id stringlengths 36 36 | source stringclasses 15
values | formatted_source stringclasses 13
values | text stringlengths 2 7.55M |
|---|---|---|---|
9dfab62b-e4ae-4447-988b-703b131b0871 | trentmkelly/LessWrong-43k | LessWrong | Fix it and tell us what you did
The main danger for LW is that it could become rationalist-porn for daydreamers.
I suggest a pattern of counterattack:
1. Find a nonrational aspect of your nature that is hindering you right now.
2. Determine privately to fix it.
3. Set a short deadline. Do the necessary work.
... |
ebc234d5-eca2-4a80-8b39-79d585b19d56 | trentmkelly/LessWrong-43k | LessWrong | Introducing the Basic Post-scarcity Map
A society in which all basic human needs are provided at zero or very low cost without significant human work can be defined as basic post-scarcity.
Our civilization is getting close to the point where this is technically feasible.
To reach this milestone as soon as possible, ... |
f90bf902-054d-435c-93e7-b754ac4735e4 | trentmkelly/LessWrong-43k | LessWrong | Marketing Rationality
What is your opinion on rationality-promoting articles by Gleb Tsipursky / Intentional Insights? Here is what I think:
Trying to teach someone to think rationally is a long process -- maybe even impossible for some people. It's about explaining many biases that people do naturally, demonstrating... |
61925ded-f7d7-4fb8-9457-68da85b4b5db | trentmkelly/LessWrong-43k | LessWrong | Limits of Current US Prediction Markets (PredictIt Case Study)
(Disclaimers: I work in the financial industry, though not in a way related to prediction markets. Anything I write here is my opinion and not that of my employer.
This is a US-centric piece based on a case study of PredictIt: as some people have pointed o... |
f19e4e1b-54bd-4567-bc05-5f1b5f4ebc34 | StampyAI/alignment-research-dataset/blogs | Blogs | alignment is an optimization processes problem
alignment is an optimization processes problem
----------------------------------------------
i like to talk about [AI](https://www.lesswrong.com/tag/ai) alignment a lot, but the matter of alignment is really general to *optimization processes* in general.
here are so... |
8db97031-89c1-4256-8572-2df1be07a592 | trentmkelly/LessWrong-43k | LessWrong | AI #16: AI in the UK
It would make strategic sense, even from a selfish perspective, for essentially any nation to get more involved in sensible AI regulation, and to get more involved in capturing AI benefits and being the home to AI and AI-enabled economic engines.
It makes especially good sense for the UK, which ... |
8b5a313a-4fd7-4060-a399-4081e2123fbd | trentmkelly/LessWrong-43k | LessWrong | Core of AI projections from first principles: Attempt 1
I have at various times commented on AI projections and AI risk, using my own models of the key issues. However I have not really crystallized any overview of what my models imply, partly because I am uncertain about some key issues, but also partly because I jus... |
12270c5a-5f96-46a5-bc92-6043e93b4f9f | trentmkelly/LessWrong-43k | LessWrong | Airportpourri
A random grab bag of vacation photos, travel observations, and geopolitical opinions. |
7f5791ef-922c-4693-a0d3-4a3574011c8d | trentmkelly/LessWrong-43k | LessWrong | Very Unnatural Tasks?
In popular AGI takeoff scenarios, AGIs end up doing something "unnatural" where the complexity of the goal is completely orthogonal to their ability to fulfill that goal. I'm looking for a scenario in which an AGI's objective function is "very unnatural". Not maximizing paperclips, building compa... |
2f127490-e6be-43f8-867a-0c101adbebb2 | trentmkelly/LessWrong-43k | LessWrong | Publication of "Anthropic Decision Theory"
My paper "Anthropic decision theory for self-locating beliefs", based on posts here on Less Wrong, has been published as a Future of Humanity Institute tech report. Abstract:
> This paper sets out to resolve how agents ought to act in the Sleeping Beauty problem and various ... |
2fa75605-6143-468e-81f8-8e8d965aea72 | trentmkelly/LessWrong-43k | LessWrong | AXRP Episode 21 - Interpretability for Engineers with Stephen Casper
YouTube link
Lots of people in the field of machine learning study ‘interpretability’, developing tools that they say give us useful information about neural networks. But how do we know if meaningful progress is actually being made? What should we ... |
37628cc9-a50a-4a38-a370-2f049cc38876 | trentmkelly/LessWrong-43k | LessWrong | Whether to care about the desires of the dead
Many people claim that fulfilling a person’s desires after they are dead cannot benefit that person. Not just that it isn’t strategic to honour such desires, but that such desires can’t be fulfilled. Either it is not really possible for a person to desire for things that w... |
dc1efaba-3a4d-4061-8d2c-83777e0ced63 | trentmkelly/LessWrong-43k | LessWrong | Bayesian exercise
I am confused.
Suppose you are in charge of estimating the risk of catastrophic failure of the Space Shuttle. From engineers, component tests, and guesswork, you come to the conclusion that any given launch is about 1% likely to fail. On the strength of this you launch the Shuttle, and it does no... |
919c379f-9412-4c42-a6f2-5a81ba51d43e | awestover/filtering-for-misalignment | Redwood Research: Alek's Filtering Results | id: post154
The AGI Safety & Alignment Team (ASAT) at Google DeepMind (GDM) is hiring! Please apply to the Research Scientist and Research Engineer roles. Strong software engineers with some ML background should also apply (to the Research Engineer role). Our initial batch of hiring will focus more on hiring engineers... |
e68334b2-4b2f-4b0c-92b3-ea9eb3bdd602 | trentmkelly/LessWrong-43k | LessWrong | Spoiler-Free Reviews: Monster Slayers, Dream Quest and Loop Hero
Recently my love of card-based roguelikes caused me to check out these two games.
First, Loop Hero.
Loop Hero is a unique Tier 2 (Worth It) game that came out recently.
If you’re up for a unique retro-graphical semi-auto-battling rogue-like, stop read... |
7db9d30a-6f0f-4886-8ae4-76600691e610 | trentmkelly/LessWrong-43k | LessWrong | Convergent Linear Representations of Emergent Misalignment
Ed and Anna are co-first authors on this work.
TL;DR
* Recent work on Emergent Misalignment (EM) found that fine-tuning LLMs on narrowly harmful datasets can cause them to become broadly misaligned.
* We find a linear direction for misalignment in emergent... |
b99f660c-28fa-4d1e-93a5-762a962c6330 | trentmkelly/LessWrong-43k | LessWrong | Stanley Meyer's water fuel cell
I read somewhere a comment claiming that a "water engine" (Stanley Meyer's fuel cell) is possible. Wikipedia calls it a perpetual motion machine, but a quick search indicates that the guy just claimed it used water as fuel (nothing about it producing more energy than it consumed).
The ... |
f9e1af17-b8e0-475f-b9fd-580e679938eb | trentmkelly/LessWrong-43k | LessWrong | Werewolf, Cambridge UK Less Wrong Meetup April 1st 2012
There is already a post related to this meetup but it concerns a discussion which took place after I had left so I will write about the games of Werewolf. Please post your thoughts too and correct any inaccuracies.
Thoughts:
* Most people said that this was v... |
a986f474-5348-4455-9d76-89b615c20787 | StampyAI/alignment-research-dataset/blogs | Blogs | New grants from the Open Philanthropy Project and BERI
I’m happy to announce that MIRI has received two major new grants:
* A two-year grant totaling $2,112,500 from the [Open Philanthropy Project](https://www.openphilanthropy.org/about/who-we-are).
* A $600,000 grant from the [Berkeley Existential Risk Initiative](... |
2652941c-db77-48b3-b008-de1d29f9efb9 | StampyAI/alignment-research-dataset/arxiv | Arxiv | Characterizing and Detecting Mismatch in Machine-Learning-Enabled Systems
I Introduction
---------------
Despite advances in frameworks for machine learning (ML) model development and deployment, integrating models into production systems still remains a challenge [[4](#bib.bib27 "How to Bridge Machine Learning and... |
a8de7a25-4232-4d0e-9ce6-646771ef5024 | trentmkelly/LessWrong-43k | LessWrong | The two-headed bacterium
Cross-posted from substack.
I like to see categories as fish nets we use to capture ideas. We classify things into categories like individuals, nation or species, and of course it is all arbitrary and doesn’t correspond to anything in the real world. But categories still form useful chunks we... |
e9bce8d1-51f2-4745-90de-cd192359a0f8 | StampyAI/alignment-research-dataset/blogs | Blogs | New report: “UDT with known search order”
[](https://intelligence.org/files/UDTSearchOrder.pdf)Today we release a new technical report from MIRI research associate Tsvi Benson-Tilsen: “[UDT with known sea... |
a70dbf1b-c93b-4f07-a25f-6702b622046c | StampyAI/alignment-research-dataset/lesswrong | LessWrong | GPT-3: a disappointing paper
*[Note: I wrote this post in late May 2020, immediately after the GPT-3 paper was released.]*
This post is a compilation of two posts I recently made on tumblr.
For context: I have been an enthusiastic user of GPT-2, and have written a lot about it and transformer models more generally. ... |
5ee6cd1b-75b8-46d6-912b-55cebc5a1567 | trentmkelly/LessWrong-43k | LessWrong | History's Biggest Natural Experiment
Faced with the threat of COVID-19, most countries in the world are suddenly, dramatically improving their hygiene and infectious disease control. This won't just slow the spread of COVID-19. As a byproduct, it will also suppress almost every other contagious disease. This creates a... |
08093a69-56cc-47a6-a9dc-afe2155a286a | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Military Artificial Intelligence as Contributor to Global Catastrophic Risk
***Note**:* *the below is a chapter by Matthijs Maas, Kayla Matteucci, and Di Cooke (CSER), for a forthcoming book on new avenues of research on Global Catastrophic Risk. The chapter explores whether, or in what ways, the military use of AI co... |
3718d6f7-d512-4468-8d8d-8509ac80c008 | trentmkelly/LessWrong-43k | LessWrong | Parenting and Happiness
If you're trying to decide whether to have kids it would be helpful to know whether you should expect yourself to be happier childfree or as a parent. The press reports it both ways, passing on the results of conflicting studies. I would like to find a good summary by someone who has read all t... |
ab935c7e-086b-4a92-8005-20f841c8f87b | trentmkelly/LessWrong-43k | LessWrong | Book Review (mini): Co-Intelligence by Ethan Mollick
A brief and pleasant exploration of how AI is shaping the world around us, with tips and suggestions for educators, workers, and students to use AI as a Coworker, Tutor, Coach, or another Person. It is best seen as a business book or read by those with little AI exp... |
caa56329-6a55-415b-8d26-bbc7d23c7d45 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Regulate or Compete? The China Factor in U.S. AI Policy (NAIR #2)
*This post was published as part of the 2nd edition of* [*Navigating AI Risks*](https://navigatingairisks.substack.com/)*, a newsletter about the governance of transformative AI risks written by* [*@simeon\_c*](https://www.lesswrong.com/users/simeon_c?m... |
ab2d66da-c717-4b04-8280-d47e712a2251 | trentmkelly/LessWrong-43k | LessWrong | Book Review: Denial of Death
Denial of Death is the 1973 summation of anthropologist Ernest Becker's life's work studying human nature, building upon the work of the great psychologists of the 20th Century. It basically aims to be a grand unifying theory of psychology, and against all odds it kind of succeeds.
I came... |
dd6d29c2-b88b-4d52-993f-cdd78b9c6d5a | StampyAI/alignment-research-dataset/lesswrong | LessWrong | AISC project: SatisfIA – AI that satisfies without overdoing it
This project is part of the upcoming round of [AI Safety Camp](http://www.aisafety.camp/).
SatisfIA – AI that satisfies without overdoing it
=================================================
Summary
-------
This project will make a contribution to s... |
8bcf8c6b-ca29-4866-a001-8158140f8dfb | StampyAI/alignment-research-dataset/blogs | Blogs | April files
*By Katja Grace, 1 April 2021*
Today we are sharing with our blog readers a collection of yet-to-be-published drafts, in the hope of receiving feedback. We are especially looking for methodological critique, but all comments welcome!
[Human-level performance estimate](https://docs.google.com/document/d... |
3c53bc87-96a7-433d-950f-6d5b23ae46fc | trentmkelly/LessWrong-43k | LessWrong | Relevant to natural abstractions: Euclidean Symmetry Equivariant Machine Learning -- Overview, Applications, and Open Questions
Talk by Tess E. Smidt (MIT). Abstract:
> Atomic systems (molecules, crystals, proteins, etc.) are naturally represented by a set of coordinates in 3D space labeled by atom type. This is a ch... |
fc18506f-d58e-49db-84ae-5a5c2bc2be4d | StampyAI/alignment-research-dataset/blogs | Blogs | do not form your own opinion
> "your mind is a labyrinth of anti-cognitive-bias safeguards, huh?"
>
>
a friend, to me
do not form your own opinion
----------------------------
when confronted with an idea, it can be tempting to try to evaluate it. in fact, the brain does this by default.
however, i try to av... |
3e5bf22b-9bf4-466d-9643-e66554f079c7 | trentmkelly/LessWrong-43k | LessWrong | March-April 2025 Progress in Guaranteed Safe AI
Say hi at ICSE in Ottawa, I’ll be at the reception Thursday, this colocated event on Friday, and the LLM4Code workshop on Saturday.
As usual there are no benefits to the paid subscription.
Sorry for consolidating two months into one post again after I said I wouldn’t.... |
f5675ee3-09f7-4c71-91b5-f107bb12d785 | trentmkelly/LessWrong-43k | LessWrong | Cryonic suspension where?
I want to sign up for cryonic suspension. I haven't done so yet because I haven't been able to decide which organization to use. I'm not expecting you guys to choose for me, but it would be very helpful if those of you who are signed up (or will sign up) would say which organization you went ... |
64b349f6-d619-4d34-9fd4-1d8eafd10d74 | trentmkelly/LessWrong-43k | LessWrong | Nozick's fluff, or Omega the trickster.
"OK, so here is the plan: You and I bet on whether Nozick one-boxes or not, and then I tell him that I can predict what he chooses and set up the opaque box accordingly ahead of time." Hermes looked extremely pleased with his new idea.
"You can't seriously expect even someone a... |
dad588dd-b214-4977-b01c-fa6fbd7c837b | trentmkelly/LessWrong-43k | LessWrong | Two Cult Koans
A novice rationalist studying under the master Ougi was rebuked by a friend who said, “You spend all this time listening to your master, and talking of ‘rational’ this and ‘rational’ that—you have fallen into a cult!”
The novice was deeply disturbed; he heard the words You have fallen into a cult! reso... |
5447647f-cd7b-4b2b-86ca-24dbe49df6c2 | trentmkelly/LessWrong-43k | LessWrong | Watch Elon Musk’s Neuralink presentation
"Electric vehicles, rockets... and now brain-computer interfaces. Elon Musk's newest venture, Neuralink, aims to bridge the gap between humans and artificial intelligence by implanting tiny chips that can link up to the brain. At a press conference on July 16, Neuralink's ambit... |
581230f9-7baa-4c38-a948-7c3f1cb54f44 | trentmkelly/LessWrong-43k | LessWrong | How French intellectuals ruined the West - Postmodernism and its impact, explained
|
c2ded659-0ac4-4bfe-a9d1-266cfc136de8 | trentmkelly/LessWrong-43k | LessWrong | What are the best online tools for meetups and meetings?
tl;dr – If you have tried at least two telepresence tools, explain their relative strengths, and which one you prefer.
A key question for both remote organizations, and apparently all social life now, is "what tools can best keep me connected to people in the w... |
1cb0128d-0e8a-46f6-98cc-d2102298e12d | trentmkelly/LessWrong-43k | LessWrong | Question Gravity
Earlier today, I was in the shower. I had just finished a bottle of shampoo, and I tried balancing it on the thin railing of the shower door. I expected it to fall down, but instead it stayed there. From the side, it certainly looked like it was sticking out more than halfway over the edge. (Picture) ... |
0e6ff778-2361-48ae-b3d5-ab2b9aba9fb7 | trentmkelly/LessWrong-43k | LessWrong | Is Interpretability All We Need?
An LLM is a simulator for token-generation processes, generally ones that are human-like agents. You can fine-tune or RLHF it to preferentially create some sorts of agents (to generate a different distribution of agents than was in its pretraining data), such as more of ones that won't... |
f7e98a01-f5c9-4335-9e74-cf6b2ac84bae | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Robustness of Model-Graded Evaluations and Automated Interpretability
TL;DR Many evaluations and automated interpretability rely on using multiple models to evaluate and interpret each other. One model is given full access to the text output of another model in OpenAI's automated interpretability and many model-graded... |
112a221c-08c7-4d65-a567-545b44ccc0f9 | StampyAI/alignment-research-dataset/arxiv | Arxiv | RAFT: A Real-World Few-Shot Text Classification Benchmark
1 Introduction
---------------
 and Financial Times, among others, recently reviewed the book Race Against the Machine by economists Erik Brynjolfsson and Andrew McAfee. The FT reviewer writes:
> Pattern recognition, the authors think, will quickly allow machine... |
26f5b480-807e-47e1-89d8-64ec978a16eb | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Everything's normal until it's not
*Crossposted from the EA Forum:* [*https://forum.effectivealtruism.org/posts/2hduXN5MXCZPqKjSv/everything-s-normal-until-it-s-not*](https://forum.effectivealtruism.org/posts/2hduXN5MXCZPqKjSv/everything-s-normal-until-it-s-not)
*Note: This is a quick write-up of some thoughts I'... |
e4d7c7cd-7941-4f7c-a9e0-f0b11b1c3f5a | trentmkelly/LessWrong-43k | LessWrong | Intro to Naturalism: Orientation
A note on how to approach this sequence:
If you were exactly like me, I would ask you to savor this sequence, not scarf it. I would ask you to approach each of these essays in an expansive, lingering, thoughtful sort of mood. I would ask you to read them a little bit at a time, perhap... |
0c754813-4ad5-4416-8213-3edc1071cac0 | trentmkelly/LessWrong-43k | LessWrong | Wizards Implements Ochlocracy
Today, a player who recently made the Top 8 at a Grand Prix was banned for life from tournament Magic: The Gathering. His account on Magic Online was seized, and he will be given a check for what Wizards thinks that account was worth.
Wizards does not discuss the reasons for bans, but in... |
aa7e6fd1-96d8-4d75-9937-79d9eff46df6 | trentmkelly/LessWrong-43k | LessWrong | Intelligence Explosion analysis draft: From digital intelligence to intelligence explosion
Again, I invite your feedback on this snippet from an intelligence explosion analysis Anna Salamon and myself have been working on. This section is less complete than the others; missing text is indicated with brackets: [].
___... |
d9bdbbdf-8d6e-426d-9502-0c7197f1afcb | trentmkelly/LessWrong-43k | LessWrong | Derek Parfit, "On What Matters"
Derek Parfit has published his second book, "On What Matters". Here are reviews by Tyler Cowen and Peter Singer. |
867bb545-fea0-42b9-b93c-7df067b3202b | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Accurate Models of AI Risk Are Hyperexistential Exfohazards
(Where "an exfohazard" is [information which leads to bad outcomes if known by a large fraction of society](https://www.lesswrong.com/posts/yET7wbjjJZtpz6NF3/don-t-use-infohazard-for-collectively-destructive-info).)
Let us suppose that we've solved the techn... |
9694c9bb-ff3c-46b4-84e3-0d2b0ac63d37 | trentmkelly/LessWrong-43k | LessWrong | Dollars for Brains
Putanumonit is now accepting donations, and donations now let you rent my brain if you have use for it.
Original: https://putanumonit.com/2017/11/15/dollars-for-brains/ |
7f014301-85a7-48e5-9e52-9720c7a3cae7 | trentmkelly/LessWrong-43k | LessWrong | Antisocial personality traits predict utilitarian responses to moral dilemmas
So says the title of an interesting recent paper I stumbled on yesterday (ungated link; h/t Chris Bertram). Here's the abstract:
> Researchers have recently argued that utilitarianism is the appropriate framework by which to evaluate moral... |
7cac34a6-ddf9-440c-b51a-b981bfc15d88 | trentmkelly/LessWrong-43k | LessWrong | Lessons After a Couple Months of Trying to Do ML Research
I’m a 17 y/o who started doing transformer interpretability work around October 2021, mainly for the learning experience, but also with the goal of potentially finding something cool and interesting about transformers. I’m writing this post to consolidate some ... |
4926320a-d479-40bb-81ae-a20e12c2f29f | StampyAI/alignment-research-dataset/special_docs | Other | Anthropic uncertainty in the Evidential Blackmail
I’m currently writing a piece on anthropic uncertainty in Newcomb problems. The idea is that whenever someone simulates us to predict our actions, this leads us to have anthropic uncertainty about whether we’re in this simulation or not. (If we knew whether we were in ... |
1b8993e2-8921-4dd4-a2a1-97f01190dd9c | StampyAI/alignment-research-dataset/blogs | Blogs | right to death, therefore
right to death, therefore
-------------------------
because i like [freedom](defining-freedom.html) so much, i think people should be generally able to do what they want. but this immediately raises a conundrum: should someone be able to do an action that hampers their *future* freedom?
o... |
1cdeaff2-f6a6-409f-acd2-366a3204250a | trentmkelly/LessWrong-43k | LessWrong | Headphones hook
Some months ago I got a 3D printer. (An Anycubic Kobra Go, which was pretty high up the list of "best value for money for a starter printer" that I found through /r/3Dprinting at the time.) I haven't used it much, but recently I used it to solve an itch. I wanted my headphones to be in easy reach, so I... |
915e6e67-1d79-4a9b-8c54-91e2d3aef941 | trentmkelly/LessWrong-43k | LessWrong | [Link] Wavefunctions: from Linear Algebra to Spinors
I wrote this blogpost because I thought it took an excessive amount of digging to understand what a spinor was. My original motivation was to understand wavefunctions more concretely since I recently discovered that wavefunctions are spinor-valued, not (necessarily)... |
7328fc57-19cd-4042-b115-9a3146243069 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Melbourne Social Meetup
Discussion article for the meetup : Melbourne Social Meetup
WHEN: 20 July 2012 06:30:00PM (+1000)
WHERE: Ben's house - see mailing list - Carlton, VIC
Melbourne's next social meetup is on Friday, 20th of July. Once again, it'll be at Ben's house - see the mailing list for location d... |
2ae1ce0e-58d1-4016-adc5-9006d7f7e61b | trentmkelly/LessWrong-43k | LessWrong | Worrying less about acausal extortion
I've wanted a good version of this post to exist for awhile. I'm not sure how to write that post. But here's a okay-ish first draft, which seemed hopefully better than nothing. I've made it possible for anyone to write comments on the LessWrong doc editor version of the post, so i... |
bdbc73ef-8b76-4fd6-b02a-3e9c6fbc06c1 | trentmkelly/LessWrong-43k | LessWrong | How I got so excited about HowTruthful
This is the script for a video I made about my current full-time project. I think the LW community will understand its value better than the average person I talk to does.
Hi, I'm Bruce Lewis. I'm a computer programmer. For a long time, I've been fascinated by how computers can ... |
b3d9ffed-8547-4b57-a09e-744bf8b04dc1 | trentmkelly/LessWrong-43k | LessWrong | [Book Review] "Suffering-focused Ethics" by Magnus Vinding
(UPDATED DECEMBER 2021)
Magnus Vinding’s “Suffering-Focused Ethics” is a heartfelt, yet occasionally overreaching work laying out a case for why we should focus primarily on suffering in our ethical decisions, and exactly how we can approach the world with th... |
f699e6ed-bc7d-4f16-bae0-d067d361dd1c | trentmkelly/LessWrong-43k | LessWrong | Meetup : Stockholm: Mental contrasting
Discussion article for the meetup : Stockholm: Mental contrasting
WHEN: 21 October 2016 04:00:00PM (+0200)
WHERE: Lindstedtsvägen 3, Room 1537
Last week, we had only enough time for an intro to TAPs. This week, we will dive into mental contrasting, in a workshop held by David_... |
7b0a134c-2208-4972-83db-17440d273d38 | trentmkelly/LessWrong-43k | LessWrong | Coronavirus: Justified Practical Advice Thread
(Added: To see the best advice in this thread, read this summary.)
This is a thread for practical advice for preparing for the coronavirus in places where it might substantially grow.
We'd like this thread to be a source of advice that attempts to explain itself. This i... |
7367c16e-c700-409d-8e0a-f03b92395207 | trentmkelly/LessWrong-43k | LessWrong | How to discover the nature of sentience, and ethics
Note: This is a very rough draft of a theory. Any comments or corrections are welcome. I will try to give credit to useful observations.
To discover the nature of sentience and ethics, we can actually use scientific principles even though they are sometimes consider... |
8b25c386-015b-48f3-a625-acdf36fdc292 | trentmkelly/LessWrong-43k | LessWrong | Elon Musk and Solar Futurism
2024 is the year it became clear that we're actually going to do battery energy storage. That it was time to stop soberly reminding anyone too excited about solar that storage is an unsolved problem. In fact I personally became very excited about solar, and briefly thought that this year's... |
bf28939b-4c5e-48b1-a7d6-2bd78a38c90a | trentmkelly/LessWrong-43k | LessWrong | Sunlight is yellow parallel rays plus blue isotropic light
When you look up the color temperature of daylight, most sources will say 6500K, but if you buy an LED with that color temperature, it will not look like the sun in the sky. It will seem bluer (or, less yellow-y). Yet, 6500K is arguably the correct number. Wha... |
bca95a99-3609-467b-a721-7f391045d5ab | trentmkelly/LessWrong-43k | LessWrong | Based Beff Jezos and the Accelerationists
It seems Forbes decided to doxx the identity of e/acc founder Based Beff Jezos. They did so using voice matching software.
Given Jezos is owning it given that it happened, rather than hoping it all goes away, and people are talking about him, this seems like a good time to co... |
ba90cca9-fa44-46a8-b04c-8dad0e22f9ee | trentmkelly/LessWrong-43k | LessWrong | Goals vs. Rewards
Related: Terminal Values and Instrumental Values, Applying behavioral psychology on myself
Recently I asked myself, what do I want? My immediate response was that I wanted to be less stressed, particularly for financial reasons. So I started to affirm to myself that my goal was to become wealthy, an... |
1b63b4d5-7018-4c0c-b3dc-65eac93948c5 | StampyAI/alignment-research-dataset/blogs | Blogs | New paper: “Formalizing convergent instrumental goals”
[](https://intelligence.org/files/FormalizingConvergentGoals.pdf)Tsvi Benson-Tilsen, a MIRI associate and UC Berkeley PhD candidate, has written a paper with contributions from MIRI Ex... |
5eb9f451-9c6f-4894-9de0-8179769cf93a | trentmkelly/LessWrong-43k | LessWrong | Don’t warn nonspecifically!
This is a decent warning sign. Image via Wikipedia
I hate safety warnings. It’s not that I’m hurt by someone out there’s condescending belief that I can’t work out whether irons are for drying children. And I welcome the endless mental accretion of terrifying facts about obscure ways one c... |
287e57ec-2b32-469c-9bac-8f08edb87978 | trentmkelly/LessWrong-43k | LessWrong | Self-fulfilling misalignment data might be poisoning our AI models
> Your AI’s training data might make it more “evil” and more able to circumvent your security, monitoring, and control measures. Evidence suggests that when you pretrain a powerful model to predict a blog post about how powerful models will probably ha... |
fc94ad4e-6340-42f3-91cd-6981e2b202a3 | trentmkelly/LessWrong-43k | LessWrong | AI #55: Keep Clauding Along
Things were busy once again, partly from the Claude release but from many other sides as well. So even after cutting out both the AI coding agent Devin and the Gladstone Report along with previously covering OpenAI’s board expansion and investigative report, this is still one of the longest... |
da1f0080-fc0b-4a37-9b4f-288da7d972c3 | trentmkelly/LessWrong-43k | LessWrong | Weekly LW Meetups
This summary was posted to LW Main on August 1st. The following week's summary is here.
Irregularly scheduled Less Wrong meetups are taking place in:
* Perth, Australia: Introducing Less Wrong: 05 August 2014 06:00PM
The remaining meetups take place in cities with regular scheduling, but involve ... |
4168695d-b3d0-4f7d-a955-0cc0ea86b5f7 | StampyAI/alignment-research-dataset/special_docs | Other | Special issue on learning for human–robot collaboration
Autonomous Robots
https://doi.org/10.1007/s10514-018-9756-z
Special issue on learning for human–robot collaboration
Leonel Rozo1·Heni Ben Amor2·Sylvain Calinon3·Anca Dragan4·Dongheui Lee5,6
© Springer Science+Business Media, LLC, part of Springer Nature 2018
Once... |
520c3cce-b0f1-4263-9c33-dc5e8a63a919 | LDJnr/LessWrong-Amplify-Instruct | LessWrong | "“Psychophysics,” despite the name, is the respectable field that links physical effects to sensory effects. If you dump acoustic energy into air—make noise—then how loud does that sound to a person, as a function of acoustic energy? How much more acoustic energy do you have to pump into the air, before the noise sound... |
8d17b0d2-9e39-4df5-8e41-969d8e95a42c | trentmkelly/LessWrong-43k | LessWrong | AI Safety Newsletter #3: AI policy proposals and a new challenger approaches
Welcome to the AI Safety Newsletter by the Center for AI Safety. We discuss developments in AI and AI safety. No technical background required.
Subscribe here to receive future versions.
---
Policy Proposals for AI Safety
Critical industr... |
dfc8dedd-d588-4ed5-9442-3b05cd27e683 | trentmkelly/LessWrong-43k | LessWrong | Notes from a conversation on act-based and goal-directed systems
I had a conversation with Nate about different possible goal systems for agents, and I think some people will be interested in reading this summary.
Goal specification
I started by stating my skepticism about approaches to goal specification that rely ... |
7f51ebc0-95c7-4c32-baf9-42506cf7de31 | trentmkelly/LessWrong-43k | LessWrong | Two Challenges for ELK
This post sketches two challenges to ARC's project around eliciting latent knowledge that differ somewhat in kind from the challenges ARC is most concerned about. They relate to the difficulty in distinguishing beliefs from other representations.
Introduction
The problem of ELK, as outlined in... |
31abcd4d-226b-4f22-b546-31cfc16a4261 | StampyAI/alignment-research-dataset/arbital | Arbital | Join and meet: Exercises
Try these exercises to test your knowledge of joins and meets.
Tangled up
--------------------
Determine whether or not the following joins and meets exist in the poset depicted by the above Hasse diagram.
$c \vee b$
%%hidden(Show so... |
cd9de747-2610-40ad-8b22-809fe9c75520 | trentmkelly/LessWrong-43k | LessWrong | How familiar is the Lesswrong community as a whole with the concept of Reward-modelling?
I initially assumed that the concept of reward-modelling would be something most Lesswrongers were very familiar with. After all, this is one of the best communities for conversing on the topic. And a large percentage of all posts... |
ab245955-2700-42a2-9fe4-5caa33c9b8e2 | trentmkelly/LessWrong-43k | LessWrong | LW resources on childhood experiences?
Does LW have a curated or tagged set of resources on childhood experiences including "trauma"? Personally I'm interested in perspectives from developing world countries (I belong to one), but I'm totally interested in different people coming at this topic from different angles.
... |
45e192c0-5281-4820-bba5-71f49b84fa81 | trentmkelly/LessWrong-43k | LessWrong | Why “Solving Alignment” Is Likely a Category Mistake
A common framing of the AI alignment problem is that it's a technical hurdle to be overcome. A clever team at DeepMind or Anthropic would publish a paper titled "Alignment is All You Need," everyone would implement it, and we'd all live happily ever after in harmoni... |
ee3149fc-014d-42bd-a7e8-7afd5c08361c | StampyAI/alignment-research-dataset/arxiv | Arxiv | Dynamic Safe Interruptibility for Decentralized Multi-Agent Reinforcement Learning
1 Introduction
---------------
Reinforcement learning is argued to be the closest thing we have so far to reason about the properties of
*artificial general intelligence* [[8](#bib.bib8)]. In 2016, Laurent Orseau (Google DeepMind) an... |
5cebb426-5317-4608-bf01-37f2d1efebd1 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Takeoff speeds have a huge effect on what it means to work on AI x-risk
The slow takeoff hypothesis predicts that AGI emerges in a world where powerful but non-AGI AI is already a really big deal. Whether AI is a big deal right before the emergence of AGI determines many super basic things about what we should think o... |
4a7ed9b4-211c-4a80-a314-cdb2f1493b0f | trentmkelly/LessWrong-43k | LessWrong | Safe Search is off: root causes of AI catastrophic risks
Epistemic status: My best guess
When I look at advanced AI development, I see three general conditions that seem to be the root causes of all catastrophic risks:
* reliance on deep learning without knowing how to do it safely,
* pressure to make progress on ... |
00e00afc-6a4b-42a9-8337-392874d1c36a | trentmkelly/LessWrong-43k | LessWrong | Show, not tell: GPT-4o is more opinionated in images than in text
Epistemic status: This should be considered an interim research note. Feedback is appreciated.
Introduction
We increasingly expect language models to be ‘omni-modal’, i.e. capable of flexibly switching between images, text, and other modalities in ... |
de39e220-d07a-474f-8330-c260faac565f | trentmkelly/LessWrong-43k | LessWrong | COVID/Delta advice I'm currently giving to friends
[Epistemic status: US-centric. gathering various impressions even though I haven't deeply investigated this topic and am not an expert, so it's all in one place and others can better use or critique the claims.]
1. Figure things out for yourself
Figure out what r... |
5193731b-5f2c-4383-bc88-a176e34c9ba1 | trentmkelly/LessWrong-43k | LessWrong | New Rationality Blog: 'Measure of Doubt'
Measure of Doubt is a new rationality blog from Julia and Jesse Galef (a sister-brother team), aimed at a much broader audience than Less Wrong.
Julia is the author of the recently-linked article Are Intuitions Good Evidence? and many other writings. I know less about Jesse, b... |
e6655bc3-7161-4ff0-90d3-647868a9061e | trentmkelly/LessWrong-43k | LessWrong | Weekly LW Meetups
This summary was posted to LW Main on February 5th. The following week's summary is here.
Irregularly scheduled Less Wrong meetups are taking place in:
* Australia Online Hangout: 06 February 2016 07:30PM
* European Community Weekend: 02 September 2016 03:35PM
* San Antonio Meetup Discussion (To... |
4847a7e1-8855-41a4-b104-d0583ea58933 | trentmkelly/LessWrong-43k | LessWrong | Against Utilitarianism: Sobel's attack on judging lives' goodness
Luke tasked me with researching the following question
> I‘d like to know if anybody has come up with a good response to any of the objections to ’full information’ or ‘ideal preference’ theories of value given in Sobel (1994). (My impression is “no.”)... |
fe7d879a-74e0-433f-80d2-8ec804a1af60 | trentmkelly/LessWrong-43k | LessWrong | Anthropic is Quietly Backpedalling on its Safety Commitments
The company released a model it classified as risky — without meeting requirements it previously promised
This is the full text of a post first published on Obsolete, a Substack that I write about the intersection of capitalism, geopolitics, and artificial i... |
f3071262-f78a-4dea-b3dd-939d77218a4d | trentmkelly/LessWrong-43k | LessWrong | Arbitrage Drains Worse Markets to Feeds Better Ones
Also, why one account keeps running dry when you try to arbitrage two markets.
I was thinking about how inter-market arbitrage might affect one's account balances & the total amount of money on the markets arbitraged.
## The arbitrage.
Let's say I have accounts o... |
686272a9-c5c0-4d8b-82d5-1c2c52ffca56 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Thinking soberly about the context and consequences of Friendly AI
The project of Friendly AI would benefit from being approached in a much more down-to-earth way. Discourse about the subject seems to be dominated by a set of possibilities which are given far too much credence:
* A single AI will take over the world... |
2cf831ea-271e-4398-a324-5e49a9302299 | trentmkelly/LessWrong-43k | LessWrong | Meetup : BYU-I Meetup
Discussion article for the meetup : BYU-I Meetup
WHEN: 29 April 2015 03:00:00PM (-0600)
WHERE: Rexburg, Idaho: McKay Library Commons, by the dictionaries
Optimization
Discussion article for the meetup : BYU-I Meetup |
6648f217-8e59-477e-824c-b49bf092bb1b | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Conditioning Predictive Models: Large language models as predictors
*This is the first of seven posts in the [Conditioning Predictive Models Sequence](https://www.alignmentforum.org/s/n3utvGrgC2SGi9xQX) based on the paper “[Conditioning Predictive Models: Risks and Strategies](https://arxiv.org/abs/2302.00805)” by Eva... |
b2116501-c718-401e-84f5-5d609dcae7d1 | trentmkelly/LessWrong-43k | LessWrong | Virtual Machine Learning Conferences: The Good and the Bad
We are now almost two years into the COVID-19 pandemic: international travel has slowed to a trickle and all machine learning conferences have moved online. By now, most of my PhD research has taken place during the pandemic, which I’ve presented at four onlin... |
d927da49-1387-4480-a4e4-62639facc327 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Formalizing the "AI x-risk is unlikely because it is ridiculous" argument
There is a lot of good writing on technical arguments against AI x-risk (such as [Where I agree and disagree with Eliezer](https://www.lesswrong.com/posts/CoZhXrhpQxpy9xw9y/where-i-agree-and-disagree-with-eliezer) (which mostly argues for more u... |
49fc8f54-9ba1-444c-906e-75cc8f29c3a1 | trentmkelly/LessWrong-43k | LessWrong | Issues with Giving Multiplier
See the end of this post for a response from Giving Multiplier.
Giving Multiplier launched last fall to promote effective giving. You may have read about it at the time in Future Perfect, but I learned about it more recently, from Giving What We Can. Since it aims to convince people to d... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.