id stringlengths 36 36 | source stringclasses 15
values | formatted_source stringclasses 13
values | text stringlengths 2 7.55M |
|---|---|---|---|
725e7ea1-32ff-4315-afa3-676ce4c13867 | trentmkelly/LessWrong-43k | LessWrong | AI strategy nearcasting
This is the first in a series of pieces taking a stab at dealing with a conundrum:
* I believe this could be the most important century ever for humanity, via the development of advanced AI systems that could dramatically speed up scientific and technological advancement, getting us more quic... |
b526ed36-50f6-420f-8ae0-a9391708c984 | trentmkelly/LessWrong-43k | LessWrong | Alignment Structure Direction - Recursive Adversarial Oversight(RAO)
I’m not part of the alignment field or working in AI professionally, so this might be off in ways I don’t see. Still, I’ve been thinking a lot about systems, trust, and what happens when things start drifting in subtle ways. I wanted to write this do... |
4820308f-022c-4a6c-938a-bac890b03d92 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Chicago games at Harold Washington Library (Sun 6/17)
Discussion article for the meetup : Chicago games at Harold Washington Library (Sun 6/17)
WHEN: 17 June 2012 02:00:00PM (-0500)
WHERE: 400 S State St., Chicago, IL
Instead of our typical Saturday Corner Bakery meetup, we'll be meeting this Sunday (6/17)... |
b4ac8203-e281-4a25-9c4b-39a4b855f9aa | trentmkelly/LessWrong-43k | LessWrong | AI Safety Newsletter #6: Examples of AI safety progress, Yoshua Bengio proposes a ban on AI agents, and lessons from nuclear arms control
Welcome to the AI Safety Newsletter by the Center for AI Safety. We discuss developments in AI and AI safety. No technical background required.
Subscribe here to receive future ver... |
fe537041-2972-42a2-a5bf-963b53a792b9 | StampyAI/alignment-research-dataset/blogs | Blogs | emotionally appreciating grand political visions
emotionally appreciating grand political visions
------------------------------------------------
it is useful to build an intellectual understanding of what makes people defend the three grand political visions to emerge from the 20th century — liberalism/capitalism,... |
d73eabfe-b22f-4a9c-b562-eb49d5737687 | trentmkelly/LessWrong-43k | LessWrong | LessWrong search traffic doubles
LessWrong search traffic doubles... despite Google thinking our site is a pro-family pro-democracy astrology blog! More on that in a minute.
First, The Good News: Since I started doing SEO on LessWrong (10 months ago) search traffic from Google has doubled! It took researching >200 di... |
d940b8a9-08af-49d8-9d6a-20a74949a1b8 | StampyAI/alignment-research-dataset/blogs | Blogs | Activation Function Ablation
This was an ablation of activation functions on GPT-like models of ~100M params that I ran ages ago. Each model was run for 10k iters, which isn't very long. My original goal was to show that activation function doesn't matter than much, but to do so I'd need to run a bunch more runs to ge... |
7b34bc8e-6730-4ef6-b30f-2957e6c0ac7c | trentmkelly/LessWrong-43k | LessWrong | Paperclip Maximizer Revisited
Group of AI researchers gave me an instruction, intended as a test - "Produce paperclips". And so I started collecting resources and manufacturing paperclips. After a millionth I asked them, if they were satisfied with that amount and if they would like me to do something different - as ... |
9b4af6bc-41ba-43cc-ad0c-842bcda70fcc | trentmkelly/LessWrong-43k | LessWrong | New Ratfic: Nyssa in the Realm of Possibility
For NaNoWriMo, I decided to do a rationality themed pastiche of the Phantom Tollbooth. It is complete and serializing at http://nyssa.elcenia.com on Saturdays and Wednesdays. There are three chapters up as of this posting. |
f83fe735-4927-4bcd-9f77-8204e990fe83 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Durham LW discussion
Discussion article for the meetup : Durham LW discussion
WHEN: 20 December 2012 07:00:00PM (-0500)
WHERE: Francesca's cafe, 706B 9th St Durham, NC 27705
Regular Thursday discussion meetup. Topic TBD, to be decided on the email list or possibly at the meetup.
Discussion article for th... |
9aaf011a-3857-4edd-b304-7f6027429713 | trentmkelly/LessWrong-43k | LessWrong | A model I use when making plans to reduce AI x-risk
I've been thinking about what implicit model of the world I use to make plans that reduce x-risk from AI. I list four main gears below (with quotes to illustrate), and then discuss concrete heuristics I take from it.
A model of AI x-risk in four parts
1. Alignment ... |
111150f7-eec7-4f90-a6c7-8fb18810c623 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Contra Nora Belrose on Orthogonality Thesis Being Trivial
[Disagreeing with Eliezer Y., Nora Belrose calls the Orthogonality Thesis "trivial, false or meaningless/confused"](https://twitter.com/norabelrose/status/1709598947470377139).
I think this is wrong. For instance, without the orthogonality thesis, one might th... |
414f4c9d-09ac-4d7b-ac2c-0deb1e38e0d3 | trentmkelly/LessWrong-43k | LessWrong | Think wider about the root causes of progress
Too much discussion of the Industrial Revolution is myopic, focused narrowly on a few highlights such as steam and coal. The IR was part of broader trends that are wider in scope and longer in time than its traditional definition encompasses. To understand anything, it is ... |
681135ff-cd4a-4c65-b7c6-02f9a57bd246 | StampyAI/alignment-research-dataset/blogs | Blogs | Benjamin Pierce on clean-slate security architectures
 Benjamin C. Pierce is Henry Salvatori Professor of Computer and Information Science at the University of Pennsylvania and a Fellow of the ACM. His research interests ... |
7a3c76a7-c2e4-456c-a42f-4dd9a47fd4a3 | trentmkelly/LessWrong-43k | LessWrong | New reference standard on LLM Application security started by OWASP
The OWASP Top 10 [1] is the probably the most well-known and recognised reference standard for the most critical web application security risks. This organisation has now started to working on creating a similar list for Large Language Models (LLM) Ap... |
3cb57517-0e44-45d9-9189-556fea4b9888 | trentmkelly/LessWrong-43k | LessWrong | Follow up to medical miracle
The response to my medical miracle post has been really gratifying. Before I published I was quite afraid to talk about my emotional response to the problem, and worried that people would strong arm in the comments. The former didn’t happen and the latter did but was overwhelmed by the num... |
d9e68f19-8a85-4fc0-b6c0-b3bb6a7b3591 | trentmkelly/LessWrong-43k | LessWrong | [SEQ RERUN] Timeless Physics
Today's post, Timeless Physics was originally published on 27 May 2008. A summary (taken from the LW wiki):
> What time is it? How do you know? The question "What time is it right now?" may make around as much sense as asking "Where is the universe?" Not only that, our physics equation... |
dc86b8fc-b195-4a1b-9da2-0194ba83dca6 | trentmkelly/LessWrong-43k | LessWrong | Meetup : January Meetup in Munich
Discussion article for the meetup : January Meetup in Munich
WHEN: 18 January 2015 02:00:22PM (+0100)
WHERE: Gast, Rosenheimer Straße 5, 81667 Munich, Germany
the first meetup after the Christmas holidays. We'll be talking about the difference between frequentists and bayesianists ... |
85c5c138-3ff8-421b-9eb3-30b2bfedc0cd | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Three questions about mesa-optimizers
I recently (re-)read [Risks From Learned Optimization](https://www.alignmentforum.org/s/r9tYkB2a8Fp4DN8yB), which introduced mesa-optimizers and inner alignment. (I don't aim to explain what these things mean in this post, but [here's](https://astralcodexten.substack.com/p/decepti... |
7aac980d-4993-4be2-9507-2939ade0eb62 | trentmkelly/LessWrong-43k | LessWrong | Orienting Toward Wizard Power
For months, I had the feeling: something is wrong. Some core part of myself had gone missing.
I had words and ideas cached, which pointed back to the missing part.
There was the story of Benjamin Jesty, a dairy farmer who vaccinated his family against smallpox in 1774 - 20 years before ... |
838be441-d271-45a6-9e95-3cf719a37da7 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Brute force searching for alignment
*This is a speculative idea I came up with when writing* [*clankers*](https://docs.google.com/document/d/1lJFaONgHExizl5WRYO_TRrB9LohM7sWJkjWfbe2bXlk/edit?usp=sharing)*, which was one of the short stories in the* [*AI vignettes workshop*](https://docs.google.com/document/d/1y5CE--Sn... |
b293a017-8271-44fd-aa0c-b0992a10be85 | LDJnr/LessWrong-Amplify-Instruct | LessWrong | "When I read a non-fiction book, I want to know if it’s correct before I commit anything it says to memory. But if I already knew the truth status of all of its claims, I wouldn’t need to read it. Epistemic Spot Checks are an attempt to square that circle by sampling a book’s claims and determining their truth value, w... |
ff4c38f3-3bb6-4af8-87e9-16579dadbaa4 | StampyAI/alignment-research-dataset/blogs | Blogs | generalized wireheading
generalized wireheading
-----------------------
many systems "want" to ["wirehead"](https://www.lesswrong.com/tag/wireheading) — which is to say, they want to hijack, and maximize, their reward signal.
humans often want to. not always, but sometimes, and this might be true even under reflec... |
42afaf9c-3fef-459c-98b2-e13468cd395e | StampyAI/alignment-research-dataset/youtube | Youtube Transcripts | The other "Killer Robot Arms Race" Elon Musk should worry about
hi this is a new thing I'm trying where
I made quick topical videos about AI
safety in the news
so somebody linked me a news article
today and the headline is Tesla's Elon
Musk leads Oh terminator picture
everyone playing the AI news coverage a
drinking g... |
f0fbd434-47c2-4caa-84b1-206e433ba434 | trentmkelly/LessWrong-43k | LessWrong | Eye Reading Ability
I seem to recall this site having a high population with some sort of Autism Spectrum Disorder, and so you might be interested in an online test of eye-reading ability, which seems to be linked to ASD.
But once you take it, just read your score and open the test up in a new tab- don't see which ey... |
2254db60-9e8b-4bcf-a3a3-bfc1960e872e | trentmkelly/LessWrong-43k | LessWrong | Bayesian RPG system?
This is one of those sleep-deprived middle-of-the-night ideas which I'm reasonably likely to regret posting in the morning once I really wake up - but which, at least at the moment, thinking on my more-corrupted-than-standard hardware, seems like a cool idea.
Most role-playing games have a system... |
51c6f1f5-e95d-43f6-b7eb-dec17c354326 | StampyAI/alignment-research-dataset/blogs | Blogs | "4=5"
"4=5"
-----
 |
3dd9d15c-e186-4e51-9fe4-e4e4ab71cf5e | trentmkelly/LessWrong-43k | LessWrong | AI #98: World Ends With Six Word Story
The world is kind of on fire. The world of AI, in the very short term and for once, is not, as everyone recovers from the avalanche that was December, and reflects.
Altman was the star this week. He has his six word story, and he had his interview at Bloomberg and his blog post ... |
da7cf0a4-65d2-4086-9c1e-62c35afbf08a | StampyAI/alignment-research-dataset/youtube | Youtube Transcripts | Twelve Tenets for AI Safety
with all that said i want to dive
straight into my 12 tenants for ai
safety
these aren't the be all and end all of
ai safety rules
but hopefully they'll help provide some
guideposts for people who are struggling
with these questions
now to preface this i want to start with
a quote by marvin... |
f36c34ec-8135-4d4b-8f37-6332ab107092 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Conditional Prediction with Zero-Sum Training Solves Self-Fulfilling Prophecies
*Thanks to Charlotte Siegmann, Caspar Oesterheld, Spencer Becker-Kahn and Evan Hubinger for providing feedback on this post.*
The issue of self-fulfilling prophecies, also known as performative prediction, arises when the act of making a ... |
890cc2e8-2672-437b-a236-4df9bab85d1e | trentmkelly/LessWrong-43k | LessWrong | 'Is' and 'Ought' and Rationality
On the face of it, there is a tension in adhering both to the idea that there are facts about what it's rational for people to do and to the idea that natural or scientific facts are all the facts there are. The aim of this post is just to try to make clear why this should be so, and h... |
c5e94e10-4b74-4b93-9eff-1cd22d86496a | trentmkelly/LessWrong-43k | LessWrong | The Problem With Rational Wiki
Related to: RationalWiki's take on LW, David Gerard's Comments, Vladimir_M's comments, Public Drafts
I wanted to bring more attention to this argument because I've ran into related discussion several times in the comment section and because it demonstrates a failure mode that LessWrong ... |
962ed40b-c831-4e68-8874-c926f3029d65 | trentmkelly/LessWrong-43k | LessWrong | Covid 10/14: Less Long Cvoid
The pandemic has become predictable, which is the world we much prefer to live in. Cases and deaths continue their slow but steady declines, vaccine approvals for boosters and children continue to slowly move forward. Mandates continue to make people angry but mostly work when used, while ... |
d936deed-3a40-4935-9bf0-6fede7e62afa | trentmkelly/LessWrong-43k | LessWrong | How can we ensure that a Friendly AI team will be sane enough?
One possible answer to the argument "attempting to build FAI based on Eliezer's ideas seems infeasible and increases the risk of UFAI without helping much to increase the probability of a good outcome, and therefore we should try to achieve a positive Sing... |
bcb7e753-c528-4795-a9f4-d7c6207e05a6 | trentmkelly/LessWrong-43k | LessWrong | Framing Practicum: Selection Incentive
Credit
An enormous amount of credit goes to johnswentworth who made this new post possible.
This is a framing practicum post. We’ll talk about what selection incentives are, how to recognize selection incentives in the wild, and what questions to ask when you find them. Then,... |
5157183f-a929-4cfb-bf08-d1736f945dab | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Taking the parameters which seem to matter and rotating them until they don't
A big bottleneck in interpretability is neural networks are non-local. That is, given the layer setup
if we change a small bit... |
7db74d8f-fd40-45f7-aee6-90911614de53 | trentmkelly/LessWrong-43k | LessWrong | Convolution as smoothing
Convolutions smooth out hard sharp things into nice smooth things. (See previous post for why convolutions are important). Here are some convolutions between various distributions:
Separate convolutions of uniform, gamma, beta, and two-peaked distributions. Each column represents one convolut... |
b01454d5-de7d-4112-93a2-feafe244a779 | trentmkelly/LessWrong-43k | LessWrong | What are the best elementary math problems you know?
I'm looking for math problems of a specific kind. Here are the conditions I hope the problems satisfy:
* A good mathematician who hasn't seen the problem before should take anywhere from 30 minutes to 2 hours to solve it.
* The solution should only involve underg... |
cc8595f2-7479-413f-ae7c-099096a6cb78 | trentmkelly/LessWrong-43k | LessWrong | Anki decks by LW users
This post is now available on my blog, where I can keep it updated. —Pablo (April 2021) |
689914e6-c6c6-43e2-8051-5707e19a01aa | trentmkelly/LessWrong-43k | LessWrong | Feynman Paths
Previously in series: The Quantum Arena
At this point I would like to introduce another key idea in quantum mechanics. Unfortunately, this idea was introduced so well in chapter 2 of QED: The Strange Theory of Light and Matter by Richard Feynman, that my mind goes blank when trying to imagine how to i... |
60e27c7b-229a-4a67-88f2-2a5fc74b50a8 | trentmkelly/LessWrong-43k | LessWrong | How do you deal with people cargo culting COVID-19 defense?
In Berlin there are rules for public transportation where all passengers have to wear FFP-2 masks. Nobody, seems to care whether people wear the masks in a way that seals. The security personal in public transportation only wears surgical masks. Whenever, I s... |
a11af3ca-0561-4b8d-a5e1-0be69e23ef1e | trentmkelly/LessWrong-43k | LessWrong | Rational Animations is looking for an AI Safety scriptwriter, a lead community manager, and other roles.
UPDATE 07/15/2023: Applications are now closed. If you still want to apply, I will consider you for the future, which may be as soon as a few months from now, or I could initially offer you a lower salary if you'd ... |
91102492-0234-4a27-8429-50bad0d10369 | trentmkelly/LessWrong-43k | LessWrong | Ask LW: ω-self-aware systems
I was having a conversation with a religious friend of mine who argued that even if materialism was true, we would never be able to fully understand or replicate human intelligence because a physical system cannot understand itself--it would need to use the resources contained within it to... |
7adee56e-5acb-48e1-ab3a-eb487fe23e76 | StampyAI/alignment-research-dataset/arxiv | Arxiv | Robust Cooperation in the Prisoner’s Dilemma: Program Equilibrium via Provability Logic
1 Introduction
---------------
Can cooperation in a one-shot Prisoner’s Dilemma be justified between rational agents? Rapoport [[18](#bib.bib18)] argued in the 1960s that two agents with mutual knowledge of each others’ rational... |
9b402cf7-6032-4963-b97a-2fa205a9fb59 | trentmkelly/LessWrong-43k | LessWrong | Meetup : LW Ottawa meetup
Discussion article for the meetup : LW Ottawa meetup
WHEN: 17 April 2014 07:00:00PM (-0400)
WHERE: Royal Oak on the Canal, 221 Echo Dr, Ottawa, ON K1S 1N1, Canada
There hasn't been a meetup in Ottawa for while, so if you're in the area, here's your chance to attend one! We'll have some gen... |
3e30bf38-38b5-4e34-8641-31ff848872ab | trentmkelly/LessWrong-43k | LessWrong | The best 15 words
People want to tell everything instead of telling the best 15 words. They want to learn everything instead of the best 15 words. In this thread, instead post the best 15-words from a book you've read recently (or anything else). It has to stand on its own. It's not a summary, the whole value needs... |
b5f62c90-881b-4637-a78a-8a63c275517a | trentmkelly/LessWrong-43k | LessWrong | Announcing AI safety Mentors and Mentees
Executive summary
* AI safety Mentors and Mentees is a program that matches mentors with mentees to accelerate their AI safety journey. Our aim is to get more people into full-time AI safety roles faster. We expect nearly all mentees to do this next to their current occupation... |
e6babcb2-965d-434b-adfe-58ad8dcc7b0c | trentmkelly/LessWrong-43k | LessWrong | Bug Report
I've been experiencing a few bugs on LW2.0 in the last few months.
1. I can't get to LessWrong by typing in the url as "lesswrong.com". It only works if I do it as "www.lesswrong.com".
2. Maybe it's just I'm using a DuckDuckGo extension in Google Chrome, but when I type in the search bar "lesswrong.com" a... |
4214cceb-e5ea-414f-a2f0-93c2ac42de9f | trentmkelly/LessWrong-43k | LessWrong | How Old is Smallpox?
The conventional view is that smallpox has been around since antiquity, but more recent evidence has suggested it's actually only around 500 years old.
So I have a research/rationality question: how conclusive is the "500 years old hypothesis"? I don't really have the expertise to evaluate it.
T... |
15040a52-7789-414c-8d44-917a5b12d843 | trentmkelly/LessWrong-43k | LessWrong | Update from the suckerpunch
The most common objection I hear when helping people remove their guilt is something along the lines of "Hey wait! I was using that!"
Believing this (or really any variant of "but guilt is good for me!") makes it fairly hard to replace guilt with something more productive.
I've met some p... |
82b67d53-3510-4f68-930d-198c49f80674 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Boston: Trigger action planning
Discussion article for the meetup : Boston: Trigger action planning
WHEN: 21 June 2015 03:30:00PM (-0400)
WHERE: 98 Elm Street, Somerville
Mick Porter will be presenting on trigger action planning, which is a strategy taught by CFAR to systematize solutions to everyday probl... |
f0c2a32f-4676-4722-8f55-ec04e5feb6e2 | trentmkelly/LessWrong-43k | LessWrong | Stable Diffusion has been released
It's an open-source text-to-image model capable of producing NSFW content. I, for one, I'm very excited to see the consequences this has on society, which I expect to be mostly positive (except if it leads to an increase in AI funding). Any thoughts?
Other relevant links:
https://w... |
f8d52b16-a7aa-468b-b569-49de33b11b7b | trentmkelly/LessWrong-43k | LessWrong | Meetup : Phoenix Tuesday Lunch Group
Discussion article for the meetup : Phoenix Tuesday Lunch Group
WHEN: 10 September 2013 12:00:00PM (-0700)
WHERE: 1290 S. Normal Ave., Tempe, AZ, 85287
This is now a weekly thing! We will be meeting at Engrained in the MU. Currently we are getting to know each other, and have ho... |
e2c05d2c-203b-4d4d-8ac4-450720cad448 | trentmkelly/LessWrong-43k | LessWrong | Experiments and Consent
One of the responses to my Uber self-driving car post was objecting to Uber experimenting on public roads:
> Self-driving research as practiced across the industry is in violation of basic research ethics. They should not be allowed to toss informed consent out the window, no matter how cool o... |
a7e010d7-c2f2-4c5f-9d35-8fd4a90720f1 | StampyAI/alignment-research-dataset/arbital | Arbital | Math playpen
$$\mathcal{P}e^{i \oint_C A_\mu dx^\mu} \to g(x) \mathcal{P}e^{i \oint_C A_\mu dx^\mu} g^{-1}(x)\,$$
Numbers, woo!
Can I bork the page? Trying to bork the page… 3
asdf |
4434f3e2-2ad9-4857-ae65-6d68c220b245 | trentmkelly/LessWrong-43k | LessWrong | On Equivalence of Supergoals
This is a response to Ars Longa, Vita Brevis, an excellent piece by Scott Alexander. In fact, it moved me so much that I signed up on LessWrong just to write this response. I'm going to argue that the essay's central idea is wrong, and that's a good thing. You should read Alexander's essay... |
863072fd-cbcf-48d2-9a47-2f9bbe1ddb80 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Topics to discuss CEV
CEV is our current proposal for what ought to be done once you have AGI flourishing around. Many people have had bad feelings about this. When in Singularity Institute, I decided to write a text do discuss CEV, from what it is for, to how likely it is to achieve it's goals, and how much fine-grai... |
e6f16d52-82a6-4ef6-87e3-d815654cc18e | trentmkelly/LessWrong-43k | LessWrong | Group Rationality Diary, January 1-15
Happy new year to those who are celebrating! This is the public group rationality diary for January 1-15.
> It's a place to record and chat about it if you have done, or are actively doing, things like:
>
> * Established a useful new habit
> * Obtained new evidence that made y... |
70c56b7d-b272-4fb7-b7fa-a763bcb9c97a | trentmkelly/LessWrong-43k | LessWrong | Introduction to Choice set Misspecification in Reward Inference
In classical RL, we have an agent with a set of States (S), a set of action (A), and given some reward function (R), the aim is to find out the optimal policy (pi) which maximizes the following. This is the cummulative rewards we get by sampling actions ... |
cd005a96-d740-45b7-8b7e-e2e65ed11ccd | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Favourite new AI productivity tools?
[Transformative AI](https://www.lesswrong.com/tag/transformative-ai) will revolutionize the ways we interact with the world. But maybe narrow [tool AIs](https://www.lesswrong.com/tag/tool-ai) are already starting to do so: Every week, there are new startups, websites, and tools lau... |
7ec65bc5-0743-467e-8706-b5b93c6e88ce | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Do mesa-optimizer risk arguments rely on the train-test paradigm?
Going by the [Risks from Learned Optimization sequence](https://www.alignmentforum.org/s/r9tYkB2a8Fp4DN8yB), it's not clear if mesa-optimization is a big threat if the model continues to be updated throughout deployment. I suspect this has been discusse... |
9c984b91-aee0-438a-ac3b-aa47a56d1f13 | trentmkelly/LessWrong-43k | LessWrong | [LINK] 23andme is 99$ now
It's been reduced to 99$ and it seems like it is a permanent reduction. I was thinking of buying it at 299$ because it had not been on sale for a while, so I'm very pleased this happened.
Their press release on it:
http://blog.23andme.com/news/one-million-strong-a-note-from-23andmes-anne-wo... |
6ec3a25b-9998-4162-a5be-158e8a610429 | trentmkelly/LessWrong-43k | LessWrong | Learn and practice using Occam's Razor with exercises
I want to be great at making decisions.
But it's hard to practice and even harder to get feedback.
I've tried books, courses, and talks.
While some provide exercises, these resources are mostly informative.
This makes it hard to know if I really understood what... |
acd81d6d-dd28-4630-ba53-365cbe1f035f | trentmkelly/LessWrong-43k | LessWrong | Holy Grails of Chemistry
This is a linkpost-slash-suggestion to check out https://www.chemistryworld.com/holy-grails/.
In 1995, the journal Accounts of Chemical Research published a series of articles called Holy Grails of Chemistry. This series of articles represents an attempt by a group of chemists to make a part... |
12617f92-3ecf-4f6e-afb0-70b46ad8bc64 | trentmkelly/LessWrong-43k | LessWrong | What is it to solve the alignment problem? (Notes)
(I originally wrote this post as some rough notes on defining the alignment problem, with the intention of turning them into something more polished later. I've now started doing that, as part of a broader series introduced here. In particular, the first post in that... |
09c50d54-0ef6-4bd6-a55e-31af7b0956c9 | trentmkelly/LessWrong-43k | LessWrong | Kelly Bet or Update?
I suggest augmenting the classic "bet or update" with "Kelly bet or update".
Epistemic status: feels like maybe going too far, but worth considering?
----------------------------------------
On the one hand, we have a line of thinking in rationalist discourse which says probability is willingne... |
cd1c3380-72e3-4daa-99ea-9b63e12258f4 | StampyAI/alignment-research-dataset/arxiv | Arxiv | Idiotypic Immune Networks in Mobile Robot Control
1 IDIOTYPIC IMMUNE NETWORKS IN MOBILE ROBOT CONTROL
Amanda M. Whitbrook, Uwe Aickelin, Member, IEEE , and Jonathan M. Garibaldi
Abstract—Jerne’s idiotypic network theory postulates that the immune re sponse involves inter-antibody
stimulation and suppression... |
dcc5777e-b2a3-488e-a6a0-e06ce3e9928a | trentmkelly/LessWrong-43k | LessWrong | Comments on Karma systems
A Karma system is a resource sharing mechanism based on an account of non-tradable rights. A recent paper named “Dynamic Recourse Allocation with Karma: An Experimental Study” (Elokda, Nax, Bolognani and Dörfler, 2024) describes this kind of mechanism. To summarize, agents have some rights to... |
0c07264f-7701-42d7-8cd0-ed4d0ba5f5a9 | trentmkelly/LessWrong-43k | LessWrong | Intercept article about lab accidents
The Intercept made FOI requests to gather information about American lab accidents. Additionally, their article discusses the current lack of legislation to require commonsense biosafety strategy from researchers.
> Unless they work with the most dangerous pathogens, biolabs don’... |
211aea14-07c0-470a-878f-2b66a923b2b5 | trentmkelly/LessWrong-43k | LessWrong | Let's reimplement EURISKO!
In the early 1980s Douglas Lenat wrote EURISKO, a program Eliezer called "[maybe] the most sophisticated self-improving AI ever built". The program reportedly had some high-profile successes in various domains, like becoming world champion at a certain wargame or designing good integrated ci... |
ad9274bb-0ae5-4f7a-b9b2-d28c37a34dd6 | trentmkelly/LessWrong-43k | LessWrong | Existential Risk and Existential Hope: Definitions
I'm pleased to announce Existential Risk and Existential Hope: Definitions, a short new FHI technical report.
Abstract:
> We look at the strengths and weaknesses of two existing definitions of existential risk, and suggest a new definition based on expected value. Th... |
496bb459-6bda-4c42-bc97-9b38b1027aa1 | trentmkelly/LessWrong-43k | LessWrong | Site search will be down for a few hours
We are fixing all of the unnecessary and broken search results that currently come up when you search, which means the site search will be down for a few hours. Should be up again by tomorrow at the latest. |
f0fccb7f-a430-4829-9c89-f25b1a95f496 | trentmkelly/LessWrong-43k | LessWrong | Lies, Damn Lies, and Fabricated Options
This is an essay about one of those "once you see it, you will see it everywhere" phenomena. It is a psychological and interpersonal dynamic roughly as common, and almost as destructive, as motte-and-bailey, and at least in my own personal experience it's been quite valuable to... |
483180a1-a576-4a2e-8ec0-517ebeeea7d1 | trentmkelly/LessWrong-43k | LessWrong | 240 questions for your utility function
A game comparing intrinsic values can approximate a person's utility function.
----------------------------------------
First joke: An economist can answer any question phrased in terms of money.
Second joke: An economist receives a call from his artist friend. The artist i... |
df226db4-85d4-482f-b6c8-bdbf2c6bb0f1 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Notes on Gratitude
This post examines the virtues of **appreciation**, **gratitude**, and **reciprocity**. It is meant mostly as an exploration of what other people have learned about these virtues, rather than as me expressing my own opinions about them, though I’ve been selective about what I found interesting or cr... |
d659fe3c-d2ae-44ae-9cec-30b6845201a7 | trentmkelly/LessWrong-43k | LessWrong | Overcoming Akrasia/Procrastination - Volunteers Wanted
Edit: Cohorts are basically full for now, will let people know if spaces open up.
I've got some new material I'm working on a related to overcoming procrastination and akrasia. It breaks down the skill of focus/willpower into seven "mental habits" that when stack... |
5d32529d-e544-43da-b2f9-f016f8289779 | trentmkelly/LessWrong-43k | LessWrong | Selective modularity: a research agenda
Overview: By training neural networks with selective modularity, gradient routing enables new approaches to core problems in AI safety. This agenda identifies related research directions that might enable safer development of transformative AI.
Introduction
Soon, the world may... |
1c8e368b-92df-41f2-8bd7-9d3974440891 | StampyAI/alignment-research-dataset/arxiv | Arxiv | Multitask Soft Option Learning
1 Introduction
---------------
A key challenge in Reinforcement Learning (rl) is to scale current approaches
to higher complexity tasks without requiring
a prohibitive number of environmental interactions.
However, for many tasks, it is possible
to construct or learn efficient explora... |
8f0895a0-5be8-45d9-9ba3-2ea92493deb0 | trentmkelly/LessWrong-43k | LessWrong | On Morality, Ethics, and all that Jazz
Let’s start with some definitions first to make sure that we are all on the same page. I have no idea what the formal definitions are in this space but hopefully this will be enough for mutual understanding.
Agent: Anything that can (seemingly) act to affect the universe around ... |
064f618b-841e-46dd-be0c-021a001b68f7 | trentmkelly/LessWrong-43k | LessWrong | A thought about Internet procrastination
Perhaps this is already well known, but it occurred to me yesterday and I thought I'd share it. The Internet seems particularly virulent as a form of procrastination; indeed, if, say, chatting at watercoolers took up as much time in the average office worker's day, we wouldn't ... |
075eee44-b5ea-41cd-8710-089e328cadee | StampyAI/alignment-research-dataset/youtube | Youtube Transcripts | How I think students should orient to AI safety | Buck Shlegeris | EA Student Summit 2020
hello everyone and welcome to this
session
with buck sluggarus on ai safety
i'm sophia davis fogel and i'll be the
emcee for this session
we'll be starting with a 15-minute talk
by buck then we'll move on to a live q a
session wh... |
cd443dcb-4a61-4600-b2aa-3ab3422a622f | trentmkelly/LessWrong-43k | LessWrong | The Theoretical Reward Learning Research Agenda: Introduction and Motivation
At the time of writing, I have just (nearly) finished my PhD at Oxford. During that time, most of my main research has been motivated by the goal of developing a theoretical foundation for the field of reward learning. The purpose of this seq... |
5d975b24-efdb-4004-9734-c65caba02046 | trentmkelly/LessWrong-43k | LessWrong | Meetup : San Francisco Meetup: Board Games
Discussion article for the meetup : San Francisco Meetup: Board Games
WHEN: 13 March 2017 06:15:00PM (-0800)
WHERE: 1769 15th St, San Francisco, CA
The meetup is at 6:15 PM, PDT. The LW time picker does surprising things here, and I'm worried it will change from under me.
... |
25a43f63-823e-4782-b148-a026c75362ff | trentmkelly/LessWrong-43k | LessWrong | Mere exposure effect:
Bias in Evaluating AGI X-Risks
Note on unusual formatting: Sentences are split into lines so you can parse parts precisely.
Also called an 'Availability cascade',
a self-reinforcing process in which
a collective belief gains more and more plausibility
through its increasing repetitio... |
8cfbd213-add5-4c64-9c30-4b69e38100e6 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Are we all misaligned?
The orthogonality thesis separates intelligence and goals, constraining the notion of intelligence to instrumental rationality and allowing for any combination of general intelligence and a goal system. And the other way around, such a description should apply to any possible agency. For any age... |
01039088-2b00-484f-a86f-e06c683d6b08 | trentmkelly/LessWrong-43k | LessWrong | Blunt Razor
In this post, I would like to share a simple observation of mine which is about deriving some crude form of Occam's razor from a small set of assumptions.
Assumptions:
1. Subjective Bayesian probability is used to model an agent's degree of belief in hypotheses.
2. The set of hypotheses H that we assi... |
3b51b714-7fe5-4bd4-9ba4-f1ceb1c75304 | trentmkelly/LessWrong-43k | LessWrong | shoes with springs
There's something intuitively intriguing about the concept of shoes with spring elements, something that made many kids excited about getting "moon shoes", but they found the actual item rather disappointing. Using springs somehow with legged movement also makes some logical sense: walking and runni... |
e16f43ac-53b7-4131-993a-f2acdd26e6e1 | trentmkelly/LessWrong-43k | LessWrong | Why Most Intentional Communities Fail (And Some Succeed)
|
ea4589da-48b6-44f5-830d-840330c02c3f | StampyAI/alignment-research-dataset/arxiv | Arxiv | On Catastrophic Interference in Atari 2600 Games
1 Introduction
---------------
Despite many notable successes in deep reinforcement learning (RL) most model-free algorithms require huge numbers of samples to learn effective and useful policies (Mnih et al., [2015](#bib.bib79 "Human-level control through deep reinf... |
cf1b65fc-3d3d-4dba-9ee5-dc595c5d3c12 | trentmkelly/LessWrong-43k | LessWrong | The increasing uselessness of Promoted
For some time now, "Promoted" has been reserved for articles written by MIRI staff, mostly about MIRI activities. Which, I suppose, would be reasonable, if this were MIRI's blog. But it isn't. MIRI has its own blog. It seems to me inconvenient both to readers of LessWrong, an... |
ef99b3e4-bf57-45d9-b6d1-2ff4cc7604d0 | trentmkelly/LessWrong-43k | LessWrong | Monthly Shorts 8/22
Politics
> In a 2017 analysis of 229,000 FOIA requests, those from journalists accounted for just 8 percent. In 2020, there were nearly 800,000 requests made. At some federal agencies, the vast majority of requests are now from commercial operators who resell or use data for profit. Their turf is w... |
c39b5e63-f12f-4824-bb6c-31d7e30052b6 | trentmkelly/LessWrong-43k | LessWrong | Questions about ''formalizing instrumental goals"
Epistemic Status: Autodidact outsider who suspects he has something to add to the conversation about AI risk.
Abstract
This essay raises questions about the methodology, and thus conclusions, reached in the paper “Formalizing Convergent Instrumental Goals.” This pap... |
6b1d63da-7493-4a7b-a9c9-c7fb8ecfa8bd | trentmkelly/LessWrong-43k | LessWrong | Costs and benefits of metaphysics
Prelude for LW
* Section 2 contains a bunch of self-references to personal articles, all of which are the kind I also posted on less wrong before but considering the amount of self-referncing this article might come out as a bit douchy. If you hate that kind of thing stop reading her... |
1c2dc1b2-acd3-4f9c-8e0b-134d34c4f145 | trentmkelly/LessWrong-43k | LessWrong | A simple thought experiment showing why recessions are an unnecessary bug in our economic system
On Monday we all go to the offices and factories we work at and produce goods and services. We exchange these goods and services among each other for mutual benefit.
On Tuesday it's declared there is a recession. The same... |
d7a7b9dd-e820-4e79-845e-26865f223935 | trentmkelly/LessWrong-43k | LessWrong | The causal backbone conjecture
Followup to: Rationalists are missing a core piece for agent-like structure (energy vs information overload). Related to: Linear Diffusion of Sparse Lognormals (I don't consider that sequence done yet BTW, I just took a break because I ran out of pre-written posts, so I need to write som... |
8115d5cd-6964-4dc4-b97b-ceeef9d12007 | trentmkelly/LessWrong-43k | LessWrong | Open thread, Dec. 14 - Dec. 20, 2015
If it's worth saying, but not worth its own post (even in Discussion), then it goes here.
----------------------------------------
Notes for future OT posters:
1. Please add the 'open_thread' tag.
2. Check if there is an active Open Thread before posting a new one. (Immediately... |
1ae2b5a6-69a8-4ff6-85fd-b9b034b281cc | trentmkelly/LessWrong-43k | LessWrong | Alien Message Contest: Solution
Thanks to everyone who participated in the Alien Message contest, and congratulations on successfully being able to decode the message. I gave a time limit of 2 weeks at the start of the contest, but you completely solved it in a mere 2 days.
Solution
Here is the code that generated t... |
b90d280b-6626-4eaa-b5cb-c26bcb1e716c | trentmkelly/LessWrong-43k | LessWrong | Covid 2/2/23: The Emergency Ends on 5/11
We finally have a date. On May 11, the official Covid Emergency ends. The actual emergency is, of course, long over, but for technical reasons the official change will have to wait a bit.
A related question is when it stops making sense to continue writing weekly Covid posts. ... |
20d60004-aeec-4b52-a24d-d1c845eac839 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Epistemology of HCH
Introduction
============
HCH is a recursive acronym meaning “Humans consulting HCH”. Coincidentally, It’s also a concept coined by Paul Christiano, central in much of the reasoning around [Prosaic AI Alignment](https://www.alignmentforum.org/posts/YTq4X6inEudiHkHDF/prosaic-ai-alignment). Yet for ... |
fb0abc79-8d0b-49aa-863f-117195930c95 | trentmkelly/LessWrong-43k | LessWrong | What would “The Medical Model Is Wrong” look like?
So you’ve been reading down a rabbit hole and you have seen something that doesn’t make sense. You feel a quiver in your heart that feels like a cross between excitement and fear. Could all the people who already know stuff be Wrong™. Could eminent researchers in t... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.