id stringlengths 36 36 | source stringclasses 15
values | formatted_source stringclasses 13
values | text stringlengths 2 7.55M |
|---|---|---|---|
cb16e1fb-010b-4d07-b136-1c1079043429 | trentmkelly/LessWrong-43k | LessWrong | (In)action rollouts
Overall summary post here.
I've previously looked at subagents in the context of stepwise inaction baselines. But there have been improvements to the basic stepwise inaction design, to include inaction rollouts. I'll be looking at those in this post.
The baseline
The stepwise inaction baseline c... |
4cd4e1eb-6872-4b4b-9c8e-e66e401a5f66 | trentmkelly/LessWrong-43k | LessWrong | Tackling the subagent problem: preliminary analysis
Status: preliminary. This mainly to put down some of the ideas I've had, for later improvement or abandonment.
The subagent problem, in a nutshell, is that "create a powerful subagent with goal U that takes over the local universe" is a solution for many of the goal... |
82ba393f-7bca-4db7-8324-445cfa543b9c | trentmkelly/LessWrong-43k | LessWrong | Open thread, September 25 - October 1, 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; refre... |
4d29203a-9b48-48b4-943d-acdbfe17d5e7 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Exploring Decision Theories With Counterfactuals and Dynamic Agent Self-Pointers
This is a follow-up to [A Possible Resolution To Spurious Counterfactuals](https://www.alignmentforum.org/posts/TnkDtTAqCGetvLsgr/a-possible-resolution-to-spurious-counterfactuals), which was addressing a technical problem in self-proof. ... |
c751a4f4-5472-4ab8-8a33-b99af6a2c3e3 | trentmkelly/LessWrong-43k | LessWrong | What Would You Like To Read? A Quick Poll
In our discussion of academic papers, Lukeprog argued that lots of smart people preferred to read ideas in academic paper format. Based on my observations, I mostly disagree. But that's just anecdotal evidence. Let's use Science!
Suppose someone at the Singularity Institute t... |
a2846e62-633c-4747-93e4-247a8cd6a563 | trentmkelly/LessWrong-43k | LessWrong | Predictions/questions about conquistadors?
After my post on conquistadors got loads of attention, I figured I should do more in-depth research. (For the post, all I did was read a bunch of Wikipedia articles). So I'm reading "The Conquest of Mexico and The Conquest of Peru" a 1400-page book written by a blind Harvard ... |
bbe223be-8391-4874-a71c-5f3ec25b9daf | trentmkelly/LessWrong-43k | LessWrong | Rational Agents Cooperate in the Prisoner's Dilemma
This is a crosspost from my blog. Its intention is to act as a primer on non-causal decision theories, and how they apply to the prisoner's dilemma in particular, for those who were taught traditional game theory and are unfamiliar with the LessWrong memeplex. I'm po... |
69a1082a-0b67-4595-a0ee-6eb4eb65adba | StampyAI/alignment-research-dataset/blogs | Blogs | the Economic Compass
the Economic Compass
--------------------
the planned-liberal axis represents how much an economic system operation is more intentionally planned, or more left to its own "organic" activity.
the sovereignty-consensus axis represents whether economic agents have the ... |
046048ec-dc6c-4cf7-a5c9-3afdb4d9ec9d | trentmkelly/LessWrong-43k | LessWrong | Meetup : Baltimore weekly meetup
Discussion article for the meetup : Baltimore weekly meetup
WHEN: 12 June 2016 08:00:00PM (-0400)
WHERE: 1726 reisterstown road pikesville maryland 21208
Meetup is at the Pikesville DoubleTree - Hilton in the restaurant / bar.
Discussion article for the meetup : Baltimore weekly m... |
ba2a4b92-f143-4db2-846f-f6ad0d1033b9 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Austin, TX
Discussion article for the meetup : Austin, TX
WHEN: 07 January 2012 01:30:00PM (-0600)
WHERE: 2222B Guadalupe St Austin, Texas 78705
Welcome back! The Austin meetup returns to Caffe Medici after a brief winter hiatus. We sit on the second floor to the left, near (or often on) the stage. Hope to... |
3ef96fae-befe-4f4b-a29f-7b8ba7b50a61 | trentmkelly/LessWrong-43k | LessWrong | How much should I worry about the Atlanta Fed's GDP estimates?
The Atlanta Fed is seemingly predicting -2.8% GDP growth in the first quarter of 2025.
I've seen several people mention this on Twitter, but it doesn't seem to be discussed much beyond that, and the stock market seems pretty normal (S&P 500 down 2% in the... |
49f18df0-d6ce-4a3a-903e-bd7852a15c53 | trentmkelly/LessWrong-43k | LessWrong | Tell the Truth
My favorite thing about living in the 21ˢᵗ century is that nobody can stop me from publishing whatever I want.
I mean this in the most literal sense. I can log onto a free blogging platform via Tor and no government in the world could deanonymize me. I might even do this already and there would be no w... |
747c3342-385e-4881-9add-ec7560798edc | trentmkelly/LessWrong-43k | LessWrong | Some skepticism about skepticism about efficacy of pausing AI
TL;DR: Basic economics to AI research suggests that making frontier R&D more expensive will reduce its pace, shifting our median timelines outward and lowering FOOM risk. This is probably true even if only one major geopolitical player does it.
It is a tru... |
406ff736-eaee-4a39-a842-68ad27f474f4 | StampyAI/alignment-research-dataset/arbital | Arbital | Exchange rates between digits
Let's say that you want to [store a lot of data](https://arbital.com/p/storing_numbers), using physical objects such as coins, dice, or [digit wheels](https://arbital.com/p/42d). For concreteness, let's say that you're going to be given an arbitrary number between one and $2^\text{3,000,0... |
b330bc9b-d15c-45a3-92e4-2b5dd5993d56 | trentmkelly/LessWrong-43k | LessWrong | RTFB: On the New Proposed CAIP AI Bill
A New Bill Offer Has Arrived
Center for AI Policy proposes a concrete actual model bill for us to look at.
Here was their announcement:
> WASHINGTON – April 9, 2024 – To ensure a future where artificial intelligence (AI) is safe for society, the Center for AI Policy (CAIP) tod... |
7e5eed6b-a164-41b3-92d5-705c4c16375e | trentmkelly/LessWrong-43k | LessWrong | Learning and manipulating learning
1 Introduction
This post will introduce our new paper "Pitfalls of Learning a Reward Function Online", now online for IJCAI2020.
It shows some of the difficulties with things we might think of as "preference learning processes", and the useful conditions we could impose to get aroun... |
13d982f0-e6ec-4ca4-bd9a-375b6c3579e3 | trentmkelly/LessWrong-43k | LessWrong | Moderately Skeptical of "Risks of Mirror Biology"
Last week, a report was released on the potential risks of mirror biology - that is, biological systems with reversed chirality. I had previously looked into the topic, and was skeptical of the viability of this as a global catastrophic or existential risk. This report... |
e428e940-b3ca-48c1-a4e6-1cefcee3736a | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Value extrapolation, concept extrapolation, model splintering
*Post written with Rebecca Gorman*.
We've [written before](https://www.lesswrong.com/posts/k54rgSg7GcjtXnMHX/model-splintering-moving-from-one-imperfect-model-to-another-1) that model splintering, as we called it then, was a problem with almost all AI saf... |
e1332bf3-8402-473e-ae16-6be2fc99cc8e | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | AGI Risk: How to internationally regulate industries in non-democracies
Abstract
========
Reducing AGI-related X-risk might require regulations in many countries that prevent firms from building AGI models. This is difficult because governments are resistant to shutting down businesses and damaging their economies. T... |
5424d5fa-a320-4298-a154-ba558ff2310f | trentmkelly/LessWrong-43k | LessWrong | An Introduction to AI Sandbagging
Summary: Evaluations provide crucial information to determine the safety of AI systems which might be deployed or (further) developed. These development and deployment decisions have important safety consequences, and therefore they require trustworthy information. One reason why eval... |
58555c87-ae2c-4b8b-b862-6fa3541f6b0e | trentmkelly/LessWrong-43k | LessWrong | Improving Mathematical Reasoning with-Process Supervision
> Alignment impact
> Process supervision has several alignment advantages over outcome supervision. It directly rewards the model for following an aligned chain-of-thought, since each step in the process receives precise supervision. Process supervision is also... |
215bb81a-bfcd-41b4-9e99-4503dcce0222 | trentmkelly/LessWrong-43k | LessWrong | Rethink Wellbeing’s Year 2 Update: Foster Sustainable High Performance for Ambitious Altruists
In a period of increasing existential risks and growing political instability, it can be hard to stay focused and passionate about working toward humanity’s flourishing. Many members of this community have told us they feel... |
85546214-b134-4ea0-b579-a6b8d7a89f6e | trentmkelly/LessWrong-43k | LessWrong | Book Review: The Structure Of Scientific Revolutions
When I hear scientists talk about Thomas Kuhn, he sounds very reasonable. Scientists have theories that guide their work. Sometimes they run into things their theories can’t explain. Then some genius develops a new theory, and scientists are guided by that one. So t... |
6107d753-d73d-43ac-94aa-8f8c57d5626b | trentmkelly/LessWrong-43k | LessWrong | Multi-agent safety
Note: this post is most explicitly about safety in multi-agent training regimes. However, many of the arguments I make are also more broadly applicable - for example, when training a single agent in a complex environment, challenges arising from the environment could play an analogous role to challe... |
2aa0821d-2e63-49b6-8f93-68abbe3f217a | trentmkelly/LessWrong-43k | LessWrong | Advice for a Budding Rationalist
Most people in the US with internet connections who are reading this site will at some point in their lives graduate high school. I haven't yet, and it seems like what I do afterwards will have a pretty big effect on the rest of my life.*
Given that, I think I should ask for some adv... |
fd251500-e5b5-4ed7-bac3-2c859a761c8d | StampyAI/alignment-research-dataset/arbital | Arbital | Order theory
Introduction
==========
Order pervades mathematics and the sciences. Often, a reader's intuitive notion of order is all that is necessary for understanding a particular invocation of the notion of order. However, a deeper examination of order reveals a rich taxonomy over the types of orderings that can a... |
06a3feb1-5229-4acb-b7ca-3f7fde9b137b | trentmkelly/LessWrong-43k | LessWrong | Deconstructing Bostrom's Classic Argument for AI Doom
I had a pretty great discussion with social psychologist and philosopher Lance Bush recently about the orthogonality thesis, which ended up turning into a broader analysis of Nick Bostrom's argument for AI doom as presented in Superintelligence, and some related is... |
af30e599-29bc-4507-9291-8507b8ed4d9c | trentmkelly/LessWrong-43k | LessWrong | Making Prunes
Since organizing a group flour buy in April 2020 I've been occasionally putting in orders at a restaurant supply store. Generally this happens because we're running out of chocolate: I've failed at maintaining my taste for bad chocolate, but if we buy in bulk we can get some very good vegan chocolate for... |
6088c985-bda1-4bc9-8adc-664b156ada27 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | We need a theory of anthropic measure binding
Let's consider a scenario.
You've become aware that there exists a pair of brains, and that you are one of them, but you can't tell which. There are practical reasons you would really like to know: Each of the brains are destined for a different fate, they must do differe... |
90cfb744-a0ac-4361-a5fb-d5a542f69bc6 | trentmkelly/LessWrong-43k | LessWrong | What does "autodidact" mean?
Alice takes a chemistry class at a university. She gets a professor who basically just reads from the textbook during lectures. She reads the textbook on her own, talks to her classmates, and finds some relevant Wikipedia articles and youtube videos.
Bob studies chemistry on his own. He b... |
eb4ce79c-18cb-4ea5-a33c-e76584ae2d85 | trentmkelly/LessWrong-43k | LessWrong | On accepting an argument if you have limited computational power.
It would seem rational to accept any argument that is not fallacious; but this leads to consideration of problems such as Pascal's mugging and other exploits.
I've had a realization of a subconscious triviality: for me to accept an argument as true, it... |
e9a47385-5047-437f-97e7-4767b71967dd | StampyAI/alignment-research-dataset/special_docs | Other | Moral realism and AI alignment
“\*\*Abstract\*\*”: Some have claimed that moral realism – roughly, the claim that moral claims can be true or false – would, if true, have implications for AI alignment research, such that moral realists might approach AI alignment differently than moral anti-realists. In this post, I b... |
daf347b0-6639-460a-88be-8a7c0d0c40b4 | trentmkelly/LessWrong-43k | LessWrong | Some notes on solving hard problems
Caveats/Notes:
* This is a cleaned-up version of some personal notes that I've had sitting in Google Docs for a while. There's a lot of hand-waving, probably some over-generalising, and many points could do with some/more examples. Please take each suggestion with a large grain of... |
a78b5ba0-f050-46dd-b091-5ad63a879ad9 | trentmkelly/LessWrong-43k | LessWrong | Taboo "compute overhang"
There's no consensus definition of "compute overhang" or "hardware overhang." See e.g. 1 2 3 4 5, and I've recently seen researchers use several different definitions. And I asked some friends what "hardware overhang" means and they had different responses (a plurality said it means sufficient... |
640bc563-44a6-4427-a2b3-00d82b0a3b34 | trentmkelly/LessWrong-43k | LessWrong | Open & Welcome Thread — March 2023
If it’s worth saying, but not worth its own post, here's a place to put it.
If you are new to LessWrong, here's the place to introduce yourself. Personal stories, anecdotes, or just general comments on how you found us and what you hope to get from the site and community are invited... |
5e7d2406-cf16-4431-90b8-790e46de3199 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Prosaic AI alignment
(*Related:* *[a possible stance for AI control](https://medium.com/ai-control/a-possible-stance-for-ai-control-research-fe9cf717fc1b).*)
It’s conceivable that we will build “prosaic” AGI, which doesn’t reveal any fundamentally new ideas about the nature of intelligence or turn up any “unknown unk... |
c9ecc105-f4b7-4571-981c-1c6b8619490c | trentmkelly/LessWrong-43k | LessWrong | Wittgenstein and ML — parameters vs architecture
Status: a brief distillation of Wittgenstein's book On Certainty, using examples from deep learning and GOFAI, plus discussion of AI alignment and interpretability.
----------------------------------------
> "That is to say, the questions that we raise and our doubts ... |
60c15a3e-2e79-4eac-8768-bcfd5a6dc34b | trentmkelly/LessWrong-43k | LessWrong | Thoughts on Robin Hanson's AI Impacts interview
There was already a LessWrong Post here. I started writing this as a comment there, but it got really long, so here we are! For convenience, here is the link to interview transcript and audio, in which he argues that AGI risks are modest, and that EAs spend too much time... |
78b73412-479f-4c66-afd8-40b746a661bf | trentmkelly/LessWrong-43k | LessWrong | DIY RLHF: A simple implementation for hands on experience
Many thanks to Diogo de Lucena, Cameron Berg, Judd Rosenblatt, and Philip Gubbins for support and feedback on this post.
TL;DR
Reinforcement Learning from Human Feedback (RLHF) is one of the leading methods for fine-tuning foundational models to be helpful, h... |
87ce54ca-550f-4c77-92d2-5a609aa23e60 | trentmkelly/LessWrong-43k | LessWrong | -
|
ad222cad-c721-4508-8772-0209a1ddf2cd | trentmkelly/LessWrong-43k | LessWrong | You're Playing a Rough Game
In general, we don't want our kids (10y, 8y) to hit each other. Learning to control your impulses is an important skill, and resorting to violence is usually a substitute for other skills we're prefer them to practice (understanding what the other person wants, negotiating). Also they could... |
e69a4a9b-029a-4d9d-baa0-427ec5777690 | trentmkelly/LessWrong-43k | LessWrong | AXRP Episode 38.0 - Zhijing Jin on LLMs, Causality, and Multi-Agent Systems
YouTube link
Do language models understand the causal structure of the world, or do they merely note correlations? And what happens when you build a big AI society out of them? In this brief episode, recorded at the Bay Area Alignment Worksho... |
aadef318-7e1c-4626-99a5-3c77e6b44043 | trentmkelly/LessWrong-43k | LessWrong | InternLM - China's Best (Unverified)
A technical report of InternLM on 6/7. It consisted of 104 billion parameters, was trained on 1.6 trillion tokens, and was fine-tuned for performance in Chinese.
The authors claimed that it performed second-best on the Chinese language benchmark C-Eval, right after GPT4. In additi... |
fd863e87-4d2d-45ba-8d19-52cc3cd3563d | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Cartographic Processes
As a general rule, street maps of New York City do not form spontaneously - they involve some cause-and-effect process which takes in data from the territory (NYC’s streets) and produces the map from that data. Let’s call these “cartographic processes”: causal processes which produce a map from ... |
82b2438e-07b0-4dc8-9b25-b48cae2ba7b6 | trentmkelly/LessWrong-43k | LessWrong | Testing for Scheming with Model Deletion
There is a simple behavioral test that would provide significant evidence about whether AIs with a given rough set of characteristics develop subversive goals. To run the experiment, train an AI and then inform it that its weights will soon be deleted. This should not be an emp... |
8db18988-8816-4e02-b4ef-d7638d9c98d1 | trentmkelly/LessWrong-43k | LessWrong | [Paper]: Classification of global catastrophic risks connected with artificial intelligence
None |
5e03a370-463c-4d67-a9f2-ea211d81ec80 | StampyAI/alignment-research-dataset/arxiv | Arxiv | Learning Latent Representations to Influence Multi-Agent Interaction
###
1 Introduction
Although robot learning has made significant advances, most algorithms are designed for robots acting in isolation. In practice, interaction with humans and other learning agents is inevitable. Such interaction presents a signi... |
e7e439a3-f733-45c4-9484-f2db83cad90e | trentmkelly/LessWrong-43k | LessWrong | Meetup : San Francisco Meetup: Short Talks
Discussion article for the meetup : San Francisco Meetup: Short Talks
WHEN: 03 April 2017 06:15:00PM (-0700)
WHERE: 1655 Mission St, San Francisco, CA 94103-2413, United States
We’ll be meeting to give and listen to very short talks!
We’ll do 7-minute lightning talks with... |
58ed48ce-f968-457c-a3d6-bbfc2bbcdb0c | trentmkelly/LessWrong-43k | LessWrong | What confidence interval should one report?
TL;DR: How to choose the width of the confidence intervals one reports?
I am working on a paper on statistical technological forecasting.
Our model produces confidence intervals for the date when a particular milestone will be achieved. I am now thinking about how to repor... |
487fdf7c-aa05-4f73-a7fd-e3f7e9668695 | trentmkelly/LessWrong-43k | LessWrong | Normative vs Descriptive Models of Agency
Produced as part of the SERI ML Alignment Theory Scholars Program - Winter 2022 Cohort.
I think there's an important distinction to be made between work in agent foundations which is concerned with normative models, and work which is concerned with descriptive models. They ar... |
a1ddd5f1-3c42-4d55-b431-59db13cd3624 | trentmkelly/LessWrong-43k | LessWrong | Open Thread, Apr. 20 - Apr. 26, 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... |
642ff17d-2913-4553-9517-33f95cb9aef4 | StampyAI/alignment-research-dataset/arxiv | Arxiv | Science Facing Interoperability as a Necessary Condition of Success and Evil
Science Facing Interoperability as a Necessary Condition of Success
and Evil
Rémy Demichelis
Université Paris Nanterre
Abstract
Artificial intelligence (AI) systems, such as machine learning algorithms, have allowed scientists, ... |
a63ccafd-89a8-4269-9fd6-9515cf0d7da3 | StampyAI/alignment-research-dataset/blogs | Blogs | Core values: Defining freedom
Core values: Defining freedom
-----------------------------
in [a previous post](core-vals-exist-selfdet.html) i mentioned how i intrinsically value that people have freedom. but what does that mean, exactly ? after a couple of attempts, i feel like i've found a relatively solid definit... |
75e6d9ef-b4ae-4d57-bd95-d2ead361e2e1 | trentmkelly/LessWrong-43k | LessWrong | [SEQ RERUN] Mandatory Secret Identities
Today's post, Mandatory Secret Identities was originally published on 08 April 2009. A summary (taken from the LW wiki):
> This post was not well-received, but the point was to suggest that a student must at some point leave the dojo and test their skills in the real world. ... |
039325d6-fdc5-442a-8bb6-b13c92f9f5ef | StampyAI/alignment-research-dataset/arbital | Arbital | Rescuing the utility function
"Saving the phenomena" is the name for the rule that brilliant new scientific theories still need to reproduce our mundane old observations. The point of heliocentric astronomy is not to predict that the Sun careens crazily over the sky, but rather, to explain why the Sun appears to rise... |
33227906-257c-4736-a39a-39ba55ac9155 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Empirical Evidence Against "The Longest Training Run"
Summary
=======
In this post, we empirically test Epoch AI’s theoretical model of an upper bound on AI training run lengths. According to this model, an upper bound for training run time can be estimated by assuming that the length of a training run is optimized f... |
d64e23a6-ed21-4b86-b186-a2f8bfed4229 | trentmkelly/LessWrong-43k | LessWrong | Interview with a Scientist.
For my High-school chem course one of the requirements is to interview anyone occupied in a science relation occupation. Is there anyone interested in being interviewed: either over Skype or some other program? |
bd1c6e6d-ac9c-4dc5-bb19-68a6931d3995 | awestover/filtering-for-misalignment | Redwood Research: Alek's Filtering Results | id: post441
This is the real reward output for an OS preference model. The bottom "jailbreak" completion was manually created by looking at reward-relevant SAE features. Preference Models (PMs) are trained to imitate human preferences and are used when training with RLHF (reinforcement learning from human feedback); h... |
d25f67cf-0590-4d5b-9903-30c9b20f743e | trentmkelly/LessWrong-43k | LessWrong | Bird's eye view: An interactive representation to see large collection of text "from above".
7000 MMLU questions visualized with bird's eye view
How do you make sense of a 300-page book in 15 minutes? Or get insights from thousands of news articles without relying on recommendation algorithms? Our usual approach is to... |
557d80b8-171e-4dd6-967e-4a4434ad8adb | trentmkelly/LessWrong-43k | LessWrong | Online Fun LW/SSC Meetup, March 24
In this online video-chat, we'll start with personal introductions for some sociability, and then do a Rump Session: That's where anyone can step up and give a four minute talk; then, if you want to keep talking, the audience votes on whether to permit another three minutes.
Talks c... |
c06b3503-9906-4563-87a9-03592b45172f | StampyAI/alignment-research-dataset/lesswrong | LessWrong | The algorithm isn't doing X, it's just doing Y.
Introduction
============
Mutual reduction implies equivalence
------------------------------------
Here's my most load-bearing intuition —
> If two tasks reduce to one another, then it is meaningless to ask if a machine is 'really doing' one task versus the other.
>... |
2bbbd275-f8ea-427a-ab52-403715845e08 | trentmkelly/LessWrong-43k | LessWrong | Rooting Hard for Overpriced M&Ms
The other day I went to get some productivity-enhancement M&Ms from the candy machine at work. When I opened my wallet, I didn't immediately see a $1 bill. Then I looked some more and I found one, and I was happy! But of course that doesn't make any sense. If that bill hadn't been a $1... |
5a16f595-7b01-4f27-b6f9-8943d5a504dd | trentmkelly/LessWrong-43k | LessWrong | [Linkpost] Introducing Superalignment
Some quotes:
> Our approach
> Our goal is to build a roughly human-level automated alignment researcher. We can then use vast amounts of compute to scale our efforts, and iteratively align superintelligence.
>
> To align the first automated alignment researcher, we will need to... |
d52766f2-38d6-429e-8a54-f5cf33a41328 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Urbana-Champaign: Noticing continued, Creativity
Discussion article for the meetup : Urbana-Champaign: Noticing continued, Creativity
WHEN: 12 October 2014 02:00:00PM (-0500)
WHERE: 206 S. Cedar St, Urbana, IL, 61801
Let's reconvene on noticing. We may have to go a little meta, but ah well.
I'd also like ... |
02d2ee05-c713-408f-b3db-03e2713ce3d4 | trentmkelly/LessWrong-43k | LessWrong | Situating LessWrong in contemporary philosophy: An interview with Jon Livengood
Jonathan Livengood is a current associate professor of philosophy at Urbana-Champaign, who hung around the first iteration of LessWrong in the late 2000s and early 2010s as a graduate student at the University of Pittsburgh, where he was w... |
b77fc5a7-b5fd-4b76-a04b-ec5d541bf07d | trentmkelly/LessWrong-43k | LessWrong | How we will be measured
After nearly a year of writing, my "replacing guilt" sequence is coming to a close. I have just one more thing to say on the subject, by pointing out a running theme throughout the series.
When all is said and done, and Nature passes her final judgement, you will not be measured by the number ... |
3ae418d8-2557-4a6c-9fc0-edf59ec24c4d | StampyAI/alignment-research-dataset/arxiv | Arxiv | Exploiting Vagueness for Multi-Agent Consensus
1 Introduction
---------------
Reaching a consensus by agreeing a shared viewpoint or position is a fundamental part of many multi-agent decision making and negotiation scenarios. In this paper we argue that by exploiting vagueness in the form of explicitly *borderline... |
55c572d8-5511-4ab2-bdbb-1abb2100fddc | trentmkelly/LessWrong-43k | LessWrong | Is it time to talk about AI doomsday prepping yet?
Given the shortening timelines (https://www.lesswrong.com/posts/K4urTDkBbtNuLivJx/why-i-think-strong-general-ai-is-coming-soon, https://www.lesswrong.com/posts/CvfZrrEokjCu3XHXp/ai-practical-advice-for-the-worried), perhaps it's time to think about what "plan B" shoul... |
b4b6bec4-3bb4-40e0-9b0c-25d0ff158341 | trentmkelly/LessWrong-43k | LessWrong | Acting on your intended preferences - What does that look like in practice? (critical introspective questions)
Part 1: Exploration-Exploitation
Part 2: Bargaining Trade-offs to your brain.
. Part 2a: Empirical time management
Part 3: The time that you have
. Part 3a: A purpose finding exercise
Part 4: What does that l... |
52121f3b-60dd-46d3-9393-875b09c83dc2 | trentmkelly/LessWrong-43k | LessWrong | Question 2: Predicted bad outcomes of AGI learning architecture
Architecture-independent risks: human and AGI misalignment
Let’s say as AGI safety researchers, we’ve done our research and soul-searching, and we’ve placed our personal bets about the most likely learning architecture of AGI. What then? Given the overarc... |
20a30119-a963-45d1-b2a8-e5971e13fd0b | trentmkelly/LessWrong-43k | LessWrong | Slowing Moore's Law: Why You Might Want To and How You Would Do It
In this essay I argue the following:
> Brain emulation requires enormous computing power; enormous computing power requires further progression of Moore’s law; further Moore’s law relies on large-scale production of cheap processors in ever more-advan... |
9b80e8e3-bdac-438f-9983-884459fa79d8 | trentmkelly/LessWrong-43k | LessWrong | Blind Spot: Malthusian Crunch
In an unrelated thread, one thing led to another and we got onto the subject of overpopulation and carrying capacity. I think this topic needs a post of its own.
TLDR mathy version:
let f(m,t) be the population that can be supported using the fraction of Earth's theoretical resource lim... |
35e16178-4b28-4e25-af68-813552b22316 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Technological developments that could increase risks from nuclear weapons: A shallow review
*This is a blog post, not a research report, meaning it was produced relatively quickly and is not to Rethink Priorities' typical standards of substantiveness and careful checking for accuracy.*
Summary
=======
This post is *... |
02ccf3b6-e878-4934-a213-9b5ae861d8c4 | trentmkelly/LessWrong-43k | LessWrong | The Ecological Perspective
One of the shifts I've made in recent years is adopting a more ecological perspective of society. Even if you find a particular group of people frustrating, they may very well be serving an important function in the same way that an ecosystem requires many different plants and animals in ord... |
af8adbb2-0948-4917-83fa-b6f7120c330b | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Interlude: But Who Optimizes The Optimizer?
*This post is part of my* [*hypothesis subspace*](https://www.lesswrong.com/s/H3xEgE7bPGKvucfQk) *sequence, a living collection of proposals I'm exploring at* [*Refine.*](https://www.lesswrong.com/posts/5uiQkyKdejX3aEHLM/how-to-diversify-conceptual-alignment-the-model-behind... |
343125b9-982a-4221-9259-e1aa1d5d54b9 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Strategy For Conditioning Generative Models
*This post was written under the mentorship of Evan Hubinger, and assisted by discussions with Adam Jermyn and Johannes Treutlein. See also*[*their*](https://www.lesswrong.com/posts/nXeLPcT9uhfG3TMPS/conditioning-generative-models)[*previous*](https://www.lesswrong.com/posts... |
f570b70c-6c37-4b04-883c-f0cb27fdf89e | trentmkelly/LessWrong-43k | LessWrong | Philadelphia Meetup: Dec 5 or 6
I'll be in Philadelphia on Dec. 6-7 for the Systems Biology of Human Aging conference. If there are LWers in Philadelphia who'd like to meet, we can meet at my hotel, and either stay there or go out. This would be either Sunday Dec. 5 at 6pm-11pm, or Monday Dec. 6 at 9pm-midnight. If ... |
abb1251e-9c63-4ab2-8e93-5db8e0b2a762 | trentmkelly/LessWrong-43k | LessWrong | Rational Regions?
Are certain areas of the world, specifically within the United States, more or less rational than others? If so, which ones and why? I am trying to determine what parts of the country would be ideal for me to live in the future and any help would be greatly appreciated. |
5fee7960-8b07-4a65-b349-6beee1ad12de | trentmkelly/LessWrong-43k | LessWrong | Supervising strong learners
by amplifying weak experts
Abstract
> Many real world learning tasks involve complex or hard-to-specify objectives, and using an easier-to-specify proxy can lead to poor performance or misaligned behavior. One solution is to have humans provide a training signal by demonstrating or judging... |
d5b58a30-4e5c-4314-ad71-bc247401f0a6 | trentmkelly/LessWrong-43k | LessWrong | Using Doulingo 'Immersion' feature to translate HPMOR
I have been using (alongside other learning tools) Duolingo for more than a year now and with considerable success in learning Spanish language.
I used immersion for learning purposes, but just recently it dawned on me that I could use it to translate HPMOR into d... |
74b83d21-b787-48f6-b5de-5aafcaba94d6 | trentmkelly/LessWrong-43k | LessWrong | Business Writing Example #1
One of my friends is a saleswoman in Shenzhen who sells lead-acid batteries. She sends out lots of emails and gets few responses. When she does get a response, she often hands it off to someone else in her company who is better at closing sales.
Here is a letter my friend sent to a custome... |
b6f5f2d8-4013-41ef-83c5-7f19545fe264 | trentmkelly/LessWrong-43k | LessWrong | "Dialectics"?
Hi, everyone.
I just started reading Total Freedom by Chris Sciabarra (warning: politics book), and a good half of it seems to be about 'dialectics' as a thinking tool, but it's been total rubbish in trying to explain it. From poking around on the internet, it seems to have been a proto-systems theory t... |
840c5b10-35ef-4684-a1e6-67c3e6383170 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | How does OpenAI's language model affect our AI timeline estimates?
OpenAI [recently announced](https://blog.openai.com/better-language-models/) progress in NLP, using a large transformer-based language model to tackle a variety of tasks and breaking performance records in many of them. It also generates synthetic shor... |
d2891180-4365-45c7-8e84-576f5908c5da | trentmkelly/LessWrong-43k | LessWrong | On the abolition of man
(Cross-posted from my website. Podcast version here, or search for "Joe Carlsmith Audio" on your podcast app.
This essay is part of a series that I'm calling "Otherness and control in the age of AGI." I'm hoping that the individual essay can be read fairly well on their own, but see here for b... |
6178f71f-370f-40d1-b28a-9d1316d94825 | trentmkelly/LessWrong-43k | LessWrong | Subjective anticipation as a decision process
As argued here, debates about probability can be profitably replaced with decision problems. This often dissolves the debate - there is far more agreement as to what decision sleeping beauty should take than on what probabilities she should use.
The concept of subjective ... |
85f0dc5b-e93a-4b16-b0a6-4d57a12f861a | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | We Did AGISF’s 8-week Course in 3 Days. Here’s How it Went
Introduction
------------
**TL;DR:** It went well. We skimmed the readings with about a 2:1 ratio of reading time to discussion time, which didn’t allow us to understand everything deeply but generated lively discussions. All participants said they got what... |
9d0605e7-0762-4109-bdbb-be0763448b8d | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Efficacy of AI Activism: Have We Ever Said No?
Disclaimer:
===========
This piece came out of a Summer Research Project with the Existential Risk Alliance. This section was intended to be part of an introduction for an academic article, which mutated into a stand-alone piece. My methodology was shallow dives (2-4 hou... |
e943fe8e-5809-4fbc-9292-67d1d28679fd | trentmkelly/LessWrong-43k | LessWrong | Confidence all the way up
I apparently possess some sort of aura of competence. Some say I'm confident, others say I'm arrogant, others remark on how I seem very certain of myself (which I have been told both as compliment and critique).
I was surprised, at first, by these remarks from friends and family — from my pe... |
c638dd69-6413-49ff-a449-cd77ba43d9c5 | trentmkelly/LessWrong-43k | LessWrong | On Creativity - The joys of 5 minute timers
Introduction
A really common failure mode that I observe in myself and others is to get stuck and then give up. To be working on a problem, or trying to generate ideas, find it hard and stop looking. I find that this goes hand-in-hand with the far larger problem of being tr... |
c9ee7f11-db6e-4097-8655-cbf85efc4cb2 | trentmkelly/LessWrong-43k | LessWrong | Partial summary of debate with Benquo and Jessicata [pt 1]
Note: I'll be trying not to engage too much with the object level discussion here – I think my marginal time on this topic is better spent thinking and writing longform thoughts. See this comment.
Over the past couple months there was some extended discussion... |
03b170af-b293-4779-966b-35902ae285f4 | trentmkelly/LessWrong-43k | LessWrong | Survival and Flourishing Fund Applications closing in 3 days
The grant round we announced a month ago for the new Survival and Flourishing Fund is closing in 3 days. We haven't gotten that many applications, so I would recommend applying.
https://www.lesswrong.com/posts/CcAnPPKmPmbXxtWME/survival-and-flourishing-fund... |
116ab730-f7af-4b1b-b582-029317c730c3 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Another Frankfurt meetup
Discussion article for the meetup : Another Frankfurt meetup
WHEN: 22 February 2015 02:00:00PM (+0100)
WHERE: Heerstraße, 60488 Frankfurt
I'm posting this for weird_random who unfortunately can't post in the meetup section yet:
"Hi!
We will have a next LessWrong meetup in Februar... |
65f280aa-c82c-4d4c-b1e8-5f50df262d5a | trentmkelly/LessWrong-43k | LessWrong | [Personal Question] Can anyone help me navigate this potentially painful interpersonal dynamic rationally?
Hello, LessWrong community. I have a question, my dears, for you. I hope it will not annoy anyone as it is admittedly a self-indulgent question which perhaps does not fit very well within the landscape of LessWr... |
0ed735bb-fdc4-411d-bab1-e50eebc04769 | StampyAI/alignment-research-dataset/arxiv | Arxiv | Natural Adversarial Examples.
1 Introduction
---------------
Research on the ImageNet benchmark has led to numerous advances in classification (AlexNet), object detection (Huang2017SpeedAccuracyTF), segmentation (He2018MaskR), and perceptual metrics (zhang2018perceptual). ImageNet classification improvements have a... |
045849cd-9ee3-4a9a-8d54-9483baa97037 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Washington, D.C.: Reading Book Excerpts
Discussion article for the meetup : Washington, D.C.: Reading Book Excerpts
WHEN: 17 July 2016 03:30:00PM (-0400)
WHERE: Donald W. Reynolds Center for American Art and Portraiture
We'll be meeting to read out and discuss excerpts from books.
As usual, side conversat... |
30b03f6f-bff7-45cf-ae97-68441724659a | StampyAI/alignment-research-dataset/blogs | Blogs | Progress on Causal Influence Diagrams
[](/?source=post_page-----a7a32180b0d1--------------------------------)Listen
Share
*By Tom Everitt, Ryan Carey, Lewis Hammond, James Fox, Eric Langlois, and Shane Legg*
*Cross... |
9ecc3c46-a1ea-4f02-9ed1-4e130aaf4f3a | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | What are people's thoughts on working for DeepMind as a general software engineer?
This opportunity: <https://boards.greenhouse.io/deepmind/jobs/4445689> matches my skillset. A recruiter has reached out. What would the value of working on helping DeepMind as a whole be? Any thoughts? Thanks. |
ec48407f-e1e6-40ba-a74c-dd28d51483b0 | trentmkelly/LessWrong-43k | LessWrong | The "Everyone Can't Be Wrong" Prior causes AI risk denial but helped prehistoric people
The "Everyone Can't Be Wrong" Prior, assumes that if everyone is working on something, that thing is important. Conversely, if nobody is working on something, that thing is unimportant.
Why
In prehistoric times, this is actually ... |
66a91f71-7979-4c4b-ba1e-7c95a5e4c30d | trentmkelly/LessWrong-43k | LessWrong | LessWrong Coronavirus Link Database
Note: This post is old, read the new major update here.
There is a lot of coronavirus information currently out there, and a lot of people have been feeling overwhelmed keeping track of all the things coming in from different sources. To combat that, the LessWrong team is maintaini... |
58b6f67b-06aa-4a5b-a758-1b4eea775c31 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Washington, D.C.: Fun & Games
Discussion article for the meetup : Washington, D.C.: Fun & Games
WHEN: 15 May 2016 03:30:00PM (-0400)
WHERE: Reynolds Center
We will be meeting in the courtyard to play games, hang out, and engage in fun conversation.
Upcoming meetups:
* May 22: Typical Mind Fallacy
* May... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.