id
stringlengths
36
36
source
stringclasses
15 values
formatted_source
stringclasses
13 values
text
stringlengths
2
7.55M
06c48c01-5e69-46bc-8903-4d0719ac725f
StampyAI/alignment-research-dataset/eaforum
Effective Altruism Forum
Potential employees have a unique lever to influence the behaviors of AI labs [(Cross posted from my personal blog](https://oxalis.substack.com/p/potential-employees-have-a-unique)) People who have received and are considering an offer from an AI lab are in a uniquely good spot to influence the actions of that lab. ...
d732fc9b-f36b-4e64-b9f8-0e9f35390c7f
trentmkelly/LessWrong-43k
LessWrong
Sam Altman’s Chip Ambitions Undercut OpenAI’s Safety Strategy If you enjoy this, please consider subscribing to my Substack. Sam Altman has said he thinks that developing artificial general intelligence (AGI) could lead to human extinction, but OpenAI is trying to build it ASAP. Why? The common story for how AI coul...
8c87dd22-8379-47d4-a1d0-63472ea4305b
trentmkelly/LessWrong-43k
LessWrong
Visualizing Eutopia Followup to:  Not Taking Over the World "Heaven is a city 15,000 miles square or 6,000 miles around. One side is 245 miles longer than the length of the Great Wall of China. Walls surrounding Heaven are 396,000 times higher than the Great Wall of China and eight times as thick. Heaven has twelve g...
fa70b8ba-2684-4c2e-a3d0-6d02cc32f1ff
trentmkelly/LessWrong-43k
LessWrong
Unpricable Information and Certificate Hell Certificate: Expensive Shareable evidence of some desirable quality, typically of a marketable good. Increases the percieved value of that good to the market and so benefits the owner, but has a negative externality of lowering the percieved value of all similar uncertified ...
05bb1014-0f01-4967-9e73-48afc62b7920
awestover/filtering-for-misalignment
Redwood Research: Alek's Filtering Results
id: post1860 Clutching a bottle of whiskey in one hand and a shotgun in the other, John scoured the research literature for ideas... He discovered several papers that described software-assisted hardware recovery. The basic idea was simple: if hardware suffers more transient failures as it gets smaller, why not allow ...
dd7a47ce-ceb6-4763-a546-d93bca9493b1
StampyAI/alignment-research-dataset/lesswrong
LessWrong
Briefly thinking through some analogs of debate [This post is an almost direct summary of a conversation with John Wentworth.] Note: This post assumes that the reader knows what [AI Safety via Debate](https://www.alignmentforum.org/tag/debate-ai-safety-technique-1) is. I don't spend any time introducing it or explain...
3d2725af-bea8-4623-8c71-cde6cdfbfd88
trentmkelly/LessWrong-43k
LessWrong
Subsuming Purpose, Part 1   SUMMARY: The purpose of this entry is to establish the existence of local equilibriums which introduce deviations from an ends-driven organization (an organization whose primary focus is a particular purpose) to transform it into a means-driven organization (an organization whose primary ...
9fdf3cda-612e-4d2f-9e1d-6d9114795d5a
trentmkelly/LessWrong-43k
LessWrong
The Dead Cradle Theory: Why Earth May Not Survive Humanity's Expansion into Space I've been reading LessWrong for almost ten years, and finally decided to write my first post - any feedback is appreciated. This essay offers another angle on a known alignment challenge - why "AI will just leave Earth alone" is likely u...
577f66aa-eeae-426c-b9f7-edd71832bde5
StampyAI/alignment-research-dataset/lesswrong
LessWrong
AI safety university groups: a promising opportunity to reduce existential risk [*Cross-posted from the Effective Altruism Forum*](https://forum.effectivealtruism.org/posts/bL2CWihhaSjpqk8QD/ai-safety-university-groups-a-promising-opportunity-to) *Summary*: AI safety university groups[[1]](#fnvmyfd7tkhp) are a promis...
7796ed48-c61d-47af-a34b-eebf96e9368e
trentmkelly/LessWrong-43k
LessWrong
Towards Gears-Level Understanding of Agency Epistemic status: Highly speculative, much of this is probably wrong. Summary: The "centerpiece" of this post is my attempt to describe how a full-fledged consequentialist agent might incrementally grow out of a bundle of heuristics, aided by SGD or evolution. The story's d...
da97f38e-38e7-4531-b4d2-1945968d856e
trentmkelly/LessWrong-43k
LessWrong
An overview of control measures We often talk about ensuring control, which in the context of this doc refers to preventing AIs from being able to cause existential problems, even if the AIs attempt to subvert our countermeasures. To better contextualize control, I think it's useful to discuss the main countermeasures...
455c2cfb-fb73-4626-9393-4b4717f3cf71
awestover/filtering-for-misalignment
Redwood Research: Alek's Filtering Results
id: post1552 This post is a transcript of a conversation between Ege Erdil and Ronny Fernandez , recorded by me. The participants talked about a recent post by Katja Grace that presented many counterarguments to the basic case for AI x-risk. You might want to read that post first. As it was transcribed automatically b...
c2ac5845-f659-45f7-a681-ead6ad2de0af
trentmkelly/LessWrong-43k
LessWrong
An unconfortable epistemic position regarding "Verb Magic". Hello Guys, This post will act as a bottle thrown in the sea, a "Is anyone like me anywhere ?" as well as an occasion, i hope, to discuss a more serious topic with you. Some claims made here might push on your "pseudo-science" button, please read until the e...
b8672b69-33e0-4227-ac24-f0d55759c992
trentmkelly/LessWrong-43k
LessWrong
Meetup : Chicago Rationality Reading Group - Luminosity Part 3 Discussion article for the meetup : Chicago Rationality Reading Group - Luminosity Part 3 WHEN: 05 March 2017 01:00:00PM (-0600) WHERE: Harper Memorial Library Room 148, 1116 E 59th St, Chicago, IL 60637 The Chicago Rationality Reading Group meets every...
94380bb6-fc27-48c0-a2ed-3874f9fa7715
trentmkelly/LessWrong-43k
LessWrong
Orthogonality Thesis seems wrong Orthogonality Thesis (as well as Fact–value distinction) is based on an assumption that objective norms / values do not exist. In my opinion AGI would not make this assumption, it is a logical fallacy, specifically argument from ignorance. As black swan theory says - there are unknown ...
5c6c9c53-dc7a-4262-81cd-7619d5bcb8ac
trentmkelly/LessWrong-43k
LessWrong
Turning 22 in the Pre-Apocalypse Meta comment for LessWrong readers[1] Something Different This Way Comes - Part 1 In which I attempt to renegotiate rationalism as a personal philosophy, and offer my alternative—Game theory is not a substitute for real life—Heuristics over theories Introduction This essay focuses ...
bce27b94-cfb3-49dd-a770-fe33ba669e55
trentmkelly/LessWrong-43k
LessWrong
Recent Progress in the Theory of Neural Networks , It's common wisdom that neural networks are basically "matrix multiplications that nobody understands" , impenetrable to theoretical analysis, which have achieved great results largely through trial-and-error. While this may have been true in the past, recently there ...
d1225eda-2268-469c-9e85-1d6ee8b0b984
trentmkelly/LessWrong-43k
LessWrong
Self-education best practices I have a degree in a field I no longer wish to pursue in my career (English literature). Without a university and all it comes with (libraries, knowledgeable people, etc.), what are the best ways to self-educate and gain a substantive knowledge base in subjects I have little background in...
9793de03-2891-4825-bee1-aabd44ecf939
awestover/filtering-for-misalignment
Redwood Research: Alek's Filtering Results
id: post395 The coalitional frame Earlier in this sequence I laid out an argument that the goals of increasingly intelligent AIs will become increasingly systematized, until they converge to squiggle-maximization. In my last post , though, I touched on two reasons why this convergence might not happen: humans trying t...
f8066d90-5c37-41a0-bc16-dda25bd3849b
trentmkelly/LessWrong-43k
LessWrong
Covid 1/20/22: Peak Omicron The Omicron and Weekly posts are now combined, so this includes the last day’s Omicron developments, which will be how it works going forward. Next week’s will include the probability updates section as well. When there is urgent news, I’ll supplement as appropriate. This week was the peak...
d81bb067-fd14-44a7-ba8a-eb8f1171d4a9
trentmkelly/LessWrong-43k
LessWrong
Schelling Day 2.0 Schelling Day is a holiday about getting to know the people in your community that we created and celebrated in 2013. By popular request, I've revised the procedure to take into account what we learned last year. I'm aware of plans to hold Schelling Day in Boston, New York, and San Francisco on April...
5bbd5ec9-4563-434d-8f80-844d4bb314dd
trentmkelly/LessWrong-43k
LessWrong
Submit comments on Paxlovid to the FDA (deadline Nov 29th). 1,000 Americans are dying daily from COVID-19. Paxlovid is 89% effective against it. Pfizer submitted an EUA on Paxlovid to the FDA on Nov. 16th. The FDA has yet to schedule a meeting to review this EUA request, 11 days later. Thousands of lives are at stake ...
344e826e-5165-4cfe-be10-77a1fc9658a1
trentmkelly/LessWrong-43k
LessWrong
Value Loading This article was originally on the FHI wiki and is being reposted to LW Discussion with permission. All content in this article is credited to Daniel Dewey. In value loading, the agent will pick the action: Here A is the set of actions the agent can take, e is the evidence the agent has already seen,...
23764443-c014-43a8-987b-6fe307d10c02
trentmkelly/LessWrong-43k
LessWrong
Knitting a Sweater in a Burning House There's a phrase my wife and I use: "knitting a sweater in a burning house." It describes those moments when we find ourselves absorbed in trivial tasks while seemingly more important matters loom. The night before my wedding, I caught myself trying to reactivate an old Twitter pr...
dd7e2429-89bc-49be-aaf6-924beb437e2f
trentmkelly/LessWrong-43k
LessWrong
A car journey with conservative evangelicals - Understanding some British political-religious beliefs I’m heading home from a family wedding this weekend. I had a plane ticket, but in the end, decided to travel back with two of my uncles and my cousin. Most of my dad’s family are evangelicals, my aunts and uncles are ...
17533e83-7f25-4769-afd2-8ac7aad60c89
trentmkelly/LessWrong-43k
LessWrong
MIT Technology Review - Michael Hendricks opinion on Cryonics http://www.technologyreview.com/view/541311/the-false-science-of-cryonics/ Michael Hendricks is a neuroscientist and assistant professor of biology at McGill University.
bde27c01-6f50-4576-bf0d-b05f87cef7ea
trentmkelly/LessWrong-43k
LessWrong
Direct work vs. earn-to-give? I've been working in healthcare technology for 5+ years and am hitting a crossroads in terms of where I want to go in my career. Ever since discovering EA, I've been questioning whether I want to prioritize direct work vs. earn-to-give. The core reason why I work in health tech is to mak...
8b73d25a-ef2d-4dae-b921-d1da72b6b4be
StampyAI/alignment-research-dataset/arxiv
Arxiv
Very Deep Convolutional Networks for Large-Scale Image Recognition 1 Introduction --------------- Convolutional networks (ConvNets) have recently enjoyed a great success in large-scale image and video recognition (Krizhevsky et al., [2012](#bib.bib22); Zeiler & Fergus, [2013](#bib.bib34); Sermanet et al., [2014](#b...
d8917e39-c2a5-4d34-94c3-a1e6ba095ee8
trentmkelly/LessWrong-43k
LessWrong
Axis-49 to Jammer MIDI Mapper One of the risks with new instruments is that you might put a lot of work learning how to play something that then is discontinued. In the case of the jammer, things are worse because it was discontinued five years before I even got into the instrument. As I see used Axis-49s come up on r...
4fed91b4-1071-4af6-81aa-1130d98fa49a
trentmkelly/LessWrong-43k
LessWrong
[Link] Nick Bostrom on the Simulation Argument Today's Philosophy Bites features Nick Bostrom on the Simulation Argument. Here is a direct link to the MP3 file.
f36f6d99-f183-4457-b541-565af5d21ce8
trentmkelly/LessWrong-43k
LessWrong
Computer security story NPR has this story: The hacktivist group Anonymous is at it again. This time, it has humiliated an Internet security firm that threatened to out the group's hierarchy. If you remember, Anonymous has been in the news, first, because in support of WikiLeaks, it undertook cyberattacks that broug...
0509c9f5-1073-4c4d-80ff-bdbc934828d2
StampyAI/alignment-research-dataset/arxiv
Arxiv
Computational Extensive-Form Games. 1 Introduction --------------- Game-theoretic models assume that the players are completely rational. This is typically interpreted as saying that payers act optimally given (their beliefs about) other players’ behavior. However, as was first pointed out by Simon \citeyearSimon55...
41acdf4c-b4c7-40b3-b1d5-7724139c3625
StampyAI/alignment-research-dataset/arxiv
Arxiv
What are you optimizing for? Aligning Recommender Systems with Human Values 1 Introduction --------------- Recommender systems are AI systems that present users with a tailored set of items based on factors such as past user behavior, user attributes, and features of the underlying items. They rank huge numbers of ...
48dcb568-31ad-4c93-af00-124745df51ae
trentmkelly/LessWrong-43k
LessWrong
Request for Book Recommendations/Comments As you can probably guess from the title, this is a request for some brief recommendations of books of lists of books   Previous Threads: The Best Textbooks on Every Subject Book Recommendations Great Books of Failure CogSci Books
b263f488-76a5-46b8-b498-f3cf34a97f20
trentmkelly/LessWrong-43k
LessWrong
Scaffolding Skills Epistemic status: Exploratory A scaffold is a lightweight, temporary construction whose point is to make working on other buildings easier. Maybe you could live and work on a scaffold, but why would you? The point is the building. When the building is done, you can take down the scaffold, though yo...
579eb29a-87f6-449c-b6a6-8150671a9674
trentmkelly/LessWrong-43k
LessWrong
Fundamentals of Formalisation Level 6: Turing Machines and the Halting Problem Followup to Fundamentals of Formalisation Level 5: Formal Proof. First post. This is a new lesson of our online course on math formalizations required for AI safety research. ---------------------------------------- The big ideas: * Tu...
df077134-e05f-4b31-821f-cd758b228b65
trentmkelly/LessWrong-43k
LessWrong
Ten Commandments for Aspiring Superforecasters Cross-posted to the Effective Altruism Forum In the last several years, political scientist and forecasting research pioneer Philip Tetlock has made waves for the success of his research program in geopolitical forecasting, published in the form of the popular book Su...
2a1aad82-ddd8-48d9-b7c6-8c3aa2f7f713
trentmkelly/LessWrong-43k
LessWrong
A counterargument that isn't (cross-posted from my blog) Let's say that you're in a discussion, and it goes something like this: * We should do institute regulatory measures, like taxes, to inhibit the growth of the economy so that we can slow down or stop climate change. * That makes sense if you think that clima...
3a6d820c-7cfd-48c3-8b53-abd3515d83b8
trentmkelly/LessWrong-43k
LessWrong
Hidden philosophical progress Bertrand Russell: > If you ask a mathematician, a mineralogist, a historian, or any other man of learning, what definite body of truths has been ascertained by his science, his answer will last as long as you are willing to listen. But if you put the same question to a philosopher, he wi...
83df2a7e-6e10-4079-85b1-99985a6772f6
trentmkelly/LessWrong-43k
LessWrong
Extrapolating values without outsourcing I first took note of "Coherent Extrapolated Volition" in 2006. I thought it was a brilliant idea, an exact specification of how to arrive at a better future: Figure out exactly how it is that humans make their existing choices, idealize that human decision procedure according t...
23825e2b-4c57-495d-b138-e751fcee41d9
trentmkelly/LessWrong-43k
LessWrong
What Should I Do? I have been having some difficulty deciding what to do with my life. I'm not really ashamed or surprised, because the problem seems extremely difficult and worth getting right. I still don't really know. Here are some of my options, though I wouldn't be surprised if what I end up doing is not on the ...
5facde51-54ad-4780-b807-498efe1af5d6
trentmkelly/LessWrong-43k
LessWrong
Compatibilism in action A practical albeit fictional application of the philosophical conclusion that free will is compatible with determinism came up today in a discussion about a setting element from the role-playing game Exalted (5:31:44 PM) Nekira Sudacne: So during the pirmodial war, one Yozi got his fetch kille...
e8d2dab1-bd3c-4c57-b96a-b3709ef4658e
trentmkelly/LessWrong-43k
LessWrong
Prioritizing Happiness When the limiting resource is money it's quite clear that we should prioritize the uses where it goes the farthest. If there are three organizations that can distribute antimalarial nets for $5/each, $50/each, and $500/each we should just give to the first one. Similarly, if I have $5 I could us...
6a588ef9-35f8-4f23-9725-e88f0cc81b10
trentmkelly/LessWrong-43k
LessWrong
Altruistic Kidney Donation People are dying because their kidneys have failed, and you probably have an extra one: should you donate it? A dialysis machine can do most of the work of a kidney, but it's not as good. A donated kidney gives someone an average of 10 additional years. [1] (Some of them reject it right away...
9560da83-62f7-455f-826a-d7e0dcb0f4fb
trentmkelly/LessWrong-43k
LessWrong
(Rational) Decision-Making In Wartime   The martial art of rationality provides for seeing what the science of rationality means for ourselves, for our daily inner life. If so, I wonder whether war complicates exercising the muscles of sound reasoning or alters what rationality means for me. Along with the systematic ...
8de4c7b3-8453-4556-8db1-16db22dbe284
StampyAI/alignment-research-dataset/eaforum
Effective Altruism Forum
Assistant-professor-ranked AI ethics philosopher job opportunity at Canterbury University, New Zealand Canterbury University, New Zealand seeking assistant-professor-ranked philosopher of AI ethics: <https://jobs.canterbury.ac.nz/jobdetails/ajid/uUIS7/Lecturer-in-Philosophy,8123> Note that the position is for a "Lect...
27c46e96-c827-4174-8cc6-9d04ed2ab48c
StampyAI/alignment-research-dataset/arxiv
Arxiv
Efficient estimates of optimal transport via low-dimensional embeddings 1 Introduction --------------- Optimal Transport metrics (Kantorovich, [1960](#bib.bib2)) or Wasserstein distances, have emerged successfully in the field of machine learning, as outlined in the review by Peyré et al. ([2017](#bib.bib3)). They ...
34a0e02e-ca0d-4a49-90ea-cb82780ae24f
trentmkelly/LessWrong-43k
LessWrong
Verifying Rationality via RationalPoker.com Related to: Problem of verifying rationality We're excited to announce the (soft) launch of RationalPoker.com! It's a new guide developed by me, Zvi, Kevin, and patrissimo detailing how to use online poker as rationality training to conquer your cognitive biases. We want ou...
2c11b079-9557-4dd9-92a2-e3aac86b15c6
trentmkelly/LessWrong-43k
LessWrong
What's the Relationship Between "Human Values" and the Brain's Reward System? In classic, non-mesa-optimized AGI risk scenarios, an AI is typically imagined whose reward function is directly related to the optimization pressure that it exerts on the world: e.g. the paperclip maximizer. However, it seems that human val...
8da94375-54bd-4b24-b859-3d7e9bd2252b
LDJnr/LessWrong-Amplify-Instruct
LessWrong
"Suppose I were to say that the American legal system is a criminal organization. The usual response would be that this is a crazy accusation.Now, suppose I were to point out that it is standard practice for American lawyers to advise their clients to lie under oath in certain circumstances. I expect that this would st...
709dd749-da54-434b-8ba5-9c223e9eb2fa
StampyAI/alignment-research-dataset/special_docs
Other
An Introduction to Löb’s Theorem in MIRI Research An Introduction to L ob's Theorem in MIRI Research Patrick LaVictoire February 23, 2015 Contents 1 Introduction 2 2 Crash Course in L ob's Theorem 2 2.1 G odelian self-reference and quining programs . . . . . . . . . . . . . . . . . . 2 2.2 L ob's Theorem . . . . ....
da124a15-6c2b-4179-bc41-fe36e6e4f142
trentmkelly/LessWrong-43k
LessWrong
Meetup : Washington, D.C.: Fun & Games Discussion article for the meetup : Washington, D.C.: Fun & Games WHEN: 12 June 2016 03:00:00PM (-0400) WHERE: Old Patent Office Building We will be meeting in the courtyard to play games, hang out, and engage in fun conversation. Upcoming meetups: * Jun. 19: History of Sci...
facb8da2-02e2-4684-9973-b110f7d877f1
StampyAI/alignment-research-dataset/special_docs
Other
Human biases limit cumulative innovation. royalsocietypublishing.org/journal/rspb Research Cite this article: Thompson B, Griffiths TL. 2021 Human biases limit cumulativeinnovation. Proc. R. Soc. B 288: 20202752. https://doi.org/10.1098/rspb.2020.2752 Received: 3 November 2020 Accepted: 8 February 2021 Subject Categor...
4bc0c559-3b56-4ec5-b06f-0f24d9dcdd89
trentmkelly/LessWrong-43k
LessWrong
The Mere Cable Channel Addition Paradox The following is a dialogue intended to illustrate what I think may be a serious logical flaw in some of the conclusions drawn from the famous Mere Addition Paradox.  EDIT:  To make this clearer, the interpretation of the Mere Addition Paradox this post is intended to criticize...
e764e068-f44b-4270-a96b-2d702fdf955b
trentmkelly/LessWrong-43k
LessWrong
You Are Likely To Be Eaten By A Grue This was a mistaken double post.  The real post is here.  Please do not comment on this post.
56ced7bc-6da8-48cc-a42c-2d98eded2cb1
trentmkelly/LessWrong-43k
LessWrong
Chapter 28: Reductionism Whatever can go Rowling will go Rowling. This should again go without saying, but views expressed by Severus Snape are not necessarily those of the author. ---------------------------------------- "Okay," Harry said, swallowing. "Okay, Hermione, it's enough, you can stop." The white sugar ...
02ab0e67-a7d9-4490-be22-acd52e7bb33a
trentmkelly/LessWrong-43k
LessWrong
Effectiveness or altruism? Effective altruism is all the rage. And as we all know, a highly effectively altruistic thing to do is bring about more effective altruism. However, the optimal mixture of activities for promoting effectiveness is not identical to that for promoting altruism. They are upheld by different nor...
c80f8ef2-cfdc-4877-93d2-9d86faf560ee
StampyAI/alignment-research-dataset/lesswrong
LessWrong
Benchmarks for Comparing Human and AI Intelligence *Epistemic status: Been thinking about this for about a week, and done some basic googling.*   **Intro** Creating benchmarks for measuring and comparing the intelligence of AIs to humans seems useful for tracking progression and making future predictions of develo...
a2eb26f7-f623-499f-aaa4-46c777d30894
trentmkelly/LessWrong-43k
LessWrong
4. A Moral Case for Evolved-Sapience-Chauvinism Part 4 of AI, Alignment, and Ethics. This will probably make more sense if you start with Part 1. TL;DR In Parts 1 and 2 I argued that aligned AIs will not want and should not get assigned moral worth or the sorts of rights that human citizens get. In Part 3 I suggeste...
91c04334-7ace-471e-8135-1504cc1b97ad
trentmkelly/LessWrong-43k
LessWrong
What Should We Optimize - A Conversation The following is a lightly edited chat conversation from 2022-04-07. My interlocutor requested to be anonymized as Jason. This conversation sprang out of Jason reading about my transhuman dream. (The conversation starts out a bit slow, but overall covers a lot of ground.) Jas...
39b8d7f5-52c1-47a2-b3e2-407caf423f56
StampyAI/alignment-research-dataset/arbital
Arbital
Averting instrumental pressures Many subproblems of [corrigibility](https://arbital.com/p/45) involve [convergent](https://arbital.com/p/10g) instrumental [pressures](https://arbital.com/p/10k) to implement [strategies](https://arbital.com/p/2vl) that are highly anti-corrigible. Whether you're trying to maximize pape...
b39c121a-ca5c-4cf5-8604-51103be3c9b7
trentmkelly/LessWrong-43k
LessWrong
[Beta Feature] Google-Docs-like editing for LessWrong posts TL;DR: LessWrong now has similar features to Google Docs. Warning! Still rough around the edges. To enable the collaborative editor, you must check "Opt into experimental features" in your account settings and then press the green "Share" button that appears ...
a5be8211-353f-4311-a20c-b48930151c6e
trentmkelly/LessWrong-43k
LessWrong
Why no archive of refuted research? From The Atlantic's Lies, Damned Lies and Medical Science: > Still, Ioannidis anticipated that the community might shrug off his findings: sure, a lot of dubious research makes it into journals, but we researchers and physicians know to ignore it and focus on the good stuff, so wha...
9341ec5c-dc85-4605-a148-1a144e649d9d
trentmkelly/LessWrong-43k
LessWrong
Be comfortable with hypocrisy Neal Stephenson's The Diamond Age takes place several decades in the future and this conversation is looking back on the present day: > "You know, when I was a young man, hypocrisy was deemed the worst of vices,” Finkle-McGraw said. “It was all because of moral relativism. You see, in th...
1fad3885-d6f9-49c1-8e99-e5ac21bdbbc8
trentmkelly/LessWrong-43k
LessWrong
My new rationality/futurism podcast I've started a podcast called Future Strategist which will focus on decision making and futurism.  I have created seven shows so far:  interviews of computer scientist Roman Yampolskiy, LW contributor Gleb Tsipursky, and artist/free speech activist Rachel Haywire, and monologues on ...
3e559a7a-e3c4-4074-9689-3b227ec61ef1
StampyAI/alignment-research-dataset/aisafety.info
AI Safety Info
Are there any detailed example stories of what unaligned AGI would look like? Stories about the future in scientific fields always [risk being seen as sci-fi](https://www.youtube.com/watch?v=JVIqp_lIwZg), because it hasn’t happened yet and science can never say with 100% certainty that something will or won’t occur in...
d63d88dc-b574-4f8d-87ef-2663dda1422c
trentmkelly/LessWrong-43k
LessWrong
New paper: The Incentives that Shape Behaviour Abstract: Which variables does an agent have an incentive to control with its decision, and which variables does it have an incentive to respond to? We formalise these incentives, and demonstrate unique graphical criteria for detecting them in any single decision causal i...
2c03792f-a412-4310-b8b8-9f97009a4315
trentmkelly/LessWrong-43k
LessWrong
Soft skills for meetups This post is adapted from a memo I wrote for the 2022 global meetup organizer retreat, and owes some of its content to the organizers who attended. ---------------------------------------- 'Soft skills' is a broad and somewhat fuzzy term, but the OED defines it as "personal attributes that en...
d99784c3-fd31-461c-9e9a-3ab828e829ee
trentmkelly/LessWrong-43k
LessWrong
Zeckhauser's roulette Imagine you're playing Russian roulette. Case 1: a six-shooter contains four bullets, and you're asked how much you'll pay to remove one of them.  Case 2: a six-shooter contains two bullets, and you're asked how much you'll pay to remove both of them. Steven Landsburg describes an argument by Ric...
1ed416d4-5a9c-4d51-9f87-ccece10c2fa4
trentmkelly/LessWrong-43k
LessWrong
Steelmanning as an alternative to Rationalist Taboo The problem with Rationalist Taboo is that it was invented to make others clarify their thoughts, not to clarify your own. Here's how it usually goes wrong: "You can't explain the color red to me? Hahaha, I've dissolved the idea of qualia! They are just an illusion!...
12397ea0-13e0-48c2-8d48-e01d660a091d
trentmkelly/LessWrong-43k
LessWrong
Estimate the Cost of Immortality How much money would it take to engineer biological immortality for at least half of the world's population, within 20 years, with 99% confidence?  
a10d4876-dc12-4845-9f90-7e4b2e85e287
StampyAI/alignment-research-dataset/youtube
Youtube Transcripts
The Inside View #3–Evan Hubinger—Takeoff speeds, Risks from learned optimization & Interpretability to give a bit of context to the viewers um so mira is the machine intelligence research institute can you just like give a brief recap on like how long have you been working there and what you do there and what have you...
d3185460-e157-4f25-90db-0bda379b2754
trentmkelly/LessWrong-43k
LessWrong
Meetup : MIRI paper reading party Discussion article for the meetup : MIRI paper reading party WHEN: 05 April 2015 12:00:00PM (-0700) WHERE: 109 15th Ave, Seattle WA, 98122 https://www.facebook.com/events/860474784017422/ We will be following: https://intelligence.org/2014/12/23/new-technical-research-agenda-overv...
dca2e732-f82c-4e24-8990-7098ffe8b319
trentmkelly/LessWrong-43k
LessWrong
A list of Petrov buttons A thing people sometimes do in relation to Petrov day, is to have a button. And if anyone presses the button, something negative happens. I'd like to know how often the button gets pressed, so I compiled a list of all the times people have done this, that I'm aware of, and the outcome. Also, ...
fd7df517-1271-4df4-b3a8-2ea273ff70ab
StampyAI/alignment-research-dataset/eaforum
Effective Altruism Forum
Observations on the funding landscape of EA and AI safety *Epistemic status: Hot takes for discussion. These observations are a side product of another strategy project, rather than a systematic and rigorous analysis of the funding landscape, and we may be missing important considerations. Observations are also non-ex...
afa8def1-423c-4595-a5e8-346d5371023c
StampyAI/alignment-research-dataset/eaforum
Effective Altruism Forum
How Open Source Machine Learning Software Shapes AI **Acknowledgements** We would like to thank many different people for contributing to the research that comprised this post. Tim Hwang and Teddy Collins at CSET for originally sparking this research question, Helen Toner for her early encouragement and feedback on r...
aeabf292-72a5-44f9-ab7f-18c53fc18674
trentmkelly/LessWrong-43k
LessWrong
Sabbath hard and go home Growing up Jewish, I thought that the traditional rules around the Sabbath were silly. Then I forgot to bring a spare battery on a camping trip. Now I think that something like the traditional Jewish Sabbath is an important cultural adaptation to preserve leisure, that would otherwise be destr...
696dfed6-1143-4e80-925f-13d6db5ba356
trentmkelly/LessWrong-43k
LessWrong
An Apprentice Experiment in Python Programming A couple weeks ago Zvi made an apprentice thread. I have always wanted to be someone's apprentice, but it didn't occur to me that I could just ...ask to do that. Mainly I was concerned about this being too big of an ask. I saw gilch's comment offering to mentor Python pro...
e4a6faa8-d834-4269-a595-d2dd618f5891
StampyAI/alignment-research-dataset/lesswrong
LessWrong
Are you stably aligned? Epistimic status: posing questions I think are probably useful, without a strong guess at the answers. If we made an AGI just like you, would it be aligned? Would it stay aligned over a long time span? The purpose of the first question is to explore the possibility raised by shard theorya...
5587c3c9-0f07-4411-bbfc-cce0f63eb1a2
trentmkelly/LessWrong-43k
LessWrong
11 minute TED talk is about instrumental rationality Link Essentially this is about cutting your enemy, and some procedural pitfalls in his examples. I didn't spot any novel principles in the talk, but I want to learn more about how he thinks.
9f118993-afb5-436d-b50d-71fe9f54e51a
trentmkelly/LessWrong-43k
LessWrong
Meetup : Washington, D.C.: TED Talks Discussion article for the meetup : Washington, D.C.: TED Talks WHEN: 14 August 2016 03:30:00PM (-0400) WHERE: Mosaic at Metro Apartments This week, we will be meeting at the home of one of the meetup regulars (6210 Belcrest Road Apt 1131) to watch TED talks. Directions have bee...
dd65b733-4665-49d6-8fba-520c459a66b6
trentmkelly/LessWrong-43k
LessWrong
Can AI Quantity beat AI Quality? AI definitely poses an existential risk, in the sense that it can generate models with the hidden (possibly undetectable?) intention of competing against humanity for resources. The more intelligent the model, the higher its chance of success! The thought of an AI takeover is so scary...
62821d43-b2cb-4a5b-b0c4-70eebd0689a0
trentmkelly/LessWrong-43k
LessWrong
(Religion/Atheist) - Reppert on the Defeasibility Test (No (conclusive) evidence possible for a being both all-powerful and honest)
674f7fbf-1549-4709-88c6-7e46417c0897
trentmkelly/LessWrong-43k
LessWrong
Drawing Less Wrong: Should You Learn to Draw? This is the second post of the Drawing Less Wrong mini sequence, in which I discuss how to draw, how learning to draw *effectively* relates to rationality, and what the initial results were when I started running a drawing workshop, teaching people with essentially no expe...
5d5e3a0a-62bd-4c0a-91a3-960e84f0a3b5
trentmkelly/LessWrong-43k
LessWrong
asking an AI to make itself friendly   edit: I think I have phrased this really poorly and that this has been misinterpreted. See my comment below for clarification.   A lot of thought has been put into the discussion of how one would need to define the goals of an AI so that it won't find any "loopholes" and act i...
da4baf1b-c488-48b3-b944-4a907a5321c1
trentmkelly/LessWrong-43k
LessWrong
Secret Rationality Base in Europe In short, I'm wondering what place/group/organisation/activity could do for rationality in Europe what Berkeley does for rationality in the US?   Soon, we'll have LWCW in Berlin, which I hope will be an occasion to do some networking among people who think seriously about developing...
cf7fa474-48b4-47e2-a94b-a31178c37af4
trentmkelly/LessWrong-43k
LessWrong
Evil autocomplete: Existential Risk and Next-Token Predictors NOTE: The following essay was first summarized by Yitz, generated via prompting by ChatGPT, then edited by Yitz, and finally rewritten by Bing Chat. I have left all text as Bing generated it (though I did enlarge section headings, and removed some of the ea...
4aa7da9b-4e1b-4673-ac2d-436ff8b2cab2
trentmkelly/LessWrong-43k
LessWrong
A Brief Assessment of OpenAI's Preparedness Framework & Some Suggestions for Improvement This document was produced by SaferAI submitted for feedback to OpenAI. It is doing a list of remarkable improvements of the Preparedness Framework (PF) over Anthropic’s RSPs and suggests possible improvements to make the PF more ...
826b49a9-40da-44ed-a65d-163843ce514b
StampyAI/alignment-research-dataset/alignmentforum
Alignment Forum
Abstractions as Redundant Information What is it that makes the concept of “pencil” a good abstraction? One way to frame it: there are lots of “pencils” in the world - lots of little chunks of the world which I could look at which all contain information about pencils-in-general. Many different places from which I co...
726dd031-c517-4e46-8548-561aab25df28
StampyAI/alignment-research-dataset/arxiv
Arxiv
Improved Baselines with Momentum Contrastive Learning 1 Introduction --------------- Recent studies on unsupervised representation learning from images [[16](#bib.bib16), [13](#bib.bib13), [8](#bib.bib8), [17](#bib.bib17), [1](#bib.bib1), [9](#bib.bib9), [15](#bib.bib15), [6](#bib.bib6), [12](#bib.bib12), [2](#bib....
971f3941-9d3b-4089-934b-9cae504ef053
StampyAI/alignment-research-dataset/lesswrong
LessWrong
LW is to rationality as AIXI is to intelligence Apparently LW does a great job on refining rationality and dissolving confusions. But is it helpful when it comes to anything apart from designing Friendly AI, apart from a purely academic treatment of rationality? I'm currently unable to benefit from what I have so far ...
5b2e3047-d03b-4e6d-802b-b69ec9a70a07
trentmkelly/LessWrong-43k
LessWrong
Talking publicly about AI risk In the past year, I have started talking about AI risk publicly - in mainstream newspapers, public radio, national TV, some of the most popular podcasts. The twist, and reason why you probably haven't noticed is I'm doing this in Czech. This has a  large disadvantage - the positive impa...
e8badec0-d6e4-4e45-a5bb-12a35098daa0
trentmkelly/LessWrong-43k
LessWrong
[SEQ RERUN] Disputing Definitions Today's post, Disputing Definitions was originally published on 12 February 2008. A summary (taken from the LW wiki):   > You allow an argument to slide into being about definitions, even though it isn't what you originally wanted to argue about. If, before a dispute started about w...
97b0ccae-9671-4719-bd16-16561ff602c9
StampyAI/alignment-research-dataset/lesswrong
LessWrong
Do mesa-optimization problems correlate with low-slack? My understanding is that the primary problem with inner-alignment is that even outer-aligned AIs may be incredibly unpredictable due to unpredictable mesa-optimization. Put another way, there is a possibility that even air-tight objective functions could be ignor...
178272b2-bd7d-4b02-b978-943741b717e0
trentmkelly/LessWrong-43k
LessWrong
Conventional footnotes considered harmful Writers use footnotes — equally, endnotes — intending that they be optional for the reader. A note will hold a citation, technicality, or explanation, any of which is of interest to only some readers. This is a useful tactic, in principle. Footnotes are indicated with ordinal...
4cbdeb99-8968-4d8a-ab6d-44534a7c6697
trentmkelly/LessWrong-43k
LessWrong
Language Models Don't Learn the Physical Manifestation of Language Abstract We argue that there are certain properties of language that our current large language models (LLMs) don't learn. We present an empirical investigation of visual-auditory properties of language through a series of tasks, termed H-Test. This be...
5bf68256-ba68-42c6-947e-04c1db4c033b
trentmkelly/LessWrong-43k
LessWrong
What are the best books on evolutionary psychology? I'd like to divide three classes of reasons to read a discipline: 1) You are curious and want to begin reading by something 100-500 pages. I'd go for Pinker's 1990's  "How the mind works" 2) You want to screen the whole field, by reading something 500-1500 pages. I...
7bdf77b6-2eef-4ff4-898b-966835244d22
trentmkelly/LessWrong-43k
LessWrong
[LINK] Why I'm not on the Rationalist Masterlist A long blog post explains why the author, a feminist, is not comfortable with the rationalist community despite thinking it is "super cool and interesting". It's directed specifically at Yvain, but it's probably general enough to be of some interest here. http://apophe...
c83921f5-01bf-4a21-bdfa-187c0eb679d7
trentmkelly/LessWrong-43k
LessWrong
Peter Singer's first published piece on AI Hi! I am Tse Yip Fai. This is a linkpost to Peter Singer's first published piece on AI, it's also my first published paper (second author). Since the paper is open access, I won't be doing any copy-and-paste here.  This is a seminal paper resulting from an one-year research...
7e35e70b-74cc-41ae-ba0c-ef69b1268236
trentmkelly/LessWrong-43k
LessWrong
Vignettes Workshop (AI Impacts) AI Impacts is organizing an online gathering to write down how AI will go down! For more details, see this announcement, or read on. Plan 1. Try to write plausible future histories of the world, focusing on AI-relevant features. (“Vignettes.”) 2. Read each others’ vignettes and critiq...