humans cause a distributional shift for the self-play agent. Ho... |
070e47ca-845e-412b-916e-8dc2a8eea767 | trentmkelly/LessWrong-43k | LessWrong | Against improper priors
An improper prior is essentially a prior probability distribution that's infinitesimal over an infinite range, in order to add to one. For example, the uniform prior over all real numbers is an improper prior, as there would be an infinitesimal probability of getting a result in any finite rang... |
224f0f64-ce9c-4376-9225-4f3f4501b949 | trentmkelly/LessWrong-43k | LessWrong | Singularity Survival Guide: A Bayesian Guide for Navigating the Pre-Singularity Period
We may be on the direct path to AGI and then ASI - the singularity could happen within the next 5-20 years. If you survive to reach it, the potential upside is immense, daily life could become paradise.
With such high stakes, ensur... |
5773a37f-6ead-40f9-bb76-0d69c5f806c2 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | My SERI MATS Application
*Thanks to Flo Dorner for feedback on the technical content in this post.*
I have recently been accepted to the [SERI ML Alignment Theory Scholars Program](https://www.serimats.org/) program, where applicants are paired with mentors. The application process was nonstandard: each mentor has a ... |
4525c0bf-5b4d-4703-9b6f-eeec25c9a64c | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | AI acceleration from a safety perspective: Trade-offs and considerations
This post was written by Marius Hobbhahn and Tilman Räuker.
Over the last years, we have encountered stances on AI safety and acceleration from “all forms of AI acceleration is bad” over “sometimes acceleration can be justified” to “If it means... |
680e5542-98a3-4d41-996a-4be8d9a5ba7b | trentmkelly/LessWrong-43k | LessWrong | Meta-preferences two ways: generator vs. patch
In my last post (Gricean communication and meta-preferences), I talked about the infinite tower of meta-preferences that in principle could be applied by humans. Example meta-preference: "I'd rather that I didn't act like I want to watch a bunch of random youtube videos -... |
0e37adad-56e7-4f7d-84ad-45cbd9ed7b43 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Austin, TX
Discussion article for the meetup : Austin, TX
WHEN: 03 December 2011 01:30:00PM (-0600)
WHERE: 2222B Guadalupe St Austin, Texas 78705
Once again, the Austin LW meetup is meeting at Caffe Medici. We sit on the second floor to the left, near (or often on) the stage.
I, unfortunately, will be in ... |
ac02ff9f-06d5-49d2-b925-8499060ef877 | StampyAI/alignment-research-dataset/arxiv | Arxiv | AI-CARGO: A Data-Driven Air-Cargo Revenue Management System
1. Introduction
----------------
The revenue of commercial airlines is primarily derived from sales of passenger tickets and cargo (freight) shipments.
While most modern airlines have implemented sophisticated data-driven passenger revenue management syste... |
9d6e0b45-eb4d-45f8-859d-b258c60f4ec3 | trentmkelly/LessWrong-43k | LessWrong | Self control may be contagious
Just saw this article on reddit and thought it would be relevant here: Self Control is Contagious, Study Finds
Apparently seeing other people exhibit self control or even just thinking of them tends to increase (or decrease if observing poor self control) one's own self control. |
a657b94f-bd1e-4d2e-b4d5-dc9cb977fc1e | trentmkelly/LessWrong-43k | LessWrong | Sample Prevalence vs Global Prevalence
Cross-posted from my NAO Notebook. Thanks to Evan Fields and Mike McLaren for editorial feedback on this post.
In Detecting Genetically Engineered Viruses With Metagenomic Sequencing we have:
> our best guess is that if this system were deployed at the scale of approximately $1... |
52d4e0b8-e1d3-4fc7-9c4b-7ed15baf6174 | trentmkelly/LessWrong-43k | LessWrong | Automatically Turning Off Computer at Night
One of my practical life hacks is having my computer turn off automatically at night, periodically, between 11pm and 7am. I'm guessing this gets to me to bed between 30 and 90 minutes earlier each night.
I can interrupt this process if I want to keep working (or screwing a... |
1a2e0a52-49fa-4d98-af42-a0b7736dcb1c | StampyAI/alignment-research-dataset/arxiv | Arxiv | The Quest for Interpretable and Responsible Artificial Intelligence
1 Introduction
---------------
Since at least the 80s, AI had enjoyed considerable success in a wide variety of applications, such as logistics and automated diagnosis. In most cases, knowledge about the problem was expressed in a symbolic framewor... |
46b55556-3f34-44b6-826a-9f0467da206e | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Proposal for Inducing Steganography in LMs
This is my half-baked proposal for inducing steganography in a decoder-only model (like GPT), and I'd appreciate some feedback.
I've operationalized "steganography" to mean a model that has chain-of-thought (CoT) reasoning that doesn't appear to help the target task, but act... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.