id
stringlengths
36
36
source
stringclasses
15 values
formatted_source
stringclasses
13 values
text
stringlengths
2
7.55M
89b5cde0-36b9-4fc5-86c8-142fb0a3910f
trentmkelly/LessWrong-43k
LessWrong
Has the Symbol Grounding Problem just gone away? A few years ago, the symbol grounding problem was widely considered a significant challenge in AI discussions. I believed that it would likely be addressed as a side effect of capability improvements, without requiring specific breakthroughs or attention, unlike others ...
23839ee0-0e26-40e6-b93a-18741c5a4a52
trentmkelly/LessWrong-43k
LessWrong
Disappointing Table Refinishing Shaving cream on a table: good kids activity on a wet day, right? Unfortunately it turns out our table's finish is alcohol-soluble, and spending a couple hours covered with a high-alcohol foam was not good for it: The center leaf I made, and is polyurethane, but we decided the...
fbb4d078-aee5-45f9-ac2a-7ab144148dcf
trentmkelly/LessWrong-43k
LessWrong
Betadine oral rinses for covid and other viral infections Before we get started, this is your quarterly reminder that I have no medical credentials and my highest academic credential is a BA in a different part of biology (with a double major in computer science). In a world with a functional medical system no one wou...
4d11506f-9db3-4da1-9f3e-473ffd81426e
trentmkelly/LessWrong-43k
LessWrong
GiveWell event for SF Bay Area EAs Passing this announcement along from GiveWell: > GiveWell is holding an event at our offices in San Francisco for Bay Area residents who are interested in Effective Altruism. The evening will be similar to the research events we hold periodically for GiveWell donors: it will include...
621e79ec-0026-499a-bcdc-bceed5ac8f23
trentmkelly/LessWrong-43k
LessWrong
Epistemological Vigilance for Alignment This post is part of the work done at Conjecture. Nothing hampers Science and Engineering like unchecked assumptions. As a concrete example of a field ridden with hidden premises, let's look at sociology. Sociologist must deal with the feedback of their object of study (people...
9529be0b-4fe6-48ed-b171-c56888c832bc
trentmkelly/LessWrong-43k
LessWrong
What do we already have right? I was just wondering. Human minds are messed up in 1001 ways, but are there a few rational principles that most people already have down? Of course, the answers to this question are probably so extremely obvious that I haven't even considered them. But I ask all the same.
bd467f82-e640-43c8-b655-175ebd530752
StampyAI/alignment-research-dataset/lesswrong
LessWrong
Can You Upload Your Mind & Live Forever? From Kurzgesagt - In a Nutshell An interesting video on mind uploading from almost 2 years ago, focusing on explaining the concept to the wider world. It's fairly simplified, as expected of a video that has to cross large inferential distances, and the Kurzgesagt channel is bas...
53aa5862-9fae-4ba9-be97-29aa6b7fae12
trentmkelly/LessWrong-43k
LessWrong
Have the lockdowns been worth it? There’s been a lot of discussion about whether the pandemic lockdowns have been worth it. However, much of the reasoning that we’ve seen has been very motivated and un-nuanced in a way that for us has distorted a lot of the information. So this is not a thread for taking a position o...
57cc01e5-dd7e-41b3-bb97-62d8507f744f
trentmkelly/LessWrong-43k
LessWrong
This prompt (sometimes) makes ChatGPT think about terrorist organisations Yesterday, I couldn't wrap my head around some programming concepts in Python, so I turned to ChatGPT (gpt-4o) for help. This evolved into a very long conversation (the longest I've ever had with it by far), at the end of which I pasted around 6...
0c7670dd-e93d-4507-88d0-a25cbe5ce731
trentmkelly/LessWrong-43k
LessWrong
Tips and Tricks for Answering Hard Questions I've collected some tips and tricks for answering hard questions, some of which may be original, and others I may have read somewhere and forgotten the source of. Please feel free to contribute more tips and tricks, or additional links to the sources or fuller explanations....
8f64fed1-41e2-42b8-a65d-ecfdee42a23d
trentmkelly/LessWrong-43k
LessWrong
Against Time in Agent Models When programming distributed systems, we always have many computations running in parallel. Our servers handle multiple requests in parallel, perform read and write operations on the database in parallel, etc. The prototypical headaches of distributed programming involve multiple processe...
3a3067f7-c62b-47ac-aae8-40e5148b5998
trentmkelly/LessWrong-43k
LessWrong
In the Short-Term, Why Couldn't You Just RLHF-out Instrumental Convergence? To the question "Do you expect instrumental convergence to become a big pain for AGI labs within the next two years?", about a quarter of the 400 people who answered to the poll I ran said "Yes".  I would like to hear people's thoughts on ho...
0a96883a-fdff-43d4-9fdf-7588f2dcf578
trentmkelly/LessWrong-43k
LessWrong
A different take on the Musk v OpenAI preliminary injunction order The judge in the Musk v OpenAI[1] case out of the northern district of California has issued an order on Musk's motion for preliminary injunction, which asked for an order temporarily stopping OpenAI from taking further steps towards its possible conve...
c4553be3-5de0-4cd1-90ef-f4df81a2ab86
StampyAI/alignment-research-dataset/eaforum
Effective Altruism Forum
A Windfall Clause for CEO could worsen AI race dynamics Summary ------- This is a response to the [Windfall Clause proposal from Cullen O’Keefe et al](https://www.fhi.ox.ac.uk/wp-content/uploads/Windfall-Clause-Report.pdf), which aims to make AI firms promise to donate a large fraction of profits if they become extre...
c59b4049-a660-4073-8f1b-534d3cf672a2
trentmkelly/LessWrong-43k
LessWrong
A Basic Problem of Ethics: Panpsychism? Panpsychism seems like a plausible theory of consciousness.  It raises extreme challenges for establishing reasonable ethical criteria. It seems to suggest that our ethics is very subjective: the "expanding circle" of Peter Singer would eventually (ideally) stretch to encompass...
ef998cc5-941c-4e43-ae95-ea763abc70e8
trentmkelly/LessWrong-43k
LessWrong
Machines of Faithful Obedience [Crossposted on Windows On Theory]   Throughout history, technological and scientific advances have had both good and ill effects, but their overall impact has been overwhelmingly positive. Thanks to scientific progress, most people on earth live longer, healthier, and better than they ...
f0dd066c-6941-4ce8-9a6c-4a1009a9da94
trentmkelly/LessWrong-43k
LessWrong
A Look Inside a Frequentist Today I experienced the Sequences post Beautiful Probability for the first time. I will begin by quoting the quotation that Eliezer began with: > Let me introduce this issue by borrowing a complaint of the late great Bayesian Master, E. T. Jaynes (1990): > > "Two medical researchers use t...
1477093f-dcdb-4704-bdf0-655f19766171
trentmkelly/LessWrong-43k
LessWrong
[LINK] A Turing test for free will Seth Lloyd has posted a well-written pre-print, proposing a self-administered Turing test for free will but also dealing with some other aspects of the free will debate. Some excerpts: > ... the theory of computation implies that even when our decisions arise from a completely deter...
84971841-8430-4cee-83b6-bdc43f2ab680
trentmkelly/LessWrong-43k
LessWrong
The Maker of MIND After my first rebirth - my mind labile, full of that indescribable sense of renewal, everything new again, everything fresh, everything tinged with a subtle hilarity - I assisted Bartosz Sumner, now rather well-known in some parts as the author of The Making of MIND.  Bart had lived many more lives...
5fcb5529-e2a3-4454-a586-6ff68893c464
trentmkelly/LessWrong-43k
LessWrong
Meetup : Moscow meetup: discussion on community evolution; power and culture talk; rational games Discussion article for the meetup : Moscow meetup: discussion on community evolution; power and culture talk; rational games WHEN: 23 November 2015 02:00:00PM (+0300) WHERE: ул. Большая Дорогомиловская, д.5к2 Regular b...
37e9dec3-ef07-491c-a0d7-eaa72827b116
StampyAI/alignment-research-dataset/blogs
Blogs
Data, Architecture, or Losses: What Contributes Most to Multimodal Transformer Success? The ability to ground language to vision is a fundamental aspect of real-world AI systems; it is useful across a range of tasks (*e.g.*, visual question answering) and applications (*e.g.*, generating descriptions for visually impa...
6e869199-bd4d-4ed3-9909-ed92fca9c752
trentmkelly/LessWrong-43k
LessWrong
Kantian baby rats I've often wished for a list of cases where philosophy has proven useful, or has at least anticipated science in drawing correct conclusions.  Here's one for the list: The June 18 2010 Science has two very similar articles on how rat brains represent space.  Both conclude that the brain already repr...
c82981a6-08c5-4abb-bc69-540cf00b7e71
trentmkelly/LessWrong-43k
LessWrong
A Variance Indifferent Maximizer Alternative TLDR This post explores creating an agent which tries to make a certain number of paperclips without caring about the variance in the number it produces, only the expected value. This may be safer than one which wants to reduce the variance, since humans are a large source ...
a563e763-d544-4c9f-b721-8a3d0e2b651c
StampyAI/alignment-research-dataset/arbital
Arbital
0.999...=1 Although some people find it counterintuitive, the [decimal expansions](https://arbital.com/p/4sl) $0.999\dotsc$ and $1$ represent the same [https://arbital.com/p/-4bc](https://arbital.com/p/-4bc). # Informal proofs These "proofs" can help give insight, but be careful; a similar technique can "prove" that...
b08efadb-3350-4f84-814e-207c8d7aec90
trentmkelly/LessWrong-43k
LessWrong
Chapter 36: Status Differentials Wrenching disorientation, that was how it felt to walk out of Platform Nine and Three-Quarters into the rest of Earth, the world that Harry had once thought was the only real world. People dressed in casual shirts and pants, instead of the more dignified robes of wizards and witches. S...
86faf473-0e94-4d47-8935-9a22636e2748
trentmkelly/LessWrong-43k
LessWrong
Books on evolution of conscience I am looking for books on evolution of conscience. Please suggest. I searched but nothing good came out. I don't know if this is the right place for such requests. If this is not the right place, pls tell me, i will delete this post. Thanks.  
ce6a2d9f-63b0-483b-9586-9b795e64f847
trentmkelly/LessWrong-43k
LessWrong
How long should I delay my second shot? I got my first shot of Pfizer.  They automatically scheduled my second appointment for 3 weeks later, but I canceled it, with the plan to reschedule some other time.  Two reasons for this: first, while I can't unilaterally impose First Doses First, I can at least impose First Do...
058f9cb1-9f85-4248-a156-4004def50d3a
trentmkelly/LessWrong-43k
LessWrong
Should effective altruism be more "cool"? As a recent entrant to the sphere of rationality and effective altruism, I often wonder if the primarily-nerdy audience & community (which I adore) is less than a boon to the objectives of effective altruism. If a popularization (through social media platforms such as TikTok)...
417e1e41-0f53-4b36-b9ff-94c7a92f6d6f
StampyAI/alignment-research-dataset/eaforum
Effective Altruism Forum
Implications of AGI on Subjective Human Experience **Introduction** ================ Humanity is heading towards unprecedented circumstances in a destined rendezvous of two distinct forces. The continuation of artificial intelligence (AI) development toward artificial general intelligence (AGI) at the current rate an...
3e59ea4f-85f4-454e-9a5d-dab42c4f0c6e
trentmkelly/LessWrong-43k
LessWrong
Newcomblike problem: Counterfactual Informant I want to propose a variant of the Counterfactual Mugging problem discussed here.  BE CAREFUL how you answer, as it has important implications, which I will not reveal until the known dumb humans are on record. Here is the problem: Clipmega is considering whether to reve...
b69120da-5181-4725-872f-3436c80e4ad5
trentmkelly/LessWrong-43k
LessWrong
Offenders' deadly thoughts may hold answer to reducing crime
7b08b02f-5db3-4b49-8f27-af9916a0858b
StampyAI/alignment-research-dataset/lesswrong
LessWrong
Algo trading is a central example of AI risk I suspect this observation is far from original, but I didn't see it explicitly said anywhere, so it seemed worthwhile spelling it out. The paperclip maximizer is a popular example of unfriendly AI, however it's not the most realistic one (obviously it wasn't meant to be)....
3b1e415b-17dc-4482-a87a-964f65b0f532
StampyAI/alignment-research-dataset/lesswrong
LessWrong
White House Announces "New Actions to Promote Responsible AI Innovation" Highlights ========== * Vice President Kamala Harris will be meeting with officials from Alphabet, Anthropic, Microsoft, and OpenAI to "emphasize the importance of driving responsible, trustworthy, and ethical innovation with safeguards that mit...
c0f4e5c9-0a28-4b53-81e8-1336225da130
trentmkelly/LessWrong-43k
LessWrong
Meetup : Fort Collins, Colorado Meetup Wednesday 7pm Discussion article for the meetup : Fort Collins, Colorado Meetup Wednesday 7pm WHEN: 02 November 2011 07:00:00PM (-0600) WHERE: Bean Cycle, 144 North College Avenue, Fort Collins, CO 80524 Another week, another weekly meetup. Stay warm with community, now that w...
f639f2db-2705-43df-a033-fab7a2dd1f31
trentmkelly/LessWrong-43k
LessWrong
Weather and climate forecasting: how the challenges differ by time horizon Prelude: Climate change, in particular the question of anthropogenic global warming (AGW) is both an intellectually complex and a politically loaded topic. Politics has been called the mind-killer here. For a mix of both reasons (the intellectu...
bd7816d5-fd40-459d-8c47-a4b1425d2388
trentmkelly/LessWrong-43k
LessWrong
Mark Manson and Rationality As those of you on the Less Wrong chat may know, Mark Manson is my favourite personal development author. I thought I'd share those articles that are most related to rationality, as I figured that they would have the greatest chance of being appreciated.   Immediately after writing this a...
51e3c572-5468-4ca6-9d35-9f6cedeeae39
StampyAI/alignment-research-dataset/arxiv
Arxiv
Inverse reinforcement learning for video games 1 Introduction --------------- Deep reinforcement learning has achieved great success at optimizing known reward functions in a range of challenging environments. However, designing an appropriate reward function is difficult even for experienced engineers, and is impo...
8db157ec-5dff-474b-aebf-386f652f7834
trentmkelly/LessWrong-43k
LessWrong
[link] Etzioni: AI will empower us, not exterminate us https://medium.com/backchannel/ai-wont-exterminate-us-it-will-empower-us-5b7224735bf3 (Slashdot discussion: http://tech.slashdot.org/story/14/12/10/1719232/ai-expert-ai-wont-exterminate-us----it-will-empower-us) Not sure what the local view of Oren Etzioni or th...
aa08d8cc-b7ab-4555-bcc3-18facff53161
trentmkelly/LessWrong-43k
LessWrong
Elon Musk is wrong: Robotaxis are stupid. We need standardized rented autonomous tugs to move customized owned unpowered wagons. Sorry for the flashy headline, but I genuinely feel this might be the best idea I ever had. After the invention of the horseless carriage, it supposedly took people years to realize that wi...
b153be8a-ba4f-4795-aa9a-f3b6b9b36170
StampyAI/alignment-research-dataset/arxiv
Arxiv
Meta-Learning: A Survey 1 Introduction --------------- When we learn new skills, we rarely - if ever - start from scratch. We start from skills learned earlier in related tasks, reuse approaches that worked well before, and focus on what is likely worth trying based on experience (Lake et al., [2017](#bib.bib87)). ...
c6641644-3228-4416-a290-7752198b95c8
StampyAI/alignment-research-dataset/special_docs
Other
The Complexity of Decentralized Control of Markov Decision Processes 1. Introduction. Markov decision processes (MDPs) have received considerable attention, and there exist well-known algorithms for finding optimal control strategies in the case where a process is centrally controlled and the controller (or agent) has...
765c2ab2-869e-4064-a961-f5b213b5030a
trentmkelly/LessWrong-43k
LessWrong
The Case for AI Optimism This is a very comprehensive article sharing what could be considered a minority point of view for Less Wrong readers.  The article is very comprehensive, I suggest that if you are well versed in AI and labor economics, start reading from the New Work section. 
a95628d3-27a5-4ebe-a7e5-4409c1f767b0
trentmkelly/LessWrong-43k
LessWrong
Trying harder works, sometimes It seems to be commonly accepted that pressing the "Try Harder" button does not work. While I agree with this in general, I would argue that there are edge cases where trying harder is a valid option. To illustrate this, is a story: of science, and how I used it to change a particularly...
dcdf31b5-010e-49c1-a5a9-f3548c717784
trentmkelly/LessWrong-43k
LessWrong
Self-confidence and status I've seen the advice "be (more) confident" given so that the person may become more socially successful. For example self-confident get more raises. But I'm not sure if self-confidence is the cause of becoming more socially successful, or the result of it. I don't think self-confidence can b...
29a8c647-e942-4cac-a947-098754c7ca48
StampyAI/alignment-research-dataset/alignmentforum
Alignment Forum
Motivating Abstraction-First Decision Theory Let’s start with a prototypical embedded decision theory problem: we run two instances of the same “agent” from the same source code and in the same environment, and the outcome depends on the choices of both agents; both agents have full knowledge of the setup. ![](https:...
de873c04-7eab-445c-8034-9aaa7d45d23f
trentmkelly/LessWrong-43k
LessWrong
A probabilistic off-switch that the agent is indifferent to Edit: I no longer think this post deserves attention. Abstract This post presents a setup with an off-switch that is defective with probability of almost 0. The agent is indifferent to being terminated in worlds where the off-switch works. Also, the agent d...
6139e55d-4012-461f-9b27-3ae62a208b0e
trentmkelly/LessWrong-43k
LessWrong
Goals selected from learned knowledge: an alternative to RL alignment   Summary:  Alignment work on network-based AGI focuses on reinforcement learning. There is an alternative approach that avoids some, but not all, of the difficulties of RL alignment. Instead of trying to build an adequate representation of the beh...
4f68acf6-4004-4012-8cb6-74b7efc0efc0
trentmkelly/LessWrong-43k
LessWrong
Welcome to Lesswrong 2.0 Lesswrong 2.0 is a project by Oliver Habryka, Ben Pace, and Matthew Graves with the aim of revitalizing the Lesswrong discussion platform. Oliver and Ben are currently working on the project full-time and Matthew Graves is providing part-time support and oversight from MIRI. Our main goals ar...
274773ec-2bc7-4c67-9e2b-0877c1f193b1
trentmkelly/LessWrong-43k
LessWrong
Brainrot January: In early 2026, Meta launches a fleet of new AI influencers, targeting the massive audience displaced by the Xiaohongshu-TikTok wars. They are beautiful, funny, smart—whatever you want them to be. Equipped with the latest in online learning, the agents immediately begin adapting to social media trends...
e30c305d-27a3-4c5c-a9e1-57125e3283be
trentmkelly/LessWrong-43k
LessWrong
Brainstorm of things that could force an AI team to burn their lead Comments: The following is a list (very lightly edited with help from Rob Bensinger) I wrote in July 2017, at Nick Beckstead’s request, as part of a conversation we were having at the time. From my current vantage point, it strikes me as narrow and ob...
5b35a425-d25a-489b-808b-c61902271b32
StampyAI/alignment-research-dataset/blogs
Blogs
Difficulty of Predicting the Maximum of Gaussians Suppose that we have a random variable ![X \in \mathbb{R}^d](https://s0.wp.com/latex.php?latex=X+%5Cin+%5Cmathbb%7BR%7D%5Ed&bg=f0f0f0&fg=555555&s=0&c=20201002), such that ![\mathbb{E}[XX^{\top}] = I_{d \times d}](https://s0.wp.com/latex.php?latex=%5Cmathbb%7BE%7D%5BXX%...
f7cc162d-0204-4b47-b3c6-50a0b344a247
trentmkelly/LessWrong-43k
LessWrong
Fragile Universe Hypothesis and the Continual Anthropic Principle - How crazy am I? Personal Statement I like to think about big questions from time to time. A fancy that quite possibly causes me more harm than good. Every once in a while I come up with some idea and wonder "hey, this seems pretty good, I wonder if an...
2ba29a68-b989-497f-9728-11d01d75fcec
StampyAI/alignment-research-dataset/blogs
Blogs
MIRI’s August 2014 newsletter | | | | | | | | | | --- | --- | --- | --- | --- | --- | --- | --- | | | | | --- | | | | | | | --- | --- | | | | | --- | | [Machine Intelligence Research Institute](http://intelligence.org) | | | | | | | --- | --- | | | | | --- | | Dear friends, Our [summe...
089fe18f-de7d-4874-ab92-a76b9347005b
trentmkelly/LessWrong-43k
LessWrong
Interested in Cognitive Bootcamp? I'm running more 4-day "Cognitive Bootcamps" over the next couple months (during Lighthaven Eternal September season). DM me if you're potentially interested (either as an individual, or as a team). The workshop is most valuable to people who: * control their decisionmaking process...
9491b6a8-99c9-49dd-91f6-7d19acf806e4
trentmkelly/LessWrong-43k
LessWrong
lessannoying.org What does it mean?
b9fe0a58-3c60-44a6-bf88-c6ecce5dbc01
trentmkelly/LessWrong-43k
LessWrong
Progress links and tweets, 2023-06-07 Announcements & opportunities * UK Great Stagnation summit in July, apply to attend (via @s8mb) * Help Bryan Bishop create an enzyme that makes DNA based on digital instructions * The Cybersecurity Grant Program, $1M for AI-based cybersecurity (from @OpenAI) * Foundations & Fr...
e29943e7-c045-49a6-9723-90dae4daf88b
trentmkelly/LessWrong-43k
LessWrong
Bayesian Epistemology vs Popper     I was directed to this book (http://www-biba.inrialpes.fr/Jaynes/prob.html) in conversation here: http://lesswrong.com/lw/3ox/bayesianism_versus_critical_rationalism/3ug7?context=1#3ug7 I was told it had a proof of Bayesian epistemology in the first two chapters. One of the thin...
86c4dda1-6182-478b-bb44-baa542c85b13
trentmkelly/LessWrong-43k
LessWrong
The Valentine’s Day Gift That Saves Lives This is mainly of interest to Effective Altruism-aligned Less Wrongers. Thanks to Agnes Vishnevkin, Jake Krycia, Will Kiely, Jo Duyvestyn, Alfredo Parra, Jay Quigley, Hunter Glenn, and Rhema Hokama for looking at draft versions of this post. At least one aspiring rationalist w...
913ae624-1cd7-481e-82e7-8a0be8cf33ec
trentmkelly/LessWrong-43k
LessWrong
cancer rates after gene therapy Here's a new paper on cancer developing in patients after receiving gene therapy with eli-cel, brand name Skysona. What is eli-cel? Stem cells are isolated and modified by using a lentiviral vector to add the ABCD1 gene. The stem cells in bone marrow are killed with chemotherapy, and ...
c55475f9-149f-4c4f-9eea-6ae7ce7ea320
trentmkelly/LessWrong-43k
LessWrong
What does knowing the heritability of a trait tell me in practice? The concept of heritability gets misunderstood a lot, so there are several articles discussing what it doesn't mean. But reading through all of them leaves me confused about what it does mean in practical terms, outside the technical definition. For e...
69851485-5cf6-4042-8105-a44108e11a65
trentmkelly/LessWrong-43k
LessWrong
Oliver Sipple The other day I read Wikipedia arguably too much, and consequently came to know the story of Oliver Sipple. Here’s my summary of the story according to these two Wikipedia pages and this page: In the September of 1975, Oliver (‘Billy’) Sipple was an ex-marine of thirty-three, injured in Vietnam and livi...
a4ac4011-5182-431c-905b-7d21ac92ffc2
trentmkelly/LessWrong-43k
LessWrong
Meetup : Canberra Secular Solstice Discussion article for the meetup : Canberra Secular Solstice WHEN: 18 December 2016 05:00:00PM (+1100) WHERE: Nara Peace Park, Canberra Canberra’s first secular solstice celebration! Of course picnicking is an excellent way to celebrate, bring a dish! Preferably vegetarian/vegan/...
ba130e2f-046d-44b6-ab23-b6a4afd1a2f8
trentmkelly/LessWrong-43k
LessWrong
Reasonable Requirements of any Moral Theory
ca1b5324-790c-4aa7-8034-3e282cd5195d
trentmkelly/LessWrong-43k
LessWrong
A conversation about cooking, science, and creativity My girlfriend is a creative cook, and worked in the Portland food industry - renowned for its innovation by small kitchens - for ten years. I'm interested in creativity in scientific research. Where does it come from? Furthermore, what is the relationship between ...
01ac113d-d299-47be-97f4-fa37b5e35305
trentmkelly/LessWrong-43k
LessWrong
The non-painless upload   I have been trying to absorb the Lesswrong near-consensus on cryonics/quantum mechanics/uploading, and I confess to being unpersuaded by it. I'm not hostile to cryonics; just indifferent, and having a bit of trouble articulating why the insights on identity that I have been picking up from t...
1894890f-d26c-454d-bb84-c9b13fd69c93
StampyAI/alignment-research-dataset/arxiv
Arxiv
CURIOUS: Intrinsically Motivated Modular Multi-Goal Reinforcement Learning 1 Introduction --------------- In *autonomous continual learning*, agents must evolve in an open-ended, changing world and might face a variety of potential tasks. In such a realistic environment, tasks cannot always be pre-specified by engi...
31ee7116-25d2-46b7-a140-a9f7165ec9ed
trentmkelly/LessWrong-43k
LessWrong
Gödel's Legacy: A game without end Cross-posted at Brick. A fun exploration, hopefully requiring only a medium level of technical familiarity.  "Not with a bang but a whimper" It's 1930 and Kurt Gödel has just delivered a talk at Königsberg that carpet bombed Hilbert's Program. ...and no one noticed.  > Godel's an...
0e0274c2-c3da-4da3-8781-53712f11ac4e
StampyAI/alignment-research-dataset/arxiv
Arxiv
AvE: Assistance via Empowerment 1 Introduction --------------- We aim to enable artificial agents, whether physical or virtual, to assist humans in a broad array of tasks. However, training an agent to provide assistance is challenging when the human’s goal is unknown because that makes it unclear what the agent sh...
298fb2d1-62a4-4129-be32-8a5958fc3c1e
trentmkelly/LessWrong-43k
LessWrong
[Link] Cognitive bias modification as a treatment for depression This seems relevant to LessWrong, both as an extreme example of how biases can hurt people and as a possible rationality technique. Depression is presumably at the outer end of some spectrum; to the extent that it's caused by cognitive mistakes, people i...
a01494cb-3cc6-4a46-b926-ad9d2af5f009
trentmkelly/LessWrong-43k
LessWrong
Misaligned AGI Death Match Here’s my very raw, very unpolished idea of how to use the AI-in-the-box scenario to possibly solve AI alignment. It’s propably inadequate in its current form, and I didn’t even try to meticulously apply the security mindset to every step (I’m not sure I have it anyway), but maybe someone mo...
dfabd69b-082a-4248-b118-0203f5dc36e7
trentmkelly/LessWrong-43k
LessWrong
Trade makes you responsible, but why? It’s supposedly bad to exploit poor people by paying them as little as you can get away with in trade. Onlookers who don’t offer the needy anything condemn those who offer some non altruistic benefit through trade because it isn’t enough. It’s interesting that we see this as a fa...
89131b72-52fc-436b-9662-4363a3164ed0
trentmkelly/LessWrong-43k
LessWrong
At what point should CFAR stop holding workshops due to COVID-19? What information about the virus' nature and spread would cause you to believe it's too risky to continue holding workshops?
e2f0ad20-15a0-49ac-a017-7eb52b7265ff
trentmkelly/LessWrong-43k
LessWrong
[Aspiration-based designs] Outlook: dealing with complexity Summary. This teaser post sketches our current ideas for dealing with more complex environments. It will ultimately be replaced by one or more longer posts describing these in more detail. Reach out if you would like to collaborate on these issues. Multi-di...
591b2f0e-e2ae-48e0-90ab-26fdfaafcf95
trentmkelly/LessWrong-43k
LessWrong
Contra: Changing Role Terms In contra dance there are two roles, which were traditionally called "Ladies" and "Gents". A lot of people are currently trying to move the community over to calling the roles "Robins" and "Larks", to make it clear that anyone can dance either role. BIDA, the dance I help organize, switched...
8bb81d92-b776-4c12-a5d0-9c8c197722a5
trentmkelly/LessWrong-43k
LessWrong
One weird trick to turn maximisers into minimisers A putative new idea for AI control; index here. A simple and easy design for a u-maximising agent that turns into a u-minimising one. Let X be some boolean random variable outside the agent's control, that will be determined at some future time t (based on a cosmic ...
53acd5dd-49ac-470c-9282-db83f2148e97
trentmkelly/LessWrong-43k
LessWrong
Mailing List for Digitized Belief Network Discussion Hi all, This is a follow-up to a previous post of mine - 'A digitized belief network?'. I have now created a discussion group for anyone who wants to discuss the problems involved in creating a digital representation of a human's beliefs. Anyone who is interested ...
94cacb60-183e-4a87-8896-eaeef60fa69c
trentmkelly/LessWrong-43k
LessWrong
NPR show All Things Considered on the Singularity and SIAI The NPR show All Things Considered did a short story on the Singularity, including interviews with Eliezer Yudkowsky and others involved with SIAI: http://www.npr.org/2011/01/11/132840775/The-Singularity-Humanitys-Last-Invention
297cd8dc-a0de-4287-b81b-acc401fa2a62
StampyAI/alignment-research-dataset/arxiv
Arxiv
On the Veracity of Local, Model-agnostic Explanations in Audio Classification: Targeted Investigations with Adversarial Examples 1 Introduction --------------- With the rise of deep learning methods used in Music Information Retrieval (MIR), also the desire for explaining the decisions made by such models has incre...
01f58038-3bc1-4347-9afe-962a10a773ec
trentmkelly/LessWrong-43k
LessWrong
[Link] RSA Animate: extremely entertaining LW-relevant cartoons It's a brilliant idea: a lecture by a cool modern thinker, illustrated by word-by-word doodles on a whiteboard. Excellent at pulling you along the train of thought and absolutely disallowing boredom. The lectures' content is pretty great too, although th...
954af5b5-3959-4e3d-b455-c40ffe49cf30
trentmkelly/LessWrong-43k
LessWrong
A Concrete Proposal for Adversarial IDA Note: This post came out of a conversation with Geoffrey Irving and Buck Shlegeris. Epistemic Status: I suspect Paul has already thought of most or all of the ideas presented here, though I nevertheless found the exercise of carefully specifying an IDA implementation helpful an...
1ec14d30-e3be-4928-aa3a-48427fc37873
trentmkelly/LessWrong-43k
LessWrong
Perfect Competition Previously in Immoral Mazes sequence: Moloch Hasn’t Won Perfect Competition In Meditations on Moloch, Scott points out that perfect competition destroys all value beyond the axis of competition. Which, for any compactly defined axis of competition we know about, destroys all value. This is math...
cb8a93b8-2734-4323-b0de-3db62741b98d
StampyAI/alignment-research-dataset/lesswrong
LessWrong
AI Safety without Alignment: How humans can WIN against AI Here are some ideas on how we can restrict Super-human AI (SAI) **without built-in Alignment**: Overall idea: ============= We need to use unbreakable first principles, especially laws of physics to restrict or destroy SAI. We can leverage the **asymmetries ...
e88169e9-703c-421a-8f48-61496b9d5b3d
trentmkelly/LessWrong-43k
LessWrong
Snyder-Beattie, Sandberg, Drexler & Bonsall (2020): The Timing of Evolutionary Transitions Suggests Intelligent Life Is Rare Copying over Anders Sandberg's Twitter summary of the paper: > There is life on Earth but this is not evidence for life being common in the universe! This is since observing life requires livin...
f446ba23-fd11-4d02-9825-e46c0563a73c
trentmkelly/LessWrong-43k
LessWrong
Open Philanthropy Technical AI Safety RFP - $40M Available Across 21 Research Areas Open Philanthropy is launching a big new Request for Proposals for technical AI safety research, with plans to fund roughly $40M in grants over the next 5 months, and available funding for substantially more depending on application qu...
0747df7c-9153-4478-8313-a62ba637148c
trentmkelly/LessWrong-43k
LessWrong
Measuring Optimization Power Previously in series:  Aiming at the Target Yesterday I spoke of how "When I think you're a powerful intelligence, and I think I know something about your preferences, then I'll predict that you'll steer reality into regions that are higher in your preference ordering." You can quantify ...
801f7a50-66a7-47f8-8e9c-f470c91f1275
trentmkelly/LessWrong-43k
LessWrong
Manhattan project for aligned AI One possible thing that I imagine might happen, conditional on an existential catastrophe not occurring, is a Manhattan project for aligned AGI. I don’t want to argue that this is particularly likely or desirable. The point of this post is to sketch the scenario, and briefly discuss so...
1f84b13f-64e7-4c7b-abc1-3a7f06fa745c
trentmkelly/LessWrong-43k
LessWrong
Quantum Decisions CFAR sometimes plays a Monday / Tuesday game (invented by palladias). Copying from the URL: > On Monday, your proposition is true. On Tuesday, your proposition is false. Tell me a story about each of the days so I can see how they are different. Don't just list the differences (because you're alread...
fdbb84aa-f5f4-4544-913a-b020a9c73251
trentmkelly/LessWrong-43k
LessWrong
All AGI Safety questions welcome (especially basic ones) [April 2023] tl;dr: Ask questions about AGI Safety as comments on this post, including ones you might otherwise worry seem dumb! Asking beginner-level questions can be intimidating, but everyone starts out not knowing anything. If we want more people in the wor...
14a19ec4-4892-496b-ae23-c95fa05c5502
trentmkelly/LessWrong-43k
LessWrong
Readiness Heuristics Followup to: How Much Thought A trolley is hurtling towards three people. It will kill them, unless you pull a lever that diverts it onto a different track. However, if you do this, it will hit a small child and kill her. Do you pull the lever, and kill a child, or do nothing, and let three adult...
066dbf38-ab38-43a8-9d3d-c456fbcbea80
StampyAI/alignment-research-dataset/eaforum
Effective Altruism Forum
The Vitalik Buterin Fellowship in AI Existential Safety is open for applications! This is a linkpost for[https://grants.futureoflife.org/](https://forum.effectivealtruism.org/out?url=https%3A%2F%2Fgrants.futureoflife.org%2F) *Epistemic status: Describing the fellowship that we are a part of and sharing some suggestio...
9c0360fd-6eef-4d21-9ab5-64bf5edb9a77
StampyAI/alignment-research-dataset/eaforum
Effective Altruism Forum
Nuclear brinksmanship is not a good AI x-risk strategy In a [recent article](https://web.archive.org/web/20230330001659/https://time.com/6266923/ai-eliezer-yudkowsky-open-letter-not-enough/), Eliezer Yudkowsky advocates for the following measures to stop AI development: > Shut down all the large GPU clusters (the la...
5e0e1150-099b-41ac-88b3-c7b6bf116878
trentmkelly/LessWrong-43k
LessWrong
AI #71: Farewell to Chevron Chevron deference is no more. How will this impact AI regulation? The obvious answer is it is now much harder for us to ‘muddle through via existing laws and regulations until we learn more,’ because the court narrowed our affordances to do that. And similarly, if and when Congress does pa...
ad51de4a-69ce-4931-aa05-f5d3ac82ccf4
trentmkelly/LessWrong-43k
LessWrong
If a little is good, is more better? I’ve recently seen a bunch of discussions of the wisdom of publicly releasing the weights1 of advanced AI models. A common argument form that pops up in these discussions is this: 1. The problem with releasing weights is that it means that thing X can happen on a large scale, whi...
45f14ca8-3620-4a90-a3d8-9f84ab34408e
trentmkelly/LessWrong-43k
LessWrong
Are meaningful careers a cover story? It is more respectable to have a non-altruistic hobby than to have a non-altruistic job. If I say to a public servant that their career doesn’t make the world a better place, I’m being offensive. If I say to a chess player that their hobby doesn’t make the world a better place, th...
0e568fc1-c388-4703-a6d0-bf9d5305fe96
StampyAI/alignment-research-dataset/alignmentforum
Alignment Forum
Problem relaxation as a tactic .mjx-chtml {display: inline-block; line-height: 0; text-indent: 0; text-align: left; text-transform: none; font-style: normal; font-weight: normal; font-size: 100%; font-size-adjust: none; letter-spacing: normal; word-wrap: normal; word-spacing: normal; white-space: nowrap; float: none; ...
fa1d711e-1c07-4660-9bbf-c9a03ec29cf2
trentmkelly/LessWrong-43k
LessWrong
The Busy Bee Brain An old post from New Savanna. > With artificial neural nets (ANNs) all over the place these days, this post from March, 2013, warrants a bump to the top of the queue. Perhaps the chief difference between real brains and ANNs is the fact that each neuron in a real brain is a living agent, albeit of...
162adfdd-0ed1-4e7e-a9ad-9ac282552870
trentmkelly/LessWrong-43k
LessWrong
The value of x-risk reduction In his interview on the 80000 hours podcast after the publication of the Precipice, Toby Ord claimed that x-risk reduction was worth pursing regardless of weather x-risk was high or low. The claim is that when x-risk is low, it's worth working on x-risk because a small decline in the pro...
876f53d0-8d17-47cd-8648-24d16cebfb0c
trentmkelly/LessWrong-43k
LessWrong
Stabilizing logical counterfactuals by pseudorandomization Previously, we discussed the construction of logical counterfactuals in the language of optimal predictors. These counterfactuals were found to be well-behaved when a certain non-degeneracy condition is met which can be understood as a bound on the agent's abi...
e40fe1d2-e5da-4e37-a6bf-fd45bf233db7
StampyAI/alignment-research-dataset/alignmentforum
Alignment Forum
Alignment Newsletter #46 Find all Alignment Newsletter resources [here](http://rohinshah.com/alignment-newsletter/). In particular, you can [sign up](http://eepurl.com/dqMSZj), or look through this [spreadsheet](https://docs.google.com/spreadsheets/d/1PwWbWZ6FPqAgZWOoOcXM8N_tUCuxpEyMbN1NYYC02aM/edit?usp=sharing) of al...
8037668a-0182-48f5-91af-82f1e325d151
trentmkelly/LessWrong-43k
LessWrong
[Link] Belief in religion considered harmful? I've recently run across this 2007 post on the blog Unqualified Reservations (archive best read here). It is written by Mencious Moldbug, who is probably familiar to some Overcoming Bias and Lesswrong readers. He is a erudite, controversial and most of all contrarian socia...