id stringlengths 36 36 | source stringclasses 15
values | formatted_source stringclasses 13
values | text stringlengths 2 7.55M |
|---|---|---|---|
cfcbdb5f-4c6c-4e27-a004-42afb6e3f63e | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | The Parable of Predict-O-Matic
I've been thinking more about partial agency. I want to expand on some issues brought up in the comments to my [previous post](https://www.lesswrong.com/posts/4hdHto3uHejhY2F3Q/partial-agency), and on other complications which I've been thinking about. But for now, a more informal parabl... |
c51613a1-6a1c-44a6-a45e-e0d4d13b899a | StampyAI/alignment-research-dataset/blogs | Blogs | Linear algebra fact
Here is interesting linear algebra fact: let  be an  matrix and 
WHERE: 565-567 George Street, Sydney, Australia 2000
Hey all!
Just a reminder for the up-coming meet up at the Sydney RSL - level 3 at the restaurant.
Wed did not get... |
fae31c52-9008-4f2d-9111-f7b6c5a8cb2f | trentmkelly/LessWrong-43k | LessWrong | Metamorphosis
7. Metamorphosis
Summary of entire Series: An alternative approach to designing Friendly Artificial Intelligence computer systems.
Summary of this Article: Under what circumstances can an AI be trusted to negotiate in good faith an alteration to its core values?
Links to all the articles in the serie... |
0e689ea0-56b5-43a0-87a7-a00ea4c21e4b | trentmkelly/LessWrong-43k | LessWrong | [SEQ RERUN] Emulations Go Foom
Today's post, Emulations Go Foom was originally published on November 22, 2008. A summary:
> A description of what Robin Hanson thinks is the most likely scenario for a intelligence takeoff.
Discuss the post here (rather than in the comments to the original post).
This post is par... |
c5377c60-cf60-4a73-9b41-70a950514b35 | trentmkelly/LessWrong-43k | LessWrong | Philosophical Jailbreaks: Demo of LLM Nihilism
Epistemic Status: Exploratory
It was the end of August, 1991; I was leaning over a windowsill on the 4th floor of my school building in a town at Far East of USSR, looking outside on tree tops and the roof of adjacent building, and thinking about opening the window to es... |
9f148c44-434d-4e31-8f06-9f5687e41de9 | trentmkelly/LessWrong-43k | LessWrong | Human Biases (Ab)used by commerce [link]
Pretty nice article by Dan Ariely on cognitive biases benefiting certain commercial entities.
http://dl.dropbox.com/u/5506657/ArielyWired.pdf
I think he (rightly) does not claim these are all intentional, I suspect sufficient trial-and error + testing (used by many of the bus... |
d916197a-1a27-4736-afa1-57b959fceff9 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Confusion about neuroscience/cognitive science as a danger for AI Alignment
Recently, I wrote an article together with Jan Kirchner on ["brain enthusiasts" in AI Safety](https://www.lesswrong.com/posts/nfoYnASKHczH4G5pT/brain-enthusiasts-in-ai-safety) (if you find work on neuroscience/cognitive science x AI (Safety) i... |
7941063a-9430-494c-a46e-7a7552e0d912 | StampyAI/alignment-research-dataset/arbital | Arbital | Rationality
The subject domain for [epistemic](https://arbital.com/p/) and [instrumental](https://arbital.com/p/) rationality. |
bd126e48-da93-41d3-949d-12c2596d1dc9 | trentmkelly/LessWrong-43k | LessWrong | Contra Infinite Ethics, Utility Functions
Suppose you're a human, and you think we're onto something with modern physics. Humans have a brain that does not exceed a meter in radius. Humans have a brain that weighs less than 100kgs. By the Bekenstein bound, we know the average human brain could have at most 2.6*10^42 b... |
1cf116c9-1682-43b6-a674-3aa46c1ca214 | trentmkelly/LessWrong-43k | LessWrong | Privacy and Manipulation
Previously:
* “Can you keep that confidential? How do you know?”
* Parameters of Privacy
* Norm Innovation and Theory of Mind
My parents taught me the norm of keeping my promises.
My vague societal culture taught me a norm of automatically treat certain types of information as private.
... |
916850ec-4033-4b38-96af-1143e81988c6 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Reflection of Hierarchical Relationship via Nuanced Conditioning of Game Theory Approach for AI Development and Utilization
**Application of Game Theory to AI Development and Utilization**
A recent research post titled "Game Theory as an Engine for Large-Scale Data Analysis" by a Google team [(McWilliamson et al. 20... |
0ba8e9ad-6b40-4f67-84f3-4130ffe908cb | trentmkelly/LessWrong-43k | LessWrong | Thoughts on “AI is easy to control” by Pope & Belrose
Quintin Pope & Nora Belrose have a new “AI Optimists” website, along with a new essay “AI is easy to control”, arguing that the risk of human extinction due to future AI (“AI x-risk”[1]) is a mere 1% (“a tail risk worth considering, but not the dominant source of r... |
f5cd1ce3-9ab0-4397-bbc1-de103d2c0f9f | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Is it valuable to the field of AI Safety to have a neuroscience background?
First of all, I am currently studying *Philosophy, Neuroscience and Cognition* in my Bachelor's degree in Magdeburg, Germany. After being at EAxOxford last week and planning to go to EA London, I am thinking hard about my career planning proce... |
a366d6b0-19ed-4083-b70d-2ea2a990655b | StampyAI/alignment-research-dataset/arxiv | Arxiv | Few-Shot Goal Inference for Visuomotor Learning and Planning
###
1 Introduction
Reinforcement learning and planning methods assume some form of objective or reward function
that encodes the desired outcome or behavior.
There is a range of tasks where such an objective is challenging for humans to convey to robots,... |
29bd7834-c556-4376-aa11-64e22c9b44c3 | trentmkelly/LessWrong-43k | LessWrong | Birth of a Stereotype
I imagine that many ‘enlightened’ people disbelieve in stereotypes, thinking them useless, and something below them; only the ignorant masses rely on stereotypical thinking and generalisations. Stereotypes are improbable, and making such generalisations from anecdotal evidence at best is faulty. ... |
b7b47209-9036-47fc-9db9-55260f0adb20 | trentmkelly/LessWrong-43k | LessWrong | Quick thoughts on empathic metaethics
Years ago, I wrote an unfinished sequence of posts called "No-Nonsense Metaethics." My last post, Pluralistic Moral Reductionism, said I would next explore "empathic metaethics," but I never got around to writing those posts. Recently, I wrote a high-level summary of some initial ... |
340a2f40-883a-4c8f-abc1-84b952de64dd | trentmkelly/LessWrong-43k | LessWrong | The Colliding Exponentials of AI
Epistemic status: I have made many predictions for quantitative AI development this decade, these predictions were based on what I think is solid reasoning, and extrapolations from prior data.
If people do not intuitively understand the timescales of exponential functions, then mul... |
828ce97f-e2a4-49ea-b128-8fe93450599a | trentmkelly/LessWrong-43k | LessWrong | [Meta] anonymous merit or public status
Shower thought: How much are new posts evaluated based on authors’ old reputation?
We could find out if all posts were anonymous for a duration of x weeks and posts could only be upvoted during that time. [If it’s considered important that posts can be upvoted later as well, th... |
a103dd20-7ad3-426a-850f-443ab0438c26 | trentmkelly/LessWrong-43k | LessWrong | New book on s-risks
I have just published my new book on s-risks, titled Avoiding the Worst: How to Prevent a Moral Catastrophe. You can find it on Amazon, read the PDF version, or listen to the audio version.
The book is primarily aimed at longtermist effective altruists. I wrote it because I feel that s-risk prev... |
f3b149b0-6687-4863-929f-ffe39ce6ce6a | StampyAI/alignment-research-dataset/youtube | Youtube Transcripts | Yudkowsky vs Hanson — Singularity Debate
see with the festivities I forget what
the exact form of it was the question is
after all sorts of
logical things happen at some
undetermined point in the future are we
going to see a sort of very small
nucleus that the that can or does
control like all the resources or do we
s... |
103da556-d1e8-4d1f-8c76-c15959594fe9 | trentmkelly/LessWrong-43k | LessWrong | Against population ethics
The following is an excerpt from some comments I wrote to Will MacAskill about a pre-publication draft of What We Owe the Future. It is in response to the chapter on population ethics.
Chapter 8 presented some interesting ideas and did so clearly, I learned a lot from it.
That said, I could... |
e9ab590f-e874-401d-a084-48c79b50b5ce | trentmkelly/LessWrong-43k | LessWrong | Is the output of the softmax in a single transformer attention head usually winner-takes-all?
Using the notation from here: A Mathematical Framework for Transformer Circuits
The attention pattern for a single attention head is determined by A=softmax(xTWTQWKx), where softmax is computed for each row of xTWTQWKx.
Eac... |
abfb820b-1bf5-4d51-9071-6f07d09f53cc | trentmkelly/LessWrong-43k | LessWrong | What should we do about COVID in 2024?
Most of the discussion around COVID has stopped and nobody really wants to think about it. In the last months I heard of a few friends being infected by COVID and suffering from it.
What's the current state? How should we think about the risk of being infected and what should we... |
2d74fc06-16ad-450e-8dce-77fc49d181f8 | trentmkelly/LessWrong-43k | LessWrong | Deception and Jailbreak Sequence: 2. Iterative Refinement Stages of Jailbreaks in LLM
Content Warning: This blog post contains examples of harmful language generated by LLM.
Note: This blog post summarizes very preliminary results after a "weekend hackathon", inspired by the work by Ball et al. (2024).
Motivation a... |
6d250fc8-e547-44d5-860a-85f22a2a4450 | trentmkelly/LessWrong-43k | LessWrong | What do you mean by "wrong", in the moral sense?
I've been reading through the recent discussions on morality here, and I wanted to ask about how people are using the word "wrong". Something I've read here (I don't recall the specific post or comment) discussed how someone could be rationally wrong, say thinking 2+2=5... |
300d5750-39a6-486b-b7b1-2e02224caaf0 | trentmkelly/LessWrong-43k | LessWrong | GPT-X, DALL-E, and our Multimodal Future [video series]
Video series talking about how GPT-3, DALL-E, and Multimodal AI models of the future could change human creativity in foundational ways.
Total of 19 videos, so far is at #8. New Video is released every week day. |
b6592b75-ab58-4424-9f9f-676904861e87 | trentmkelly/LessWrong-43k | LessWrong | Scorable Functions: A Format for Algorithmic Forecasting
You can think of a scorable function as a "box" that can produce large clusters of complex forecasts. Credit to Dall-E 3.
Introduction
Imagine if a forecasting platform had estimates for things like:
1. "For every year until 2100, what will be the probabilit... |
c0258476-802f-42f2-87f7-e6154c103836 | trentmkelly/LessWrong-43k | LessWrong | Sustainability of Digital Life Form Societies
Hiroshi Yamakawa1,2,3,4
1 The University of Tokyo, Tokyo, Japan
2 AI Alignment Network, Tokyo, Japan
3 The Whole Brain Architecture Initiative, Tokyo, Japan
4 RIKEN, Tokyo, Japan
Even in a society composed of digital life forms (DLFs) with advanced autonomy, there is ... |
f148b3a9-6e81-4f98-ac9d-52aba5e3ecd8 | trentmkelly/LessWrong-43k | LessWrong | LW Australia's online hangout results, (short stories about cognitive biases)
In the Australia Mega-Online-hangout; a member mentioned a task/goal of his to write a few short stories to convey cognitive biases. After a while and a few more goals, someone suggested we actually write the short stories (the power of gro... |
d6262c4e-9f79-4362-a55a-de07f2522c69 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Productive Mistakes, Not Perfect Answers
*This post is part of the work done at* [*Conjecture*](https://conjecture.dev)*.*
I wouldn’t bet on any current alignment proposal. Yet I think that the field is making progress and abounds with interesting opportunities to do even more, giving us a shot. Isn’t there a contrad... |
c8de09d3-e038-41b7-97a0-4bbdabe2180e | trentmkelly/LessWrong-43k | LessWrong | Chapter 49: Prior Information
A boy waits at a small clearing at the edge of the non-forbidden forest, beside a dirt trail that runs back to the gates of Hogwarts in one direction, and off into the distance in another. There is a carriage nearby, and the boy is standing well away from it, looking at it, his eyes seldo... |
6f8a0673-fe0a-4d3b-9eac-eb55046be0ee | trentmkelly/LessWrong-43k | LessWrong | Meetup : Moscow, Chaotic Meet up
Discussion article for the meetup : Moscow, Chaotic Meet up
WHEN: 23 March 2014 04:00:00PM (+0400)
WHERE: Russia, Moscow, ulitsa L'va Tolstogo 16
We will have:
* Board games
* Non-board games
You can see proposed program in more details here (in Russian).
We gather in the Yand... |
22357b4d-724f-455b-a01d-fd4659b59a11 | trentmkelly/LessWrong-43k | LessWrong | Don't Believe Wrong Things
This is cross-posted from Putanumonit.com, you can jump in the discussion in either place.
----------------------------------------
LessWrong has a reputation for being a place where dry and earnest people write dry and earnest essays with titles like “Don’t Believe Wrong Things”. A casual... |
568cf708-7a9b-45e3-82cd-f58cc8fa3423 | LDJnr/LessWrong-Amplify-Instruct | LessWrong | "In early 2000, I registered my personal domain name weidai.com, along with a couple others, because I was worried that the small (sole-proprietor) ISP I was using would go out of business one day and break all the links on the web to the articles and software that I had published on my "home page" under its domain. Se... |
f2d69ff8-1a30-4a2c-963c-8414ac34335c | trentmkelly/LessWrong-43k | LessWrong | Welcome to SSC Zurich [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 does... |
a6a47bfd-a48e-46ed-81f1-94d99833c8d7 | StampyAI/alignment-research-dataset/blogs | Blogs | formalizing the QACI alignment formal-goal
*this work was done by [Tamsin Leake](https://carado.moe) and [Julia Persson](https://www.lesswrong.com/users/juliahp) at [Orthogonal](https://orxl.org).*
*thanks to [mesaoptimizer](https://mesaoptimizer.com/) for his help putting together this post.*
formalizing the QAC... |
d4d329e5-c353-4f9d-bcd4-ba845ee7954b | trentmkelly/LessWrong-43k | LessWrong | Free copy of Feynman's autobiography for best corny rationalist joke
I have an extra copy of Richard Feyman's autobiography, "Surely You're Joking, Mr. Feynman!": Aventures of a Curious Character, which I want to give away here.
This is one of two autobiographies (along with Ben Franklin's) to actually change my life... |
b8413586-60a5-45c4-95dd-6e9bb694ae61 | trentmkelly/LessWrong-43k | LessWrong | Labelling, Variables, and In-Context Learning in Llama2
Hi LessWrong! This is my first LessWrong post sharing my first piece of mechanistic interpretability work.
I studied in-context learning in Llama2. The idea was to look at when we associate two concepts in the LLM's context — an object (e.g. "red square"), and a... |
17b73669-d631-425a-9631-072b8dcd16a7 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Second Copenhagen meetup
Discussion article for the meetup : Second Copenhagen meetup
WHEN: 26 May 2012 05:00:00PM (+0200)
WHERE: Købmagergade 52, 1150 København K
There is a desire for a second meetup in Copenhagen! This time we are meeting in Studenterhuset on Saturday at 17:00, across from the observator... |
7c3b203e-49e3-4d9f-9df0-d5095b70b3c7 | trentmkelly/LessWrong-43k | LessWrong | Three characteristics: impermanence
This is the sixth post of the "a non-mystical explanation of the three characteristics of existence" series.
Impermanence
Like no-self and unsatisfactoriness, impermanence seems like a label for a broad cluster of related phenomena. A one-sentence description of it, phrased in exp... |
1368fa37-e518-4847-bcbb-4945b1838bea | LDJnr/LessWrong-Amplify-Instruct | LessWrong | "Public schools (and arguably private schools as well; I wouldn't know) teach students what to think, not how to think.
On LessWrong, this insight is so trivial not to bear repeating. Unfortunately, I think many people have adopted it as an immutable fact about the world that will be corrected post-Singularity, rather ... |
5515923c-a342-46af-b534-7f2552f41cf0 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | The Optimizer's Curse and How to Beat It
The best laid schemes of mice and men
Go often askew,
And leave us nothing but grief and pain,
For promised joy!
- [Robert Burns](http://en.wikipedia.org/wiki/To_a_Mouse) (translated)
Consider the following question:
>
> A team of decision analysts has just... |
eb56be8a-fde6-4093-be1c-e1aa490bf143 | trentmkelly/LessWrong-43k | LessWrong | Mech Interp Challenge: November - Deciphering the Cumulative Sum Model
I'm writing this post to discuss solutions to the October challenge, and present the challenge for this November.
If you've not read the first post in this sequence, I'd recommend starting there - it outlines the purpose behind these challenges, ... |
7812b343-b594-4b26-9d4c-36d85eeb35de | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Note on algorithms with multiple trained components
**Example 1:** consider a [GAN](https://en.wikipedia.org/wiki/Generative_adversarial_network). There’s a generator and a discriminator. As an intuitive mnemonic, we can say
* The “purpose” of the generator is to trick the discriminator,
* The “purpose” of the discri... |
f612a260-5146-444e-b930-c16741893804 | trentmkelly/LessWrong-43k | LessWrong | Malice, Stupidity, or Egalité Irréfléchie?
Anyone who has decided to strike off the mainstream path has experienced this: Strong admonitions and warnings against what they were doing, and pressures not do it.
It doesn’t really matter what it is you’re trying to change. If you’re trying to become a nondrinker in a dri... |
93e73719-d3f6-4f73-9674-b22bc0c5d758 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Evaluating Existing Approaches to AGI Alignment
My read of the AI safety space is that there are currently two major approaches to AGI alignment being researched: agent foundations and agent training. We can contrast them in part by saying the ultimate goal of the agent foundations program is to figure out how to make... |
d74a11a7-4c51-4b85-b3db-e5913d02b4d4 | trentmkelly/LessWrong-43k | LessWrong | Gortha Ick
Every year my family gets together at Christmas, and in the evenings we sing carols. Up on the Housetop always ends with this parody verse:
> Pa, Ma, and Grandpa, and Grandma sad,
> All I declare have something bad
> Even the baby hates its part
> Hugging a rattlesnake close to its heart
>
> Yo ho ho, w... |
df53fc48-0282-4cef-ab6e-690f8633b858 | trentmkelly/LessWrong-43k | LessWrong | June 2014 Media Thread
This is the monthly thread for posting media of various types that you've found that you enjoy. Post what you're reading, listening to, watching, and your opinion of it. Post recommendations to blogs. Post whatever media you feel like discussing! To see previous recommendations, check out the ol... |
ef11f6fe-ebcf-4e26-b1d4-0b8595bc311f | StampyAI/alignment-research-dataset/youtube | Youtube Transcripts | MLP Neurons - 40L Preliminary Investigation [rough early thoughts]
so
one of the directions that
uh our research and anthropic has
involved has been
um just trying to understand
what different neuron neurons and
different layers of transformers are
doing
um inside the mlp layers
and it's been something that we've
have... |
09927b88-c99e-43ab-8127-390f9aed375b | trentmkelly/LessWrong-43k | LessWrong | Taking control of your happiness and productivity
Here I talk about Julian Rotter's locus of control, and the implications of the research on our health, happiness and productivity.
As per my last article(s), feel free to let me know what you think here, privately, or anonymously.
Link |
de04cdc3-679c-4999-af46-619b7b6d3f94 | trentmkelly/LessWrong-43k | LessWrong | What is the best critique of AI existential risk arguments?
If you could link to an article or other piece of media, that would be ideal. Writing one up here is fine as well. An equivalent question would be "what is the best argument for the claim that there is a <1% probability of AI existential risk?" |
739d9e53-f01b-4149-a38f-e94beacd3e91 | trentmkelly/LessWrong-43k | LessWrong | Hammer and Mask -
Wide spread use of reusable particle filtering masks as a SARS-CoV-2 eradication strategy
by Marcel Müller (M.Sc. Biological Sciences)
Contact: marcel_mueller@mail.de
Epistemic Status: I am not a virologist and this is not medical advice. That said I am fairly sure about the core claims of this pie... |
2cf56132-a3e1-4524-8331-9dc95857d528 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | How do scaling laws work for fine-tuning?
The scaling laws, at least according to the interpretation [used in Ajeya's framework](https://www.lesswrong.com/posts/KrJfoZzpSDpnrv9va/draft-report-on-ai-timelines?commentId=7d4q79ntst6ryaxWD) (and this seems to be basically endorsed by tons of people I respect on this matte... |
b794486d-22b1-47bb-917b-2175bc7004b9 | trentmkelly/LessWrong-43k | LessWrong | RLHF
I’ve been thinking about Reinforcement Learning from Human Feedback (RLHF) a lot lately, mostly as a result of my AGISF capstone project attempting to use it to teach a language model to write better responses to Reddit writing prompts, a la Learning to summarize from human feedback.
RLHF has generated some impr... |
6b6fb466-c5d9-4448-88fb-2123f9a7a6e1 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Messy personal stuff that affected my cause prioritization (or: how I started to care about AI safety)
*Crossposted from* [*Otherwise*](https://juliawise.net/)
This is the story of how I started to care about AI risk. It’s far from an ideal decision-making process, but I wanted to try to spell out the untidy real... |
49f4dd6f-4b7d-4414-96c8-a35936a01ac7 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Tucson, Arizona
Discussion article for the meetup : Tucson, Arizona
WHEN: 20 June 2012 07:00:00PM (-0700)
WHERE: 2443 North Campbell Avenue
Hey Tucson, After the fun and inspiration of the Rationality Mini-Camp in Berkeley last month, I'm pretty jazzed about rationality, and I'd like to meet some more of m... |
d78d9d3b-7cfd-4926-bc5c-005dbfc43d37 | trentmkelly/LessWrong-43k | LessWrong | AI Governance Fundamentals - Curriculum and Application
[Edit: The course has received major updates since this post was made. See here for the most recent course information and curriculum. I've removed the largely outdated information from this post.]
While many people are interested in transformative AI governance... |
25698654-111d-42f2-8584-8d51d722670a | trentmkelly/LessWrong-43k | LessWrong | Insights from Munkres' Topology
This is about the Math Textbook Topology from Miri's research guide. (You can find the pdf online for free.) I got this book about a year ago. It takes a rigorous bottom-up approach that requires almost no prior knowledge but a lot of time. It's long and there are many exercises. I've r... |
ec148b32-b0a6-4bf5-b206-511ee27e4bfd | trentmkelly/LessWrong-43k | LessWrong | The Heilmeier Questions
Funding agencies like DARPA encourage you to consider a list of questions to guide the development of a good proposal. These are often called the Heilmeier questions or Heilmeier catechism. I often look back at this checklist when starting a new research project; they provide good inspiration, ... |
3848521a-0bae-46af-b126-083b4add863d | trentmkelly/LessWrong-43k | LessWrong | What more compute does for brain-like models: response to Rohin
This is a response to a comment made by Rohin Shah on Daniel Kokotajlo's post Fun with +12 OOMs of Compute. I started trying to answer some questions and assumptions he had, then realized there was more of an inferential gap that needed filling in. Also, ... |
02eb369b-3720-44d3-8959-fda72a5c30ae | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Learning Normativity: Language
Abram Demski [has been](https://www.lesswrong.com/posts/2JGu9yxiJkoGdQR4s/learning-normativity-a-research-agenda) writing about [Normativity](https://www.lesswrong.com/posts/tCex9F9YptGMpk2sT/normativity). The suggested models so far have mostly looked at actions rather than semantics, d... |
44e32c4a-b63f-4bf8-8a08-b3fcef49f5bd | trentmkelly/LessWrong-43k | LessWrong | What is Abstraction?
Let's start with a few examples (borrowed from here) to illustrate what we're talking about:
* We have a gas consisting of some huge number of particles. We throw away information about the particles themselves, instead keeping just a few summary statistics: average energy, number of particles, ... |
9ce9418f-7189-40d5-9744-6b77dfce58e8 | trentmkelly/LessWrong-43k | LessWrong | Linkpost: Francesca v Harvard
Just when I thought the Gino case came to a close, I started reading Francesca Gino's defense. It paints her as a belabored academic unfairly penalized by misleading and incompetent investigators, as well as a biased New Yorker reporter.
I'm reposting this here as I think it's useful to ... |
5364e7e6-f952-4739-9467-30e1e139b0f0 | trentmkelly/LessWrong-43k | LessWrong | Discussion of concrete near-to-middle term trends in AI
Instead of prognosticating on AGI/Strong AI/Singularities, I'd like to discuss more concrete advancements to expect in the near-term in AI. I invite those who have an interest in AI to discuss predictions or interesting trends they've observed.
This discussion s... |
8e8753ee-5546-4e2f-9147-f94daefad135 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Confucianism in AI Alignment
*I hear there’s a thing where people write a lot in November, so I’m going to try writing a blog post every day. Disclaimer: this post is less polished than my median. And my median post isn’t very polished to begin with.*
Imagine a large corporation - we’ll call it BigCo. BigCo knows tha... |
222f9e17-aa1f-4dd2-a63f-96e9c5aad547 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | The flaws that make today's AI architecture unsafe and a new approach that could fix it
*This is a linkpost for* [*"The flaws that make today's AI architecture unsafe and a new approach that could fix it"*](https://80000hours.org/podcast/episodes/stuart-russell-human-compatible-ai/)*. You can listen to the episode on ... |
144d8510-25a0-4212-b0a8-c344731ec458 | trentmkelly/LessWrong-43k | LessWrong | The Power to Draw Better
This is Part X of the Specificity Sequence
Cats notoriously get stuck in trees because their claws are better at climbing up than down. Throughout this sequence, we’ve seen how humans are similar: We get stuck in high-level abstractions because our brains struggle to unpack them into specific... |
878e56f4-0127-4a24-88d9-89d560f43ec8 | trentmkelly/LessWrong-43k | LessWrong | Reference Post: Trivial Decision Theory Problem
A trivial decision problem is one where there is only a single option that the agent can take. In that case, the most natural answer to the answer to the question, "What action should we take?" would be "The only action that we can take!". We will call this the Trivialit... |
a8cfb1ca-d4d2-462a-94c3-179137f9c837 | trentmkelly/LessWrong-43k | LessWrong | What are the Nth Order Effects of the Coronavirus?
Some were easily foreseeable - like travel bans, lockdowns, and economic damage.
Some were made total sense in retrospect but were not an obvious think to think about - Like an increase in domestic violence (and, consequentially, shelters being overwhelmed).
Let's m... |
c9a522da-89be-43de-9479-a9b5d0d86533 | trentmkelly/LessWrong-43k | LessWrong | Proof of posteriority: a defense against AI-generated misinformation
Summary
* A proof of priority shows that a piece of data must have been created before a certain time.
* A proof of posteriority shows that a piece of data must have been created after a certain time.
* By combining the two, you can prove that the... |
f36d1c15-1f1a-4499-869c-5bf7a640ae1b | trentmkelly/LessWrong-43k | LessWrong | Replace yourself before you stop organizing your community.
Alternate title: Always be working to replace yourself in whatever capacity people rely on you.
[CN: This post has some competing access needs, and I made a tradeoff in a particular direction. I intend this as an earnest, important suggestion, not a demand. ... |
51da74e3-063f-4d32-bc57-b729e5be2090 | trentmkelly/LessWrong-43k | LessWrong | July 2011 review of experimental philosophy
Experimental philosophy is the mainstream academic field that is most directly attempting to dissolve (where possible) persistent philosophical problems by revealing the cognitive algorithms that generate old philosophical debates. Those who want to catch up with some of wha... |
33764171-de59-4867-9b2b-adab16603b23 | trentmkelly/LessWrong-43k | LessWrong | Book Review: On the Edge: The Fundamentals
The most likely person to write On the Edge was Nate Silver.
Grok thinks the next most likely was Michael Lewis, followed by a number of other writers of popular books regarding people thinking different.
I see why Grok would say that, but it is wrong.
The next most likely... |
ea764c69-ae65-46a5-8052-4324fc5d3c67 | trentmkelly/LessWrong-43k | LessWrong | [LINKPOST] Agents Need Not Know Their Purpose
https://arxiv.org/abs/2402.09734
> Ensuring artificial intelligence behaves in such a way that is aligned with human values is commonly referred to as the alignment challenge. Prior work has shown that rational agents, behaving in such a way that maximizes a utility funct... |
89285c57-f4f5-4c63-8cb2-417f21356a3f | trentmkelly/LessWrong-43k | LessWrong | Explanations of mathematical explanation
I recently read Mathematical Explanation [gated], by Mark Steiner (1978). My summary follows, and my commentary follows that. I am aware that others have written things since 1978 on this topic, but I don’t have time to read them right now.
***
We seem to think there is a dis... |
a1e7bc10-224c-4163-9656-e01c7e5eff8a | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | [AN #121]: Forecasting transformative AI timelines using biological anchors
Alignment Newsletter is a weekly publication with recent content relevant to AI alignment around the world. Find all Alignment Newsletter **[resources here](http://rohinshah.com/alignment-newsletter/)**. In particular, you can look through **[... |
be1bf5f3-ac4f-491c-9118-2684c2293b3b | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Partial Transcript of Recent Senate Hearing Discussing AI X-Risk
On Tuesday, the [US Senate Judiciary Subcommittee on Privacy, Technology and the Law](https://en.wikipedia.org/wiki/United_States_Senate_Judiciary_Subcommittee_on_Privacy,_Technology_and_the_Law) held a hearing on AI. The hearing involved 3 witnesses – D... |
629fe586-f7fd-449d-9043-a3372ad3f18f | trentmkelly/LessWrong-43k | LessWrong | Imitation Learning from Language Feedback
TL;DR: Specifying the intended behavior of language models is hard, and current methods, such as RLHF, only incorporate low-resolution (binary) feedback information. To address this issue, we introduce Imitation learning from Language Feedback (ILF), an iterative algorithm lev... |
24d019b9-098e-43ac-a93b-126f25e161b4 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Rochester LessWrong and Transhumanists
Discussion article for the meetup : Rochester LessWrong and Transhumanists
WHEN: 12 September 2015 01:00:00PM (-0400)
WHERE: 3333 West Henrietta Road, Rochester, NY 14623
Location: Jitters Cafe
I'm starting a meetup group in Rochester so some of us can hang out in me... |
fb9b99a7-1704-4323-a829-effa470ffd01 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | The conceptual Doppelgänger problem
*[Metadata: crossposted from <https://tsvibt.blogspot.com/2022/10/the-conceptual-doppleganger-problem.html>. First completed 9 October 2022.]*
Suppose we want to observe the thoughts of a mind in order to detect whether it's making its way towards a plan to harm us, and ideally al... |
8c1e7bdb-c089-48b3-bf18-e8d0ed7ce8dd | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Real-world examples of money-pumping?
Intransitive preferences are a [demonstrable](http://en.wikipedia.org/wiki/Dynamic_inconsistency) [characteristic](http://en.wikipedia.org/wiki/Allais_paradox) of human behaviour. So why am I having such trouble coming up with real-world examples of money-pumping?
"Because I'm ... |
687a16b6-c791-4a87-b535-22388f7495e2 | trentmkelly/LessWrong-43k | LessWrong | The non-indifferent behaviour of stratified indifference?
A putative new idea for AI control; index here.
This post aims to show some of the odd behaviour of stratified indifference. It seems that stratified indifference does not accomplish what I intended it to do, at least in certain situations.
Assume there are t... |
a46f774d-def5-4f4b-b10f-174ee2210382 | StampyAI/alignment-research-dataset/blogs | Blogs | CYOAs and futurism
CYOAs and futurism
------------------
there is a community for the creation and playing of games where one is to select a bunch of options, some favorable and some disfavorable, and imagine what an ensuing situation would look like. these CYOAs — named after, but not quite the same as, ["choose yo... |
22b87c46-6ba8-48f3-8099-43788bdbe565 | trentmkelly/LessWrong-43k | LessWrong | Fundamental Uncertainty: Chapter 7 - Why is truth useful?
N.B. This is a chapter in a book about truth and knowledge. It is the first draft. I have since revised it. You can find the most up-to-date info/version on the book's website.
Last chapter we proved that our knowledge of the truth is fundamentally uncertainty... |
c3751eed-f7cf-4fca-be86-0d9a5ade203e | trentmkelly/LessWrong-43k | LessWrong | Reposting previously linked content on LW
This question might be a bit specific to me, but maybe it applies to others so I'll ask publicly so the answer becomes more visible to all.
Is there are a policy or what are your thoughts on posting content on LW that was previously the subject of a link post?
In my case mos... |
12c241ae-f0fd-43d9-a236-a87b52413d20 | StampyAI/alignment-research-dataset/aisafety.info | AI Safety Info | What are some AI alignment research agendas currently being pursued?
Research at the [Alignment Research Center](http://alignmentresearchcenter.org/) is led by [Paul Christiano](https://paulfchristiano.com/), best known for introducing the [“Iterated Distillation and Amplification”](https://ai-alignment.com/iterated-d... |
e55e25d1-4616-4bf1-b0ba-149bb4cd89ef | trentmkelly/LessWrong-43k | LessWrong | Image: Another uninformed perspective on risks from AI (humor)
Here is another example of an outsider perspective on risks from AI. I think such examples can serve as a way to fathom the inferential distance between the SIAI and its target audience as to consequently fine tune their material and general approach.
v... |
b76ed4d3-9264-4798-97ad-ad113c36e477 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | A framework and open questions for game theoretic shard modeling
Background
==========
[The Shard Theory of Human Values](https://www.lesswrong.com/s/nyEFg3AuJpdAozmoX) is a new research agenda which I think has promise in serving as a model of value formation in humans and self-supervised ML agents. In brief, it cla... |
9fa5ba6c-7690-4127-bb84-2c4bd39a826d | trentmkelly/LessWrong-43k | LessWrong | Networks of Trust vs Markets
Markets are trust reducers. When functional, they make sure that you mainly need to trust one thing: that others will be self-interested. And since, as Adam Smith pointed out, you can trust even butchers and bakers to be self-interested – markets greatly increase the number of people you c... |
29753383-551c-4526-a499-e4782d89e4d5 | trentmkelly/LessWrong-43k | LessWrong | [SEQ RERUN] Cascades, Cycles, Insight...
Today's post, Cascades, Cycles, Insight... was originally published on 24 November 2008. A summary (taken from the LW wiki):
> Cascades, cycles, and insight are three ways in which the development of intelligence appears discontinuous. Cascades are when one development make... |
f32a0842-0e08-4b67-95ca-55e548995990 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | continue working on hard alignment! don't give up!
let's call "hard alignment" the (["orthodox"](https://scottaaronson.blog/?p=6821)) problem, historically worked on by MIRI, of preventing [strong agentic AIs](https://www.lesswrong.com/posts/AdGo5BRCzzsdDGM6H/contra-strong-coherence?commentId=Ph9EZRjj9CxqfcuSm) from p... |
bb6f8c52-1bb6-4b89-bf55-2d2a064016d2 | trentmkelly/LessWrong-43k | LessWrong | The two conceptions of Active Inference: an intelligence architecture and a theory of agency
I think much of the confusion about Active Inference arises because the term is used to refer to two related, yet distinct concepts (one can also call them abstractions, theories, or ontics): an intelligence architecture and a... |
083cbbf7-b468-44de-87ed-a4ad18b749d6 | trentmkelly/LessWrong-43k | LessWrong | Remote AI Alignment Overhang?
Background link: Do any AI alignment orgs hire remotely?
Disclaimer: The generator of me writing this post is my own desire for remote AI Alignment opportunities to exist. I have Alignment research in mind while writing this, but similar arguments might apply to other Alignment-related... |
71b19064-4ac8-4ea8-a527-28226f452a33 | StampyAI/alignment-research-dataset/blogs | Blogs | 2013 Winter Matching Challenge
Thanks to [Peter Thiel](http://en.wikipedia.org/wiki/Peter_Thiel), every donation made to MIRI between now and January 15th, 2014 will be **matched dollar-for-dollar**!
Also, **gifts from “new large dono... |
0bcccfd7-e4ac-416b-a612-4ae2d03ab9ee | trentmkelly/LessWrong-43k | LessWrong | GovAI: Towards best practices in AGI safety and governance: A survey of expert opinion
> A number of leading AI companies, including OpenAI, Google DeepMind, and Anthropic, have the stated goal of building artificial general intelligence (AGI) - AI systems that achieve or exceed human performance across a wide range o... |
a6e1e01a-1fce-4cd9-aa4f-471198722f0e | trentmkelly/LessWrong-43k | LessWrong | More variations on pseudo-alignment
In "Risks from Learned Optimization," we talked about a variety of different forms of pseudo-alignment—that is, ways in which a trained model's objective (its mesa-objective) can be misaligned off-distribution with the loss function it was trained under (the base objective). In part... |
962537c6-02a9-4727-90bf-4c4114031b65 | trentmkelly/LessWrong-43k | LessWrong | Barack Obama on Artificial Intelligence, Autonomous Cars, and the Future of Humanity
|
4e929ad5-7b3d-467b-9baf-8cd4a0bca250 | trentmkelly/LessWrong-43k | LessWrong | Can GPT-4 play 20 questions against another instance of itself?
Seems like a neat way to check understanding / world-model. Not my idea, saw this post: https://evanthebouncy.medium.com/llm-self-play-on-20-questions-dee7a8c63377 |
91bf9969-dac6-4426-baa9-355a66a91ff1 | trentmkelly/LessWrong-43k | LessWrong | AI #10: Code Interpreter and Geoff Hinton
The big capabilities news this week is a new ChatGPT mode (that I do not have access to yet) called Code Interpreter. It lets you upload giant data files, analyzes them automatically, can even write papers about its findings, many are impressed.
The big discourse news is that... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.