id
stringlengths
36
36
source
stringclasses
15 values
formatted_source
stringclasses
13 values
text
stringlengths
2
7.55M
408d0d3c-a548-4699-991c-cc68e6d8ac33
trentmkelly/LessWrong-43k
LessWrong
Meetup : Less Wrong NH Meet-up Discussion article for the meetup : Less Wrong NH Meet-up WHEN: 11 August 2015 07:00:00PM (+0100) WHERE: 269 Pearl St., Manchester, NH (Posting this for a friend because he can't comment) The second NH meet-up is this Tuesday, 8/11, in Manchester, NH at 7 pm at a private residence. L...
980c1966-6da4-4d07-834a-72cfd204d0f4
trentmkelly/LessWrong-43k
LessWrong
2024 in AI predictions Follow-up to: 2023 in AI predictions. Here I collect some AI predictions made in 2024. It's not very systematic, it's a convenience sample mostly from browsing Twitter/X. I prefer including predictions that are more specific/testable. I'm planning to make these posts yearly, checking in on pred...
8edec6aa-1269-4b2d-904a-10659d851ba1
StampyAI/alignment-research-dataset/lesswrong
LessWrong
Quick Nate/Eliezer comments on discontinuity This isn't a proper response to [Paul Christiano](https://www.lesserwrong.com/posts/AfGmsjGPXN97kNp57/arguments-about-fast-takeoff) or [Katja Grace](https://aiimpacts.org/likelihood-of-discontinuous-progress-around-the-development-of-agi/)'s recent writings about takeoff sp...
4024cb0d-27a1-4da3-8df8-f84ce64a1381
trentmkelly/LessWrong-43k
LessWrong
What would the world be like if everyone was a rationalist? None
55d4296e-be86-4289-9a47-c8585556460c
trentmkelly/LessWrong-43k
LessWrong
How Can People Evaluate Complex Questions Consistently? I'm doing a project on how humans can evaluate messy problems and come up with consistent answers (consistent with both themselves over time and with other people), and what the trade off is with accuracy. This isn't a single unified field, so I need to go poking...
55d3e9ff-c3df-4f11-9a4b-76a0f80a6363
trentmkelly/LessWrong-43k
LessWrong
You'd Want The Right Answer It's quite common to see people bending the truth if it works for them. It happens far more often than you'd like to think. For example, seeing the weak point of a belief but you choose to ignore it and rehearse the strong points. What's the harm in that? What if the setting changed - You'...
0fea5acd-93be-4d40-a58e-2da9715213b8
trentmkelly/LessWrong-43k
LessWrong
HPMOR Anniversary Parties: Coordination, Resources, and Discussion Summary As some of you may already know, it's been ten years since Harry Potter and the Method of Rationality concluded and wrap parties were held around the world. I've taken the baton of coordinating parties from Habryka, and now over 20 parties on 4...
6fb5de44-3ed2-41a0-99a0-27cfa9bc9ed4
StampyAI/alignment-research-dataset/arxiv
Arxiv
Measurement in AI Policy: Opportunities and Challenges ### 1 Introduction The rapid development of Artificial Intelligence and its growing effects on science, the economy and society have increased the need to better assess its technical progress, and economic and societal impact. On October 30, 2019, the [AI ...
67b3bc02-140a-49c5-b2f2-a26989cefbcd
StampyAI/alignment-research-dataset/special_docs
Other
AI Insights Dataset Analysis AI Insights Dataset Analysis Colleen McKenzie and J. Bryce Hidysmith 1 AI Insights Dataset Analysis 1.1 Overview This dataset collects a list of discoveries essential to the development of the current generation of artificial intelligence. We trace this development back as early as we have ...
770c655d-9399-4107-813c-2a8e5ce6eeac
StampyAI/alignment-research-dataset/arbital
Arbital
Proof of Bayes' rule: Probability form Let $\mathbf H$ be a [variable](https://arbital.com/p/random_variable) in $\mathbb P$ for the true hypothesis, and let $H_k$ be the possible values of $\mathbf H,$ such that $H_k$ is [https://arbital.com/p/-1rd](https://arbital.com/p/-1rd). Then, Bayes' theorem states: $$\mathbb...
65c24741-223a-4569-bf8d-c86b1192f8e9
StampyAI/alignment-research-dataset/lesswrong
LessWrong
We don’t trade with ants When discussing advanced AI, sometimes the following exchanges happens: “Perhaps advanced AI won’t kill us. Perhaps it will trade with us” “We don’t trade with ants” I think it’s interesting to get clear on exactly why we don’t trade with ants, and whether it is relevant to the AI situat...
778f8c2e-31ee-45b4-bfce-62a32aac5631
trentmkelly/LessWrong-43k
LessWrong
Examples of How I Use LLMs I often talk to people who don't find LLMs useful. Paraphrasing: > Why would I want to ask questions to a thing that might give me the right answer but might also make something up that fools me into thinking it's right? I've played with them, but they don't seem like they'd save me any tim...
439f2549-0ba4-4060-8182-c1d1df64c69a
StampyAI/alignment-research-dataset/lesswrong
LessWrong
Two Stupid AI Alignment Ideas Epistemic status: this is not my field.  I am unfamiliar with any research in it beyond what I've seen on LW.  These are the best ideas I could come up with when stopping to think about the problem, and aren't things I've seen on LW, but I don't imagine that they're original and I would h...
187ac1f1-cf25-4770-bfa9-4d21d5fd6ae2
trentmkelly/LessWrong-43k
LessWrong
Three Axes of Prohibitions The Game of Thrones board game is similar to Diplomacy (so I hear: I've never actually played Diplomacy). You often need to make alliances to survive, but these alliances are weak. It is both expected and required that you will eventually break your alliances, otherwise you will lose. My fir...
1269130b-84a0-4765-aa18-942c3f904d1b
trentmkelly/LessWrong-43k
LessWrong
Is Metaethics Unnecessary Given Intent-Aligned AI? While you can read this post as a standalone, I'm writing it partially in response to Wei Dai's recent post Meta Questions about Metaphilosophy. I had already been considering writing a top-level post about metaethics and whether "solving" it is a tractable problem th...
6d6bb427-4ba1-4319-a0bd-2a24eb44f5aa
trentmkelly/LessWrong-43k
LessWrong
Decision Theory Paradox: PD with Three Implies Chaos? Prerequisites: Familiarity with decision theories (in particular, Eliezer's Timeless Decision Theory) and of course the Prisoner's Dilemma. Summary: I show an apparent paradox in a three-agent variant of the Prisoner's Dilemma: despite full knowledge of each other...
54f006f6-c826-4514-a85f-9d6cf64845f2
trentmkelly/LessWrong-43k
LessWrong
How I'd Introduce LessWrong to an Outsider Note/edit: I'm imagining explaining this to a friend or family member who is at least somewhat charitable and trusting of my judgement. I am not imagining simply putting this on the About page. I should have made this clear from the beginning - my bad. However, I do believe t...
e8f09a23-5e0f-4f15-b5d3-52e9b73c1335
trentmkelly/LessWrong-43k
LessWrong
Regulatory lags for New Technology [2013 notes] I found some old notes from June 2013 on time delays in how fast one can expect Western political systems & legislators to respond to new technical developments. In general, response is slow and on the order of political cycles; one implication I take away is that a tak...
318fd52b-725a-42d4-bdea-c285f3893905
trentmkelly/LessWrong-43k
LessWrong
The News is Never Neglected Dear Lsusr, I am inspired by your stories about Effective Evil. My teachers at school tell me it is my civic responsibility to watch the news. Should I reverse this advice? Or should I watch the news like everyone else, except use what I learn for evil? Sincerely, [redacted] -----------...
74d48537-7267-4ab4-93ab-bb0d708ae3c9
trentmkelly/LessWrong-43k
LessWrong
The Monty Maul Problem In his Coding Horror Blog, Jeff Atwood writes about the Monty Hall Problem and some variants. The classic problem presents the situation in which the game show host allows a contestant to choose one of three doors, one of which opens to reveal a prize while the other two reveals goats. The host ...
1a4fb003-a649-474a-88bf-977b64a78173
StampyAI/alignment-research-dataset/arxiv
Arxiv
Reinforcement Learning with Competitive Ensembles of Information-Constrained Primitives 1 Introduction --------------- Learning policies that generalize to new environments or tasks is a fundamental challenge in reinforcement learning. While deep reinforcement learning has enabled training powerful policies, which ...
dc7f0d08-2cdf-4a04-93f0-b2c9bb4740b2
trentmkelly/LessWrong-43k
LessWrong
Transformative AI and Scenario Planning for AI X-risk This post is part of a series by the AI Clarity team at Convergence Analysis. In our previous post, Corin Katzke reviewed methods for applying scenario planning methods to AI existential risk strategy. In this post, we want to provide the motivation for our focus o...
eaf80ff0-497c-443b-ab8d-19f73000835f
trentmkelly/LessWrong-43k
LessWrong
Mortal: A Transponyist Fanfiction I recently published Mortal, a novella-length My Little Pony fanfiction meant to introduce anti-death concepts to an unfamiliar audience. Short description: > Twilight Sparkle's friends have lived long and happy lives. Now their time is coming to an end, but Rainbow Dash, at least, w...
80dfd562-cfc1-4a30-adc9-a72a963ded25
trentmkelly/LessWrong-43k
LessWrong
Towards reflection with relative optimal predictor schemes We construct a family or oracles relative to which all estimation problems which have approximate solutions in exponential time are generatable. Relatively to these oracles, we can use the optimal predictor scheme Λ for assigning expectation values to arbitrar...
ff5c2706-47af-4461-8ee1-943284bce512
trentmkelly/LessWrong-43k
LessWrong
Dependability I have become very, very interested in developing a skill that I call Dependability. I believe the skill exists on a spectrum, and you can have less or more of it. It’s not a binary where you either have it or you don’t. I believe this skill can be trained on purpose. I will briefly describe each attr...
7c8099ef-7d7b-4969-800e-2fcf3999fea7
StampyAI/alignment-research-dataset/blogs
Blogs
Energy efficiency of Vickers Vimy plane *Updated Nov 5, 2020* The Vickers Vimy: * flew around 0.025—0.050 m/kJ * and moved mass at around 0.081 – 0.25 kg.m/J Details ------- The *Vickers Vimy* was a 1917 British WWI bomber.[1](https://aiimpacts.org/energy-efficiency-of-vickers-vimy-plane/#easy-footnote-bottom-...
d1005ec2-c79b-46c4-b5bb-9f238c66816d
trentmkelly/LessWrong-43k
LessWrong
rough draft on what happens in the brain when you have an insight Epistemic status: It is better to be wrong than to have no model at all. I recommend the footnotes.[1]     🍵 On my current models of theoretical[2] insight-making, it looks something like this: 1. A gradual build-up and propagation of salience wrt ...
9154674c-dc88-43ea-98a8-4fcb950c28c6
trentmkelly/LessWrong-43k
LessWrong
Goals for which Less Wrong does (and doesn't) help Related to: Self-Improvement or Shiny Distraction: Why Less Wrong is anti-Instrumental Rationality We’ve had a lot of good criticism of Less Wrong lately (including Patri’s post above, which contains a number of useful points). But to prevent those posts from confusi...
57c0939f-dbed-4e14-aaf4-5184a0e81b47
trentmkelly/LessWrong-43k
LessWrong
AGI is likely to be cautious According to Professor Stuart Russell, and with a sentiment I have seen re-expressed often in the AI safety community: > A system that is optimizing a function of n variables, where the objective depends on a subset of size k<n, will often set the remaining unconstrained variables to extr...
c9efcbc7-222d-4b4c-ac06-fec0b34bf207
trentmkelly/LessWrong-43k
LessWrong
Does GenZ have a shorter attention span? intuitive answer: yes! * reasoning: harder to focus on one thing at a time. * TLDR; frequent context switching to ensure the constant or increasing flow of dopamine. * it doesn't matter if I am consuming low or high-quality dopamine (low: watching reaction videos, high: ...
307708ba-e802-4423-9ed8-086c634f7977
trentmkelly/LessWrong-43k
LessWrong
Decoherent Essences Followup to:  Decoherence is Pointless In "Decoherence is Pointless", we talked about quantum states such as > (Human-BLANK) * ((Sensor-LEFT * Atom-LEFT) + (Sensor-RIGHT * Atom-RIGHT)) which describes the evolution of a quantum system just after a sensor has measured an atom, and right before a ...
7c54905b-520d-41b4-ac3d-33e22374d3fc
trentmkelly/LessWrong-43k
LessWrong
[Link] Sarah Constantin: "Why I am Not An AI Doomer" This is a good post from Sarah Constantin explaining why her expected timeline to agentic AGI is long (> 10 years). Topics discussed include whether LLMs and other current major research directions will endow AIs with adequate world models, causal inference, and goa...
a89cf6ee-eff8-4e1d-9080-573849a9a628
trentmkelly/LessWrong-43k
LessWrong
AISN#14: OpenAI’s ‘Superalignment’ team, Musk’s xAI launches, and developments in military AI use 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. --------------------------------...
2ad70528-8a41-46ba-aad9-47964fd02cd7
trentmkelly/LessWrong-43k
LessWrong
Announcing the SoS Research Collective for independent researchers (and academics thinking independently) “Study hard what interests you the most in the most undisciplined, irreverent and original manner possible.” (Richard Feynman) We (the founding gardeners Roger, Dario, and Sergey) have always conceived of Seeds o...
08fe5805-d22c-4af2-a0a3-ff8f1bd417e9
trentmkelly/LessWrong-43k
LessWrong
Open Thread, Jun. 1 - Jun. 7, 2015 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 b...
49554a3b-eefc-43ca-bd8c-fa196365707a
StampyAI/alignment-research-dataset/eaforum
Effective Altruism Forum
When reporting AI timelines, be clear who you're deferring to It's fashionable these days to ask people about their AI timelines. And it's fashionable to have things to say in response. But relative to the number of people who report their timelines, I suspect that only a small fraction have put in the effort to form...
969a154c-5f9a-4d6a-9216-e6c055be44e1
trentmkelly/LessWrong-43k
LessWrong
How will they feed us I was wondering how in more optimistic future scenarios people would be fed or get money to buy food et cetera. A common topic in discussions between pro and anti doomers is how the AI will kill us. It is a fair point, killing 8 billion people seems difficult at first glance. But what if the AI d...
63df4330-1b50-47e3-9194-5e7b3975a2d7
trentmkelly/LessWrong-43k
LessWrong
Other people's procedural knowledge gaps There's been a recent heavily upvoted and profusely commented post about things people want to learn. It's close to having so many comments in a single day that it should probably have a part 2. However, the subject seems to inspire thoughts about what *other* people ought to ...
0f9750cc-c401-4e63-8cd1-b8336d1fab41
trentmkelly/LessWrong-43k
LessWrong
Apply to LASR Labs: a London-based technical AI safety research programme Edit: Applications for this round are now closed! If you are interested in future rounds, you can express interest here.  TLDR; apply by April 24th 23:59 GMT+1 to join a 12-week programme and write a technical AI safety paper in a team of 4 wit...
8c028b09-7a03-446a-beff-80814998335b
trentmkelly/LessWrong-43k
LessWrong
Quantum Physics Revealed As Non-Mysterious This is one of several shortened indices into the Quantum Physics Sequence. Hello!  You may have been directed to this page because you said something along the lines of "Quantum physics shows that reality doesn't exist apart from our observation of it," or "Science has disp...
bbd7315f-5032-4d20-bccc-ab8a9786f2bb
StampyAI/alignment-research-dataset/alignmentforum
Alignment Forum
A positive case for how we might succeed at prosaic AI alignment *This post is my attempt at something like a response to [Eliezer Yudkowsky’s recent discussion on AGI interventions](https://www.lesswrong.com/posts/CpvyhFy9WvCNsifkY/discussion-with-eliezer-yudkowsky-on-agi-interventions).* I tend to be relatively pe...
94fc5548-b36b-4dc6-a7cd-3860a53524e9
trentmkelly/LessWrong-43k
LessWrong
What criterion would you use to select companies likely to cause AI doom? (Asking for a project which involves collecting data from the major AI companies.) Potential ideas found during a brainstorm: - Threshold on number of parameters of biggest model produced. - Threshold on number of money invested in AI research. ...
70a24e38-5205-4af2-b4d9-c3a51b38af1a
trentmkelly/LessWrong-43k
LessWrong
Anthropic Atheism (Crossposted from my blog) I've been developing an approach to anthropic questions that I find less confusing than others, which I call Anthropic Atheism (AA). The name is a snarky reference to the ontologically basic status of observers (souls) in other anthropic theories. I'll have to explain myse...
d65b6c53-8e54-4a8e-8e8e-99ff592705fd
trentmkelly/LessWrong-43k
LessWrong
Open Thread None
79c87699-504f-4293-94da-abb7f8b774c1
trentmkelly/LessWrong-43k
LessWrong
Defining AI wireheading What does it mean for an AI to wirehead its reward function? We're pretty clear on what it means for a human to wirehead - artificial stimulation of part of the brain rather than genuine experiences - but what does it mean for an AI? We have a lot of examples of wireheading, especially in info...
9aa42c66-59a1-440e-999c-a2592bc6cc26
trentmkelly/LessWrong-43k
LessWrong
New report: A review of the empirical evidence for existential risk from AI via misaligned power-seeking Visiting researcher Rose Hadshar recently published a review of some evidence for existential risk from AI, focused on empirical evidence for misalignment and power seeking. (Previously from this project: a blogpos...
708e655e-e1b9-4762-be73-bc4539d6e845
trentmkelly/LessWrong-43k
LessWrong
[LINK] Brain region changes shape with learning the layout of London General Intro: http://www.bbc.co.uk/news/health-16086233 Older research: http://www.fil.ion.ucl.ac.uk/Maguire/Maguire2006.pdf Abstract of latest research Possible implications for WBE (Is it possible to get short term function correct without ha...
c11d52e6-88b7-4f18-8012-7b0d83de6e7d
trentmkelly/LessWrong-43k
LessWrong
SuperBabies podcast with Gene Smith This is a linkpost to the latest episode of The Bayesian Conspiracy podcast. This one is a 1.5 hour chat with Gene Smith about polygenic screening, gene-editing for IVF babies, and even some gene-editing options for adults. Likely of interest to many Less Wrongians.
c84475dc-7aa5-4bff-b946-33b2e76a639d
trentmkelly/LessWrong-43k
LessWrong
Meetup : Cambridge UK Discussion article for the meetup : Cambridge UK WHEN: 19 February 2012 11:00:00AM (+0000) WHERE: JCR, Trinity College, Cambridge, CB2 1TQ, UK Meet at the Great Gate if you dont know where the JCR is. Great Gate is on St. John's Street opposite the bookshop "Heffers". Join the google group at ...
86bbb75b-b0b0-46aa-84da-1b34cf357a6b
StampyAI/alignment-research-dataset/arxiv
Arxiv
Adversarial Logit Pairing 1 Introduction --------------- Many deep learning models today are vulnerable to adversarial examples, or inputs that have been intentionally optimized to cause misclassification. In the context of computer vision, object recognition classifiers incorrectly recognize images that have been ...
21158a36-27ae-4c59-b5b8-64704183277a
awestover/filtering-for-misalignment
Redwood Research: Alek's Filtering Results
id: post1857 This post will try to explain what I mean by continuity assumptions (and discontinuity assumptions), and why differences in these are upstream of many disagreements about AI safety. None of this is particularly new, but seemed worth reiterating in the form of a short post. What I mean by continuity Start ...
2b119835-8800-4fb9-a40d-fa3292a3b263
trentmkelly/LessWrong-43k
LessWrong
Meetup : Games in Kocherga club: FallacyMania, Tower of Chaos, Scientific Discovery Discussion article for the meetup : Games in Kocherga club: FallacyMania, Tower of Chaos, Scientific Discovery WHEN: 22 February 2017 07:40:00PM (+0300) WHERE: Moscow, B.Dorogomilovskaya, 5-2 Welcome to Moscow LW community makeshift...
5a6cafa0-75df-4615-ba53-445115e0e014
trentmkelly/LessWrong-43k
LessWrong
D&D.Sci June 2022 Evaluation and Ruleset This is a followup to the D&D.Sci post I made ten days ago; if you haven’t already read it, you should do so now before spoiling yourself. Here is the web interactive I built to let you evaluate your solution; below is an explanation of the rules used to generate the dataset (...
6a1b0308-7172-4905-b0a9-d401143c24da
trentmkelly/LessWrong-43k
LessWrong
Databases of human behaviour and preferences? Hey there, lesserwrongers! Starting on some of the computer science/neuroscience of my ideas for deducing human preference. To do this, it would be useful to have datasets of human behaviour in relatively restricted situations. Possibly datasets of people playing simple ...
647ff8c7-6c91-4e65-8b6b-7754b9398e2e
trentmkelly/LessWrong-43k
LessWrong
[LINK] What are some stupid things smart people do? On Quora: What are some stupid things smart people do? Examples of common types of stupidity that are typical of otherwise very smart people. Lee Semel's answer in particular would make a great post here: a "to don't" list. You may wish to go through and identify th...
059868a5-de7b-4015-9cd5-93f692d3669f
trentmkelly/LessWrong-43k
LessWrong
Welcome to Rationality [Edit With Your Details] (The following are our suggestions for what kind of information is best to include in the welcome post of your group, feel free to replace them with whatever you think is best) What kind of events does your group usually run? What does it usually do? How frequently doe...
2e9cbdf6-9b83-43e4-83db-06270058653d
trentmkelly/LessWrong-43k
LessWrong
A Data limited future This is a prediction of what I think the near future might be like. Suppose the trends in scaling laws roughly continue. Deep learning can do anything if it has enough data. But getting that data is hard. So large language models get better, but not that much better. They are using most of the ...
e55be8cf-dde7-4ccc-b690-15ef8db89cf6
trentmkelly/LessWrong-43k
LessWrong
Summary of my Participation in the Good Judgment Project Follow-Up to Good Judgment Project, Season Three. During the last forecasting season I took part in the Good Judgment Project (GJP; see also the blog) and this is a short summary of my participation (actually triggered by hamnox comment). The GJP estimates wor...
a3cb70bd-21b2-4d53-9b56-e4a15c1012dd
trentmkelly/LessWrong-43k
LessWrong
"Outside View!" as Conversation-Halter Followup to:  The Outside View's Domain, Conversation Halters Reply to:  Reference class of the unclassreferenceable In "conversation halters", I pointed out a number of arguments which are particularly pernicious, not just because of their inherent flaws, but because they attem...
b17c2529-d510-466b-b24e-ccb899d0456b
trentmkelly/LessWrong-43k
LessWrong
The ones who endure Content warning: death, long-lasting suffering. This story was rough to write, and may well be rough to read. It's intended as a response to Le Guin's The Ones Who Walk Away From Omelas, although it can be read standalone. > If you can force your heart and nerve and sinew > To serve your turn long...
e6b18410-1a0f-44e3-975a-2f75235a3852
trentmkelly/LessWrong-43k
LessWrong
Splitting Debate up into Two Subsystems In this post I will first recap how debate can help with value learning and that a standard debater optimizes for convincingness. Then I will illustrate how two subsystems could help with value learning in a similar way, without optimizing for convincingness. (Of course this ne...
08fe9c97-bc33-45c3-b7a1-f6af2bcbb53f
trentmkelly/LessWrong-43k
LessWrong
Positive Directions Let's set the scene - Nihilism We are chunks of self important jelly staggering about on the surface of a tiny nugget of rock in a second rate solar system. Sol's system is ~ four billion years old We have been “Ourselves” (Anatomically modern) for ~200,000 years Countless people have lived and ...
46d22e00-4162-4f78-864b-9cf655cde3dd
trentmkelly/LessWrong-43k
LessWrong
D&D.Sci(-fi): Colonizing the SuperHyperSphere [Evaluation and Ruleset] This is a followup to the D&D.Sci post I made ten days ago; if you haven’t already read it, you should do so now before spoiling yourself. Here is the web interactive I built to let you evaluate your solution; below is an explanation of the rules ...
b62e676f-6cae-426f-9ec7-d63914febe69
StampyAI/alignment-research-dataset/lesswrong
LessWrong
Taboo "human-level intelligence" Maybe calls for tabooing terms can become a bit inflationary. However, talking about if and when AI will reach or surpass “human-level intelligence” can misleadingly frame the discussion, so I recommend avoiding that term.  What is "human-level intelligence" and when will AI surpa...
fb3a619b-4f19-456a-8a35-3a9269f30ecc
trentmkelly/LessWrong-43k
LessWrong
What do we know about the AI knowledge and views, especially about existential risk, of the new OpenAI board members? They have announced three new board members in addition to Altman, but we seem to know almost nothing about their views or knowledge on any AI-related subjects? What if anything do we know? From OpenA...
6626699e-32cf-406d-acc0-2ab7436528d5
trentmkelly/LessWrong-43k
LessWrong
Bringing Agency Into AGI Extinction Is Superfluous Epistemic status: Maybe I am missing something here, normally I'd publish this on my own blog, but I'd actually love to get some feedback before I signal boost it. I’ve seen a lot of debates around AGI risk hedge on the question of whether or not AGI will be “agentic...
db0a5103-ee1e-4128-989b-a575747375da
awestover/filtering-for-misalignment
Redwood Research: Alek's Filtering Results
id: post206 Work that I’ve done on techniques for mitigating risk from misaligned AI often makes a number of conservative assumptions about the capabilities of the AIs we’re trying to control. (E.g. the original AI control paper , Adaptive Deployment of Untrusted LLMs Reduces Distributed Threats , How to prevent collu...
da034554-bc27-49fc-827b-d23990a31e53
trentmkelly/LessWrong-43k
LessWrong
Differential reproduction for men and women. There's an idea I've seen a number of times that 80% of women have had descendants, but only 40% of men. A little research tracked it back to this, but the speech doesn't have a cite and I haven't found a source. The reproduction rates for men and women (possibly for the w...
685fa971-c4e2-417a-b810-58361ec259e0
trentmkelly/LessWrong-43k
LessWrong
What I Learned (Conclusion To "The Sense Of Physical Necessity") This is the concluding post in a sequence that demonstrates a complete naturalist study, specifically a study of query hugging (sort of), as described in The Nuts and Bolts of Naturalism. For context on this sequence, see the intro post. Here is where m...
9442c1c2-a1ab-46ee-b7c8-5b69c782342b
trentmkelly/LessWrong-43k
LessWrong
Oracle paper Available on the arXiv, my paper on two types of Oracles (AIs constrained to answering questions only), and how to use them more safely. > An Oracle is a design for potentially high power artificial intelligences (AIs), where the AI is made safe by restricting it to only answer questions. Unfortunately m...
a8bbcb9f-4396-4268-8224-921ae369ea54
trentmkelly/LessWrong-43k
LessWrong
Find a study partner For reasons mentioned in So8res article as well as for other reasons: studying with a partner can be very good. In November, Adele_L had posted an article for people wanting to find a study partner. It got 17 comments, but only 1 since November 16th. So I thought we (I) should make a monthly threa...
f87b22a6-3bf1-4700-9990-27e7f56579f7
trentmkelly/LessWrong-43k
LessWrong
To Be Decided #4 TBD is a quarterly-ish newsletter about deploying knowledge for impact, learning at scale, and making more thoughtful choices for ourselves and our organizations. On Stories vs. Data The reason stories work for us as human beings is because they are few in number. We can spend two hours watching a d...
29d89cb8-2bc9-4e33-9ec8-2e1672c6f0f5
trentmkelly/LessWrong-43k
LessWrong
Announcing ILIAD — Theoretical AI Alignment Conference We are pleased to announce ILIAD — a 5-day conference bringing together 100+ researchers to build strong scientific foundations for AI alignment. ***Apply to attend by June 30!*** * When: Aug 28 - Sep 3, 2024 * Where: @Lighthaven (Berkeley, US) * What: A mix...
d0856cda-2344-4dd3-ad99-1e7abaff9e56
StampyAI/alignment-research-dataset/lesswrong
LessWrong
Bayeswatch 7: Wildfire "You don't seem surprised," said Vi. "The very existence of Bayeswatch increases the odds of catastrophic disaster. The world is like a forest. Put out the little fires and detritus accumulates until every spark causes a conflagration," said Miriam. Around the world, 99% of Internet-connecte...
7edcd244-84cf-4444-a453-4154fa667b4c
trentmkelly/LessWrong-43k
LessWrong
Locality of goals Introduction Studying goal-directedness produces two kinds of questions: questions about goals, and questions about being directed towards a goal. Most of my previous posts focused on the second kind; this one shifts to the first kind. Assume some goal-directed system with a known goal. The nature o...
f67a14bb-6644-4b85-9877-ccec10da78ef
StampyAI/alignment-research-dataset/lesswrong
LessWrong
Global online debate on the governance of AI Hi guys, For background, I’m a French EA, attended a CFAR workshop, and recently decided to work on AI policy as it is a pressing and neglected issue. I’ve been working for The Future Society for a few weeks already and would like to share with you this opportunity to imp...
8b1962d6-222d-419b-b864-74e16b270969
trentmkelly/LessWrong-43k
LessWrong
The Fatal Gift of Beauty: The Trials of Amanda Knox Charles Petzold writes about Nina Burleigh's "The Fatal Gift of Beauty: The Trials of Amanda Knox".
166f7c07-f8da-49ea-a392-c799619eb7e6
StampyAI/alignment-research-dataset/lesswrong
LessWrong
[April Fools] User GPT2 is Banned For the past day or so, user [GPT2](https://www.lesswrong.com/users/gpt2) has been our most prolific commenter, replying to (almost) every LessWrong comment without any outside assistance. Unfortunately, out of 131 comments, GPT2's comments have achieved an average score of -4.4, and ...
24ad34fa-98b8-4d5a-a49a-166b5a7453a0
trentmkelly/LessWrong-43k
LessWrong
Open Thread By request of the community, an Open Thread for free-form comments, so long as they're still related to the basic project of this blog. A word on post requests:  You're free to ask, but the authors can't commit to posting on requested topics - it's hard enough to do the ones we have in mind already.
c0c11d94-d9d9-4953-84b9-6a659d961f7b
trentmkelly/LessWrong-43k
LessWrong
Harry Potter and the Methods of Rationality discussion thread, February 2015, chapter 111 This is a new thread to discuss Eliezer Yudkowsky’s Harry Potter and the Methods of Rationality and anything related to it. This thread is intended for discussing chapter 111. There is a site dedicated to the story at hpmor.com,...
192ff849-04fe-4c49-8da8-ee705e865d13
trentmkelly/LessWrong-43k
LessWrong
[SEQ RERUN] The Failures of Eld Science Today's post, The Failures of Eld Science was originally published on 12 May 2008. A summary (taken from the LW wiki):   > A short story set in the same world as "Initiation Ceremony". Future physics students look back on the cautionary tale of quantum physics. Discuss the p...
07e0f565-638e-4843-9ef9-216f4af86c6e
StampyAI/alignment-research-dataset/alignmentforum
Alignment Forum
New Scaling Laws for Large Language Models On March 29th, DeepMind published a paper, ["Training Compute-Optimal Large Language Models"](https://arxiv.org/abs/2203.15556), that shows that essentially everyone -- OpenAI, DeepMind, Microsoft, etc. -- has been training large language models with a deeply suboptimal use o...
5c30b717-f6f5-4012-8596-892cdcf8959d
StampyAI/alignment-research-dataset/special_docs
Other
Reconciliation between factions focused on near-term and long-term artificial intelligence Introduction Artificial intelligence (AI) experts are-to generalize-of two minds about long-term AI, especially regarding the potential for radically transformative types of future AI such as artificial superintelligence (ASI) a...
c26ee505-b232-4ec6-b7d9-1b278bdcb1aa
trentmkelly/LessWrong-43k
LessWrong
[Linkpost] Visual roadmap to strong human germline engineering (x.com, bsky.) This is a linkpost for "Visual roadmap to strong human germline engineering". That page gives a quick introduction to the technical problem of strong germline engineering, and gives a mildly-interactive visual roadmap. Here's a screenshot: ...
1fbbc505-d8f9-4b2b-aaa6-b33039ad35f4
trentmkelly/LessWrong-43k
LessWrong
What would you like from Microcovid.org? How valuable would it be to you? Microcovid.org was a vital tool to many of us during the pandemic- I made a whole speech about it back at summer solstice. The back and forth over how finished covid is, plus a dependence entirely on volunteers, has pushed microcovid into someth...
10bed20d-10cd-4498-a7f2-f94b0492706d
trentmkelly/LessWrong-43k
LessWrong
Meta-discussion from "Circling as Cousin to Rationality" A lot of the discussion on this post ended up being about LessWrong norms. I've moved that particular thread over to the comments here, and left a comment there pointing over here. (Some of those comments, including the initial one, were object-level relevant t...
41d00e99-f849-4007-a688-f32f48466538
trentmkelly/LessWrong-43k
LessWrong
Appropriateness of Discussing Rationalist Discourse of a Political Nature on LW? There was recently a back-and-forth between Slate Star Codex and Nathan Robinson of Current Affairs, a major national news magazine on the political left in the United States. It didn't end well. I think it would serve as a good example i...
f9f8b9d8-7be2-4f0f-8ebf-c23374edda88
trentmkelly/LessWrong-43k
LessWrong
10 Public "I was wrong" Admissions by Scientists and Intellectuals “I was wrong” Why are these three words so hard for the human tongue? This question has fascinated for a very long time…. Talk about the importance of integrity is common in intellectual circles but actual cases that demonstrate that are much harder t...
45701153-e9d0-4c58-9e9f-fe81f7fdae44
trentmkelly/LessWrong-43k
LessWrong
Meetup : Socializing and board games Discussion article for the meetup : Socializing and board games WHEN: 13 October 2015 07:00:00PM (+0300) WHERE: Google, Electra Tower, 98 Yigal Alon Street, Tel Aviv, Israel We're going to have a board game/socializing event. We haven't had them in a while because too many Tuesd...
153fa616-a205-4c47-b535-ae16dd6c959f
StampyAI/alignment-research-dataset/blogs
Blogs
against AI alignment ? against AI alignment ? ---------------------- [i usually consider AI alignment to be pretty critical](were-all-doomed.html). that said, there are some ways in which i can see the research that is generally associated with alignment to have more harmful potential than not, if it is applied. t...
da33de3e-4dbf-42a4-aaeb-edda4d3f2d98
StampyAI/alignment-research-dataset/eaforum
Effective Altruism Forum
My thoughts on nanotechnology strategy research as an EA cause area **Two-sentence summary:** Advanced nanotechnology might arrive in the next couple of decades (my wild guess: there’s a 1-2% chance in the absence of transformative AI) and could have very positive or very negative implications for existential risk. Th...
d7e39519-9607-4349-876a-07e834d384eb
awestover/filtering-for-misalignment
Redwood Research: Alek's Filtering Results
id: post1934 This post is written in my personal capacity, and does not necessarily represent the views of OpenAI or any other organization. Cross-posted to the Effective Altruism Forum In the previous post of this sequence , I argued that intent-aligned AIs would, by default, have incentives to break the law. This po...
143bc850-6ee7-459f-8566-9212cb164116
trentmkelly/LessWrong-43k
LessWrong
Text Posts from the Kids Group: 2021 Another round of liberating kid posts from Facebook. For reference, in 2021 Lily turned 7, Anna turned 5, and Nora was born. (Some of these were from me; some were from Julia. Ones saying "me" could mean either of us.) 2021-01-02 Anna: Hello, I'm Mr. Hamburger. Me: It's time t...
90597f3a-674f-441e-8c59-fc2fdd9401d8
trentmkelly/LessWrong-43k
LessWrong
[Link] Admitting to Bias Summary: Current social psychology research is probably on average compromised by political bias leftward. Conservative researchers are likely discriminated against in at least this field. More importantly papers and research that does not fit a liberal perspective faces greater barriers and b...
ed05ef24-4420-46dc-8fe5-33d478f833c6
trentmkelly/LessWrong-43k
LessWrong
A hermeneutic net for agency [Metadata: crossposted from https://tsvibt.blogspot.com/2023/09/a-hermeneutic-net-for-agency.html. First completed September 4, 2023.] A hermeneutic net for agency is a natural method to try, to solve a bunch of philosophical difficulties relatively quickly. Not to say that it would work....
3bdf0f8c-e6c1-46b7-892e-0875cbe474af
trentmkelly/LessWrong-43k
LessWrong
Why Do You Keep Having This Problem? One thing I've noticed recently is that when someone complains about how a certain issue "just keeps happening" or they "keep having to deal with it", it often seems to indicate an unsolved problem that people may not be aware of. Some examples: * Players of a game repeatedly ask...
f01bfb3e-ab5d-4090-8ee6-8ff35cbbdef3
trentmkelly/LessWrong-43k
LessWrong
Meetup : December Rationality meetup featuring Double Crux Discussion article for the meetup : December Rationality meetup featuring Double Crux WHEN: 10 December 2016 05:00:00PM (+0100) WHERE: Tibits Oerlikon, Zurich, Switzerland Double Crux is a strategy for resolving and illuminating disagreements by investigati...
d51eef69-7f3e-44b0-8aa8-c11eecb42813
trentmkelly/LessWrong-43k
LessWrong
A fun calibration game: "0-hit Google phrases" Here's a simple calibration game: propose some phrase(s), like "the ultimate east care pant" (something one of my pairs of pants say) and ask "How likely is it that Google return no search results for this phrase (in quotes)?"
c10260d3-007e-458b-b0ab-dae01c2d7a0d
trentmkelly/LessWrong-43k
LessWrong
Disentangling four motivations for acting in accordance with UDT Introduction In this post, I examine a number of different arguments for acting like an updateless agent. Concretely, I try to provide answers to the following question: What could move someone who isn’t already born a perfectly coherent UDT agent to ado...
1c386c73-4fdb-46eb-9c79-10a919fedc70
trentmkelly/LessWrong-43k
LessWrong
Intent-aligned AI systems deplete human agency: the need for agency foundations research in AI safety This is a summary of our recent preprint Intent-aligned AI systems deplete human agency: the need for agency foundations research in AI safety and submission to the Open Philanthropy AI Worldviews Contest.[1]. With re...