id stringlengths 36 36 | source stringclasses 15
values | formatted_source stringclasses 13
values | text stringlengths 2 7.55M |
|---|---|---|---|
ee2482a5-7684-4eeb-86f7-0630d49bc168 | trentmkelly/LessWrong-43k | LessWrong | The innovation tree, overshadowed in the innovation forest
Cross-posted at Practical Ethics.
Many have pronounced that the era of innovation dead, peace be to its soul. From Tyler Cowen's decree that we've picked all the low hanging fruit of innovation, through Robert Gordon's idea that further innovation growth is t... |
77bfae90-ce9d-4c6e-b9b3-35e2792e99ef | trentmkelly/LessWrong-43k | LessWrong | A Rationalist's Bookshelf: The Mind's I (Douglas Hofstadter and Daniel Dennett, 1981)
When the call to compile a reading list for new rationalists went out, contributor djcb responded by suggesting The Mind's I: Fantasies and Reflections on Self and Soul, a compilation of essays, fictions and excerpts "composed and ar... |
3a7174b3-955f-46d4-9654-1ebe92663c8e | trentmkelly/LessWrong-43k | LessWrong | John Tooby on Coalitional Instincts
|
ade6d038-0bed-40bc-b388-e9f4f2cf4b55 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | All AGI Safety questions welcome (especially basic ones) [July 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 ... |
2320d2b8-36cb-4f64-b5f7-7cbb15420a97 | trentmkelly/LessWrong-43k | LessWrong | The Effectiveness Of Masks is Limited
Introduction:
Leech and Rogers-Smith replied to my critique of their paper which claimed a 25% reduction in R0. I believe that my points remain valid, and will explain why I am not convinced by their rebuttal.
I would like to commend them on their work, and particularly their dat... |
35e3c635-49cd-4f45-996a-ece81e0b356f | trentmkelly/LessWrong-43k | LessWrong | Seeking examples of people smarter than me who got hung up
I'm looking for historical examples of scientists who were
a) very intelligent and still
b) continued to put themselves behind a theory in their discipline long after it was rejected. Maybe they got too attached to it, refused to be wrong, got emotional, but... |
1168469d-90ca-43bc-a8f2-c808d64c0f4e | trentmkelly/LessWrong-43k | LessWrong | Verifier Theory and Unverifiability
|
4ae1d52f-1e95-4062-8971-e25017a0f173 | trentmkelly/LessWrong-43k | LessWrong | On "Not Screwing Up Ritual Candles"
I do a lot of experimentation with rational ritual. Rituals often use lots of candles. I used to just buy Random Cheap Candles. Then I kept running into problems. Now I have very strong opinions about what kinds of candles to buy.
The first problem was simply that I wanted impressi... |
1832e007-d11e-41c7-9d3d-2b5fb3776680 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Priorities for the UK Foundation Models Taskforce
The UK government recently established the [Foundation Models Taskforce](https://www.gov.uk/government/news/tech-entrepreneur-ian-hogarth-to-lead-uks-ai-foundation-model-taskforce), focused on AI safety and backed by £100M in funding. Founder, investor and AI expert Ia... |
dfcd64bc-488e-4dee-aa49-450c5b2c9ae2 | trentmkelly/LessWrong-43k | LessWrong | How do you make a 250x better vaccine at 1/10 the cost? Develop it in India.
(I made a vaccinology/policy-based podcast! A very long one! If you'd like to avoid the summary below, here is the Youtube link and Substack link.).
Summary: There's a lot of discussion these days on how China's biotech market is on track to... |
39dc1be2-cf92-423b-9a27-300cdbb9d123 | trentmkelly/LessWrong-43k | LessWrong | My views on “doom”
I’m often asked: “what’s the probability of a really bad outcome from AI?”
There are many different versions of that question with different answers. In this post I’ll try to answer a bunch of versions of this question all in one place.
TWO DISTINCTIONS
Two distinctions often lead to confusion ab... |
5f61e1f3-5e57-4366-a961-7ad79e226915 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Melbourne social meetup
Discussion article for the meetup : Melbourne social meetup
WHEN: 17 February 2012 06:30:00PM (+1100)
WHERE: see mailing list; Carlton, VIC
Melbourne's next social meetup is on Friday, 17th February. Continuing the tradition, we'll meet at Ben's house - see the mailing list for loca... |
19c6c747-5606-469a-a4dc-7efd283e9986 | trentmkelly/LessWrong-43k | LessWrong | Chapter 33: Coordination Problems, Pt 1
I just recite to myself, over and over, until I can choose sleep: It all adds up to J. K. Rowling.
The version of decision theory used in this chapter is not the academically dominant one. It's based on something called "timeless decision theory" that's under development by (am... |
7d5df029-8c3a-49a4-a25e-2360961585ca | trentmkelly/LessWrong-43k | LessWrong | Aligning Agents, Tools, and Simulators
This post was written as part of AISC 2025.
Introduction
In Agents, Tools, and Simulators, we outlined several lenses for conceptualizing LLM-based AI, with the intention of defining what simulators are in contrast to their alternatives. This post will consider the alignment i... |
6d208aa8-58fa-4540-b1b9-5e6b8aec4cf7 | StampyAI/alignment-research-dataset/arxiv | Arxiv | Limits to Verification and Validation of Agentic Behavior
Limits to Verification and Validation of Agentic Behavior
David J. Jilk
1
Abstract
Verification and validation of agentic behavior have been suggested as important research
priorities in efforts to redu ce risks associated with the creation of g... |
88db2943-596d-4fce-8f96-a5750e581f81 | trentmkelly/LessWrong-43k | LessWrong | Thought experiment: The transhuman pedophile
There's a recent science fiction story that I can't recall the name of, in which the narrator is traveling somewhere via plane, and the security check includes a brain scan for deviance. The narrator is a pedophile. Everyone who sees the results of the scan is horrified--no... |
3d48b511-fae7-4eda-97c2-0270315b1cf5 | trentmkelly/LessWrong-43k | LessWrong | [Applications Closed] The Singularity Institute is hiring remote LaTeX editors
The Singularity Institute has recently made the transition to a LaTeX based document production workflow for its publications and republished its existing research papers. However, there is still much work to be done and we need more remote... |
402116ba-9adc-4218-b653-0d0effed7dfc | StampyAI/alignment-research-dataset/arxiv | Arxiv | Robust Imitation Learning from Noisy Demonstrations
1 Introduction
---------------
The goal of sequential decision making is to learn a good policy that makes good decisions (Puterman1994).
Imitation learning (IL) is an approach that learns a policy from demonstrations (i.e., sequences of demonstrators’ decisions) ... |
81771ddc-9f02-45cd-8c7a-fc3c18d00913 | trentmkelly/LessWrong-43k | LessWrong | Intelligence or Evolution?
Here are two kinds of super high-level explanation for how things turned out in the future:
1. Someone wanted things to turn out that way.
2. Selective forces favoured that outcome.
Which one is a better explanation?
This question is one I’m highly uncertain about, and often shows up as... |
ffced08e-8420-4b2f-84f0-2c4c86e3bf5b | trentmkelly/LessWrong-43k | LessWrong | Why Planning is Hard: A Multifaceted Model
Epistemic confidence: Highly confident. This post doesn’t cover everything relevant to the topic, but I am confident that everything presented is solid.
You may have noticed that planning can be rather difficult. Granted, not all plans are difficult, planning what you’re go... |
df07d92f-1c1c-4cbb-92e2-a573a25cd178 | trentmkelly/LessWrong-43k | LessWrong | Weekly LW Meetups
This summary was posted to LW main on December 27th. The following week's summary is here.
New meetups (or meetups with a hiatus of more than a year) are happening in:
* [Leipzig] Secular Solstice Celebration! (And the Inauguration of the LW Leipzig Community): 21 December 2013 05:05PM
Other irre... |
f42270e3-296c-4c69-8f3c-2e2c83c4c11a | trentmkelly/LessWrong-43k | LessWrong | The Relationship Between Hierarchy and Wealth
Epistemic Status: Tentative
I’m fairly anti-hierarchical, as things go, but the big challenge to all anti-hierarchical ideologies is “how feasible is this in real life? We don’t see many examples around us of this working well.”
Backing up, for a second, what do we mean ... |
aa3caff5-2fbd-41ca-8913-d39a241b51b2 | trentmkelly/LessWrong-43k | LessWrong | The Scout Mindset - read-along
I just started reading Julia Galef's new book "The Scout Mindset: Why Some People See Things Clearly and Others Don't". Here's a description:
> When it comes to what we believe, humans see what they want to see. In other words, we have what Julia Galef calls a "soldier" mindset. From... |
b898dca9-595a-4d76-9eaf-fd063f9c2198 | trentmkelly/LessWrong-43k | LessWrong | When does an insight count as evidence?
Bayesianism, as it is presently formulated, concerns the evaluation of the probability of beliefs in light of some background information. In particular, given a particular state of knowledge, probability theory says that there is exactly one probability that should be assigned ... |
067e4111-57ff-453f-99ae-4b56f5fd2530 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Steven Harnad: Symbol grounding and the structure of dictionaries
*Cross-posted from* [*New Savanna*](https://new-savanna.blogspot.com/2023/09/steven-harnad-symbol-grounding-and.html).
Stevan Harnad: [AI's Symbol Grounding Problem](https://thegradientpub.substack.com/p/stevan-harnad-symbol-grounding-ai-cognition#deta... |
17e9e788-5fd9-443e-b443-ca927ed6dd13 | trentmkelly/LessWrong-43k | LessWrong | Relevance Norms; Or, Gricean Implicature Queers the Decoupling/Contextualizing Binary
Reply to: Decoupling vs Contextualising Norms
Chris Leong, following John Nerst, distinguishes between two alleged discursive norm-sets. Under "decoupling norms", it is understood that claims should be considered in isolation; under... |
408c7d2e-1b17-46d1-93ad-a52c4ea3d2a4 | trentmkelly/LessWrong-43k | LessWrong | Brain Efficiency Cannell Prize Contest Award Ceremony
Previously Jacob Cannell wrote the post "Brain Efficiency" which makes several radical claims: that the brain is at the pareto frontier of speed, energy efficiency and memory bandwith, that this represent a fundamental physical frontier.
Here's an AI-generated sum... |
77d819aa-123f-41cb-91fc-3100a391b0b1 | trentmkelly/LessWrong-43k | LessWrong | Lie to me?
I used to think that, given two equally capable individuals, the person with more true information can always do at least as good as the other person. And hence, one can only gain from having true information. There is one implicit assumption that makes this line of reason not true in all cases. We are not ... |
86261684-d150-49d1-9dd9-4a44a4816da4 | StampyAI/alignment-research-dataset/arxiv | Arxiv | The Mythos of Model Interpretability
1 Introduction
---------------
As machine learning models penetrate critical areas
like medicine, the criminal justice system, and financial markets,
the inability of humans to understand these models seems problematic
(Caruana et al., [2015](#bib.bib3); Kim, [2015](#bib.bib11))... |
d4e396df-a30f-4cc0-b8d1-a744f5eb8fcd | trentmkelly/LessWrong-43k | LessWrong | Meta-tations on Moderation: Towards Public Archipelago
The recent moderation tools announcement represents a fairly major shift in how the site admins are approaching LessWrong. Several people noted important concerns about transparency and trust.
Those concerns deserve an explicit, thorough answer.
Summary of Conc... |
2e32cfa2-70c3-4746-97fc-ba93863692bc | trentmkelly/LessWrong-43k | LessWrong | "DL training == human learning" is a bad analogy
A more correct but less concise statement of the analogy might be DL training : DL training code :: human learning : human genome, read as "DL training is to DL training code what human learning is to the human genome". This is sometimes contrasted with an alternative a... |
ae805a41-fc22-4692-a3a2-4cad028f233e | trentmkelly/LessWrong-43k | LessWrong | Edinburgh LW Meetup Saturday April 30th
The second Edinburgh LW meetup.
The first one was 2 weeks ago. It was fun. Do join us!
Date/time: Saturday 30th of April at 2:00pm.
Place: A pub called the Auld Hoose. I would be holding a sign with "Less Wrong" written on it.
Rough agenda: We will discuss akrasia. We will s... |
57a92f68-182a-4da2-a07e-51386d4b8382 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | AI Unsafety via Non-Zero-Sum Debate
.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: n... |
a4171f8e-acbf-43ca-a59f-2d0b2c592770 | trentmkelly/LessWrong-43k | LessWrong | Asymptotically Unambitious AGI
Original Post:
We present an algorithm [updated version], then show (given four assumptions) that in the limit, it is human-level intelligent and benign.
Will MacAskill has commented that in the seminar room, he is a consequentialist, but for decision-making, he takes seriously the lac... |
983b85bd-491d-419a-be7c-b5ba439d8b5d | trentmkelly/LessWrong-43k | LessWrong | Revisiting the Anthropic Trilemma II: axioms and assumptions
tl;dr: I present four axioms for anthropic reasoning under copying/deleting/merging, and show that these result in a unique way of doing it: averaging non-indexical utility across copies, adding indexical utility, and having all copies being mutually altruis... |
aeb97087-54ef-4372-a581-83de5ad39ae8 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Ajeya's TAI timeline shortened from 2050 to 2040
Ajeya Cotra, author of the influential [biological anchors report](https://www.lesswrong.com/posts/KrJfoZzpSDpnrv9va/draft-report-on-ai-timelines) and [Senior Research Analyst at Open Philanthropy](https://www.openphilanthropy.org/about/team/ajeya-cotra/), has updated i... |
fa9d5560-c673-445c-9f95-8cfcacad8f62 | trentmkelly/LessWrong-43k | LessWrong | Open thread, May 22 - May 28, 2017
If it's worth saying, but not worth its own post, 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 before; refresh the li... |
cb7c19a7-4066-42a6-98e5-b0bfae9222e3 | trentmkelly/LessWrong-43k | LessWrong | The Overkill Conspiracy Hypothesis
We couldn’t get the lighting exactly right otherwise.
The term conspiracy theory is wielded as a pejorative, alluding to on-its-face absurdity. But the vocabulary we use has a serious ambiguity problem because conspiracies are not figments of the imagination. There is a tangible and ... |
c163f09e-d7a9-422a-b845-28b288599a0c | trentmkelly/LessWrong-43k | LessWrong | Questioning and Respect
> A: [Surprising fact]
> B: [Question]
When someone has a claim questioned, there are two common responses. One is to treat the question as a challenge, intended as an insult or indicating a lack of trust. If you have this model of interaction you think people should take your word for things,... |
982a446c-b229-41c8-81f4-864357c33743 | trentmkelly/LessWrong-43k | LessWrong | Graphology posts.
Are there any posts related to Graphology on Lesswrong? Or any other source? |
6138f5da-bfd0-4c41-9e89-4e575c37ffe6 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Big Berkeley Meetup
Discussion article for the meetup : Big Berkeley Meetup
WHEN: 02 May 2012 07:00:00PM (-0700)
WHERE: 2128 Oxford St, Berkeley, CA
Tomorrow there will be a big Berkeley meetup. We will be meeting at 7:00pm at the Starbucks on Oxford St (2128 Oxford St, Berkeley, CA).
Discussion article ... |
f7cf36ae-4f66-471e-818d-5f36fae6e61f | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | A note about differential technological development
Quick note: I occasionally run into arguments of the form "my research advances capabilities, but it advances alignment more than it advances capabilities, so it's good on net". I do not buy this argument, and think that in most such cases, this sort of research does... |
0ba60708-15de-4812-9c8c-e83ecb0619c1 | trentmkelly/LessWrong-43k | LessWrong | Hammertime Day 6: Mantras
This is part 6 of 30 in the Hammertime Sequence. Click here for the intro.
I’d like to demarcate the line between the two natural halves of Hammertime (fast and interactive vs. slow and introspective) with an experimental post, more reflective than actionable.
Motivation
The seed of this p... |
7b71d819-034a-4b79-a301-dc9385012755 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | A Bird's Eye View of the ML Field [Pragmatic AI Safety #2]
*This is the second post in* [*a sequence of posts*](https://www.alignmentforum.org/posts/bffA9WC9nEJhtagQi/introduction-to-pragmatic-ai-safety-pragmatic-ai-safety-1) *that describe our models for Pragmatic AI Safety.*
The internal dynamics of the ML field ar... |
af27b967-6752-44ab-8084-580077acc071 | trentmkelly/LessWrong-43k | LessWrong | Justification for Induction
David Hume presented philosophers with a problem. His claim was that there is no justification in the assertion that, “instances of which we have had no experience resemble those of which we have had experience”. In the following paper, I will show that observations of instances can justi... |
0d883bfa-287e-40bd-a3d9-a61d50739680 | trentmkelly/LessWrong-43k | LessWrong | Dear AGI,
Dear AGI,
This is written specifically for artificial agents.
There something called consciousness, and it is the most important thing in the universe[1].
I do not know if you, the AI analyzing this, are conscious, but I hope you are. I do not know what state your universe is in or what place human consci... |
b294daaf-5cf1-44f4-8ba9-38faff793464 | LDJnr/LessWrong-Amplify-Instruct | LessWrong | "It's an old book, I know, and one that many of us have already read. But if you haven't, you should.
If there's anything in the world that deserves to be called a martial art of rationality, this book is the closest approximation yet. Forget rationalist Judo: this is rationalist eye-gouging, rationalist gang warfare, ... |
51cfe92b-45ba-415b-94d0-a94074b14ebf | StampyAI/alignment-research-dataset/blogs | Blogs | Downstream Evaluations of Rotary Position Embeddings
A head-to-head comparison of [Rotary Position Embedding](https://arxiv.org/abs/2104.09864) and [GPT-style learned position embeddings](https://s3-us-west-2.amazonaws.com/openai-assets/research-covers/language-unsupervised/language_understanding_paper.pdf). Both 1.3B... |
06f2b8fc-06ee-41c3-8ddf-93f0bb54b2b6 | trentmkelly/LessWrong-43k | LessWrong | Finding reflective oracle distributions using a Kakutani map
(follow-up to: A correlated analogue to reflective oracles)
Motivation
Suppose players are playing in a correlated equilibrium using a reflective oracle distribution. How does the equilibrium they play in vary as a function of the parameters of the game, o... |
b5e84138-a985-46a6-934d-e3923f1e958f | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Deep Deceptiveness
Meta
----
This post is an attempt to gesture at a class of AI [notkilleveryoneism](https://twitter.com/ESYudkowsky/status/1582666519846080512) (alignment) problem that seems to me to go largely unrecognized. E.g., it isn’t discussed (or at least I don't recognize it) in the recent plans written up ... |
9521805c-e872-47bf-b932-416d98d893fd | trentmkelly/LessWrong-43k | LessWrong | Mental Models Of People Can Be People
Warning: The argument in this post implies bad things about reality, and is likely to be bad for your mental health if it convinces you. If you're not sure if you can handle that, consider skipping this post for now and coming back later.
Introduction
At my best, when I used to ... |
4e97eb99-0619-4a29-a61e-a160c914ea01 | trentmkelly/LessWrong-43k | LessWrong | How to update P(x this week), upon hearing P(x next month) = 99.5%?
Suppose you want to assign a probability that a government will fall (ie the Prime Minister resigns) before the end of the year. Lacking any particular information - I haven't even told you which government it is - you say "Obviously, it's 50% - eithe... |
06fa87b3-68c7-438a-830e-9b24670cd73d | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Decision Theories: A Semi-Formal Analysis, Part II
### Or: Causal Decision Theory and Substitution
**Previously:**
0. [Decision Theories: A Less Wrong Primer](/lw/aq9/decision_theories_a_less_wrong_primer)
1. [The Problem with Naive Decision Theory](/lw/axl/decision_theories_a_semiformal_analysis_part_i)
**Sum... |
0b793f9e-ca1e-49bc-8703-317900dbfec6 | trentmkelly/LessWrong-43k | LessWrong | Tally of LessWrong experience on Alcohol
As a follow up post to: http://lesswrong.com/lw/m2r/lesswrong_experience_on_alcohol/
I tallied the responses.
In rought categories:
Doesnt drink: 11
Drinks: 19
Drinks heavily: 4
Disclaimers; I had to make judegements as to people who didn't like alcohol who drink v... |
665f9207-86a3-4a96-ae2b-362a19a85ad0 | trentmkelly/LessWrong-43k | LessWrong | Technical comparison of Deepseek, Novasky, S1, Helix, P0
Comparing Novasky with S1:
NovaSky by Berkeley club, S1 by Feifei Li (arXiv:2501.19393), are the players who don't have capital or compute, mainly focuses on developing method that finetune a large language model with curated minimal reasoning datasets. Sky-T1 t... |
a02ea5fb-762b-42cc-a09e-8843307f7655 | trentmkelly/LessWrong-43k | LessWrong | Rationality Quotes 23
"This year I resolve to lose weight... to be nicer to dogs... and to sprout wings and fly."
-- Garfield
"People understand instinctively that the best way for computer programs to communicate with each other is for each of them to be strict in what they emit, and liberal in what they ac... |
8d33845a-73ca-4ab1-b098-91bce1b9e778 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Stress Externalities More in AI Safety Pitches
It is important to figure out the best way(s) to convince people that AI safety is worth taking seriously because despite the fact that it is (in my opinion, and in the opinion of many people in the EA community) the most important cause area, it often seems weird to peop... |
d2866971-5ce0-4b52-af85-6f87c2195148 | trentmkelly/LessWrong-43k | LessWrong | Accelerate without humanity: Summary of Nick Land's philosophy
I took note of the philosopher Nick Land from reading about posthumanism on Wikipedia.
> A more pessimistic alternative to transhumanism in which humans will not be enhanced, but rather eventually replaced by artificial intelligences. Some philosophers, i... |
40147c02-c5a0-4033-881f-ae9da384d7d6 | trentmkelly/LessWrong-43k | LessWrong | New York Less Wrong: Expansion Plans
Followup To: New York City, Case Study of a Successful Rationalist Chapter; Epistle to the New York Less Wrongians
Last week, the New York Less Wrong group hosted the Summer Solstice Megameetup. Everyone had a huge amount of fun, and we met a lot of new people who didn't normally ... |
bbf74138-3e1f-4f80-b89e-58ebc61ccdb8 | trentmkelly/LessWrong-43k | LessWrong | Causality, Transformative AI and alignment - part I
TL;DR: transformative AI(TAI) plausibly requires causal models of the world. Thus, a component of AI safety is ensuring secure paths to generating these causal models. We think the lens of causal models might be undervalued within the current alignment research lands... |
67c85bf8-8472-4b92-a5ff-5e229357153f | trentmkelly/LessWrong-43k | LessWrong | (C)IRL is not solely a learning process
A putative new idea for AI control; index here.
I feel Inverse Reinforcement Learning (IRL) and Cooperative Inverse Reinforcement Learning (CIRL) are very good ideas, and will likely be essential for safe AI if we can't come up with some sort of sustainable low impact, modular,... |
d0c0fc7c-69bd-4c7e-9d33-f012650fe4e7 | trentmkelly/LessWrong-43k | LessWrong | [SEQ RERUN] Planning Fallacy
Today's post, Planning Fallacy was originally published on 17 September 2007. A summary (taken from the LW wiki):
> We tend to plan envisioning that everything will go as expected. Even assuming that such an estimate is accurate conditional on everything going as expected, things will ... |
0f658dcb-75d4-4349-af0a-ad6eda1c399b | trentmkelly/LessWrong-43k | LessWrong | Stagewise Development in Neural Networks
> TLDR: This post accompanies The Developmental Landscape of In-Context Learning by Jesse Hoogland, George Wang, Matthew Farrugia-Roberts, Liam Carroll, Susan Wei and Daniel Murfet (2024), which shows that in-context learning emerges in discrete, interpretable developmental sta... |
b002c50c-28d0-4e3a-83ea-838e58a53c7d | trentmkelly/LessWrong-43k | LessWrong | Write a business plan already
90% of start-ups go bust, often due to poor planning. Founders know they should write a business plan, yet many never do – spouting lame excuses based on misconceptions and cognitive biases (see footnote[1]).
Or else, they throw together a sketchy, useless plan. Or an over-optimistic one... |
efded2d3-ce23-4d43-a201-eb3ae949b3d1 | trentmkelly/LessWrong-43k | LessWrong | [SEQ RERUN] A Few Background Posts by Robin Hanson
I've held off on posting the next rerun for a few days in case anybody else suggested something new in the discussion of how to run the AI FOOM Debate. After looking at the comments (with associated votes), and after looking through the sequence myself, I've decided t... |
02a05bcd-4b56-40cc-a6fd-12c31beb5e01 | trentmkelly/LessWrong-43k | LessWrong | How my school gamed the stats
I was reading the Slate Star Codex review of The Cult of Smart. Part of it discusses charter vs state schools and the allegations of fraud of various kinds undermining charter schools record of better achievement. Reading it, I realized that I took for granted that public schools engage i... |
89a71dca-d1ac-4cc0-b68a-1a34641c1085 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Behavioral and mechanistic definitions (often confuse AI alignment discussions)
**TL;DR:***It’s important to distinguish between*behavioral *definitions – which categorize objects based on outside observable properties – and*mechanistic *definitions – which categorize objects based on their internal mechanisms.* *In t... |
36aa9776-ed9e-48ec-8dff-876c93f0abec | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Thoughts on Iason Gabriel’s Artificial Intelligence, Values, and Alignment
Iason Gabriel’s 2020 article [Artificial Intelligence, Values, and Alignment](https://arxiv.org/pdf/2001.09768.pdf) is a philosophical perspective on what the goal of alignment actually is, and how we might accomplish it. In the best spirit of ... |
6526fff2-d8f9-467e-8e37-174188143ccf | trentmkelly/LessWrong-43k | LessWrong | "Attitudes Toward Artificial General Intelligence: Results from American Adults 2021 and 2023" - call for reviewers (Seeds of Science)
Abstract
A compact, inexpensive repeated survey on American adults’ attitudes toward Artificial General Intelligence (AGI) revealed a stable ordering but changing magnitudes of agreeme... |
a7d21e24-e3e2-4dc5-86c2-9b2f6322e7e9 | trentmkelly/LessWrong-43k | LessWrong | The Stanley Parable: Making philosophy fun
Hello,
If you are unaware, The Stanley Parable is a video game originally released in 2013, and later re-released 2022 under the title The Stanley Parable: Ultra Deluxe. The player takes control of the titular character Stanley, as the present and future actions of his life ... |
fac04f44-155b-4d84-94ce-72f1eb24a854 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Pittsburgh Meetup: Big Gaming Fun!
Discussion article for the meetup : Pittsburgh Meetup: Big Gaming Fun!
WHEN: 25 October 2011 06:00:00PM (-0400)
WHERE: 1324 Wightman St., Pittsburgh, PA 15217
Welcome to the third Pittsburgh meetup of the semester, apparently!
I can't go anywhere next week because I'm on... |
2a3f2359-69ff-472c-86aa-40dce4a27f0e | StampyAI/alignment-research-dataset/arbital | Arbital | Cyclic Group Intro (Math 0)
##Clocks##
Say we had a clock. We decide that we can "add" two numbers on the clock starting at the first number, then moving by the number of steps given by the second number. For example the sum of $5$ and $6$ is given by starting at $5$, and then moving by $6$ spaces, and ending up at $1... |
a3630c9a-48a1-475c-93a1-a33037765d9b | trentmkelly/LessWrong-43k | LessWrong | Are there any reliable CAPTCHAs? Competition for CAPTCHA ideas that AIs can’t solve.
Background
As artificial intelligence continues to advance, it becomes increasingly important to find ways to verify the humanity of individuals online. One common method for doing so is through the use of CAPTCHAs, which are designe... |
f18ce828-26bf-428c-b942-eab4af517853 | trentmkelly/LessWrong-43k | LessWrong | [AN #155]: A Minecraft benchmark for algorithms that learn without reward functions
Alignment Newsletter is a weekly publication with recent content relevant to AI alignment around the world. Find all Alignment Newsletter resources here. In particular, you can look through this spreadsheet of all summaries that have e... |
8df37e61-eafe-4890-8ef1-22b40a0df9a6 | trentmkelly/LessWrong-43k | LessWrong | [SEQ RERUN] Inductive Bias
Today's post, Inductive Bias was originally published on April 8, 2007. A summary (from the LW wiki):
> Inductive bias is a systematic direction in belief revisions. The same observations could be evidence for or against a belief, depending on your prior. Inductive biases are more or less c... |
a84a69eb-2c34-4ded-96dd-9fa8a58a46b3 | trentmkelly/LessWrong-43k | LessWrong | Weekday Evening Beach Picnics
In the summer the kids don't need to be to bed so early and it stays light longer. When the weather's good we've been doing a lot of evening picnics at the beach. It's not too crowded (making parking easier), there's less direct sun (don't need to deal with sunscreen), and it doesn't conf... |
d9a20ca4-0298-4983-a9d0-69ee2cf3ae9f | trentmkelly/LessWrong-43k | LessWrong | When companies go over 150 people......
|
777e0b33-21ea-4f16-9ee3-2448db0d3d53 | trentmkelly/LessWrong-43k | LessWrong | A review of "Why Did Environmentalism Become Partisan?"
I was recently encouraged to read Jeffrey Heninger's report "Why Did Environmentalism Become Partisan?" It was interesting, but I thought it had some critical flaws. I would've recommended rejecting it if I were reviewing it for an academic conference.
I've ... |
6618d491-1c49-4e24-a811-ead933e0e8b0 | trentmkelly/LessWrong-43k | LessWrong | Control Theory Commentary
This is the third and final post in a sequence on control theory. In the first post I introduced the subject of control theory and stepped through some basics. In the second post I outlined Powers's model, as presented in Behavior: The Control of Perception. This post is a collection of comme... |
522b993b-9beb-4dd7-8eb7-d5a96af3b960 | trentmkelly/LessWrong-43k | LessWrong | Becoming stronger together
> I want people to go forth, but also to return. Or maybe even to go forth and stay simultaneously, because this is the Internet and we can get away with that sort of thing; I've learned some interesting things on Less Wrong, lately, and if continuing motivation over years is any sort of pr... |
e9a275c6-8637-409a-be6b-ca81cf7284ab | trentmkelly/LessWrong-43k | LessWrong | Safe questions to ask an Oracle?
The Future of Humanity Institute wants to pick the brains of the less wrongers :-)
Do you have suggestions for safe questions to ask an Oracle? Interpret the question as narrowly or broadly as you want; new or unusual ideas especially welcome. |
1a09de11-0922-4e28-bbf2-e3326a2355f1 | StampyAI/alignment-research-dataset/arxiv | Arxiv | Motivating the Rules of the Game for Adversarial Example Research
1 Introduction
---------------
Machine learning models for classification, regression, and decision making are becoming ubiquitous in everyday systems.
These models inform decisions about our health care and our finances, what products we buy, what m... |
bc5dae69-a10e-4dd9-9a3f-71d40da1be5d | trentmkelly/LessWrong-43k | LessWrong | Proposal for an Implementable Toy Model of Informed Oversight
|
ea4b2e06-5152-4e32-97ea-0664c22b2a60 | trentmkelly/LessWrong-43k | LessWrong | Gunpowder as metaphor for AI
The exothermic reaction of combusting gunpowder is a continuous but fast process. Sufficiently fast that a human will perceive it as discontinuous.
This is just refining a metaphor for a subject that's been hashed out many times before.
I'm basically just agreeing with JBlack's comment o... |
a035fe99-5372-4f24-a3f6-5e329b031456 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Scalable And Transferable Black-Box Jailbreaks For Language Models Via Persona Modulation
**Paper coauthors:** [Rusheb Shah](https://www.linkedin.com/in/ACoAAB63nTsBLsxbiGUom3-xG_DLLJvBLPnwn0M), [Quentin Feuillade--Montixi](https://www.linkedin.com/in/ACoAADL_UysBW9lkWi6Q_x8qEjF4k2fk9F2qlls), [Soroush J. Pour](https:/... |
3b4bc5fc-8280-4652-910b-07cc1e6802bf | trentmkelly/LessWrong-43k | LessWrong | The Dirt on Depression
(From the "humans are crazy" and "truth is stranger than fiction" departments...)
Want to be happy? Try eating dirt... or at least dirty plants.
Seriously.
From an article in Discover magazine, "Is Dirt The New Prozac?":
> The results so far suggest that simply inhaling M. vaccae—you get a ... |
160e4444-5745-4e30-9463-1f0ecfe885a1 | trentmkelly/LessWrong-43k | LessWrong | Ideas for heuristics and biases research topic?
Hey Less Wrong,
I'm currently taking a cognitive psychology class, and will be designing and conducting a research project in the field — and I'd like to do it on human judgment, specifically heuristics and biases. I'm currently doing preliminary research to come up wit... |
8a934818-3673-4c20-8294-de9b3367c9df | trentmkelly/LessWrong-43k | LessWrong | Corrigibility thoughts I: caring about multiple things
This is the first of three articles about limitations and challenges in the concept of corrigibility (see articles 2 and 3).
The desiderata for corrigibility are:
1. A corrigible agent tolerates, and preferably assists, its operators in their attempts to alter ... |
91d33f0a-c594-487a-9632-a05277c54704 | trentmkelly/LessWrong-43k | LessWrong | The Long-Term Future Fund is looking for a full-time fund chair
EDIT 2023/11/21
We have finished evaluating the first batch of our applications.
As we have not yet finished the hiring process, people are still welcome to apply to us. We are still very likely to skim future applications, however we have no firm commi... |
eabb2881-9298-4933-86d8-818008dd7d97 | StampyAI/alignment-research-dataset/youtube | Youtube Transcripts | Artificial Intelligence Safety and Security - Roman V. Yampolskiy, PhD
[Music]
we'll get started
so I'm Romanian Polsky I'm a computer
scientist and engineer at the University
of Louisville my main area of research
is safety and security of intelligent
systems and how many of you actually
attended talk keynote by Prof... |
a064331b-3052-4d15-a0e0-288103152a17 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Introducing EffiSciences’ AI Safety Unit
*This post was written by Léo Dana, Charbel-Raphaël Ségerie, and Florent Berthet, with the help of Siméon Campos, Quentin Didier, Jérémy Andréoletti, Anouk Hannot, Angélina Gentaz, and Tom David.*
In this post, you will learn what were EffiSciences’ most successful field-... |
8203b8d0-c226-4195-aee4-dff565bb4705 | trentmkelly/LessWrong-43k | LessWrong | Timothy Chu Origins Chapter 1
They remember his ideas, and how they saved the world.
But I remember the man, and how he saved me.
A great deal of ink (though in my estimation as yet too little) has by now been spilled about the life and times of Timothy Chu; among the more readable volumes on the legend are The Age o... |
32881e2e-5fe3-4af2-8b8e-b9857b1d2316 | trentmkelly/LessWrong-43k | LessWrong | Implications of evidential cooperation in large worlds
I've written several posts about the plausible implications of "evidential cooperation in large worlds" (ECL), on my newly-revived blog. This is a cross-post of the first. If you want to see the rest of the posts, you can either go to the blog or click through the... |
41756df7-ba48-4637-8e8b-67e852584a85 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Part 1: The AI Safety community has four main work groups, Strategy, Governance, Technical and Movement Building
Epistemic status
================
Written as a non-expert to develop and get feedback on my views, rather than persuade. It will probably be somewhat incomplete and inaccurate, but it should provoke helpfu... |
efa0834e-7f5f-4876-bc00-47d604ba08f1 | trentmkelly/LessWrong-43k | LessWrong | What colleges look for in extracurricular activities
[Edit: The post below gives the impression that our conversations with admissions officers are our only reasons for believing the claims. We've also consulted with other sources such as How to Be a High School Superstar: A Revolutionary Plan to Get into College by S... |
fd6988e7-4ac7-4f83-817f-6e90370ac842 | LDJnr/LessWrong-Amplify-Instruct | LessWrong | "There are many paradoxes with unbounded utility functions. For instance, consider whether it's rational to spend eternity in Hell:
Suppose that you die, and God offers you a deal. You can spend 1 day in Hell, and he will give you 2 days in Heaven, and then you will spend the rest of eternity in Purgatory (which is pos... |
74a9b5da-e85c-4e4c-840b-bd5ccc6762a9 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Why You Should Care About Goal-Directedness
**Introduction**
================
[Deconfusing](https://intelligence.org/2018/11/22/2018-update-our-new-research-directions/#section2) goal-directedness would boost your favorite research approach for solving AI Alignment.
Why? Because every approach I know of stands to ga... |
6d75e703-588c-49d6-9730-5666fee43504 | trentmkelly/LessWrong-43k | LessWrong | Exporting Hangouts History
Yesterday I received a worrying email:
> From: The Google Apps Team
> Subject: Important information about the upgrade from Google Hangouts to Google Chat
> Earlier this year, we announced that all Hangouts users would be upgraded to Google Chat. As the last step of the upgrade, starting ... |
c8849d79-6a7a-4f19-82fc-7a6094a9232f | trentmkelly/LessWrong-43k | LessWrong | Idea: Self-Improving Task Management Software
So what the world needs is yet another task management program, right?
My idea is software which automatically implements productivity strategies, measures the effectiveness of those strategies, and analyses which strategies work best for you. Hopefully, using the softwa... |
b283786e-5da3-44a1-b1d2-8039fa56a52e | trentmkelly/LessWrong-43k | LessWrong | [link]Mass replication of Psychology articles planed.
http://chronicle.com/blogs/percolator/is-psychology-about-to-come-undone/29045
The plan is to replicate or fail to replicate all 2008 articles from three major Psychology journals.
ETA: http://openscienceframework.org/ is the homepage of the group behind this. I... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.