id stringlengths 36 36 | source stringclasses 15
values | formatted_source stringclasses 13
values | text stringlengths 2 7.55M |
|---|---|---|---|
18edd67b-d71a-45bc-9eee-f602d6242dfe | StampyAI/alignment-research-dataset/youtube | Youtube Transcripts | 276. Universal and Transferable adversarial attacks on aligned language models
hello and welcome to session 276 in the
aicity.com reading group tonight we will
be discussing the article Universal and
transferable adversarial attacks on
aligned language models by ended so and
others
and it's all is from Carnegie melon
... |
c014efa8-4cd2-4a8e-a1e4-97d4b33608c6 | trentmkelly/LessWrong-43k | LessWrong | Place-Based Programming - Part 1 - Places
When I write machine learning software, I tend to use the Place-Based Programming (PBP) paradigm. PBP caches your computations so you rarely have to perform the same computation twice.
The fundamental unit of data is a place, which refers to a location on disk. Consider the h... |
3d18a772-5f23-4115-ab9d-2f5fead50ff3 | trentmkelly/LessWrong-43k | LessWrong | Superintelligence 16: Tool AIs
This is part of a weekly reading group on Nick Bostrom's book, Superintelligence. For more information about the group, and an index of posts so far see the announcement post. For the schedule of future topics, see MIRI's reading guide.
----------------------------------------
Welcome.... |
60958b87-8def-4af0-b402-864f0fd7f9aa | trentmkelly/LessWrong-43k | LessWrong | Ritual Report 2012: Life, Death, Light, Darkness, and Love.
One winter ago, twenty aspiring rationalists gathered in a room, ate some food, sang some songs, and lit some candles. We told some stories about why the universe is the way it is, and what kind of people we want to be.
I wrote some things about the experien... |
e59f5f9a-2ae8-40ff-af4f-031265162145 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Back to the Past to the Future
You are sitting in your bedroom, dorm, tent, sidewalk, wherever you are staying right now, reading these words, when a dark void appears in your periphera... |
140ca92b-6fed-4d14-96c0-972366f0c35c | trentmkelly/LessWrong-43k | LessWrong | Security Mindset and Ordinary Paranoia
Follow-up to: AI Alignment: Why It’s Hard, and Where to Start
----------------------------------------
(amber, a philanthropist interested in a more reliable Internet, and coral, a computer security professional, are at a conference hotel together discussing what Coral insis... |
5290ea1f-34d7-43bb-b83d-951a29c3b1f0 | trentmkelly/LessWrong-43k | LessWrong | Putting in the Numbers
Followup To: Foundations of Probability
In the previous post, we reviewed reasons why having probabilities is a good idea. These foundations defined probabilities as numbers following certain rules, like the product rule and the rule that mutually exclusive probabilities sum to 1 at most. These... |
63a9dae7-9170-429a-9c7c-ee6e22d5daea | trentmkelly/LessWrong-43k | LessWrong | Approaching Logical Probability
Followup To: Logic as Probability
If we design a robot that acts as if it's uncertain about mathematical statements, that violates some desiderata for probability. But realistic robots cannot prove all theorems; they have to be uncertain about hard math problems.
In the name of practi... |
f75bd5bf-26a6-44a6-a8cd-02cf36fe8101 | trentmkelly/LessWrong-43k | LessWrong | Progress links & tweets, 2022-09-08
Opportunities
* Nat Friedman and Daniel Gross are investing $10M in AI-first products, apply here (via @natfriedman)
* Congressman Jake Auschincloss is hiring a Chief of Staff, @MarkLutter encourages the progress community to apply
Announcements
* Superabundance: The Story of P... |
a05369de-7f0d-474f-9723-720d5f030adb | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Being an individual alignment grantmaker
I am an earlyish crypto investor who has accumulated enough to be a mid-sized grantmaker, and I intend to donate most of my money over the next 5-10 years to try and increase the chances that humanity has a wonderful future. My best guess is that this is mostly decided by wheth... |
c8d161cf-6f36-483f-8969-477fdfaac386 | trentmkelly/LessWrong-43k | LessWrong | When should you defer to expertise? A useful heuristic (Crosspost from EA forum)
One important aspect of our lives as we search for knowledge is knowing what and who to defer to, as we usually must take a lot of our knowledge on faith in expertise. However, how should you defer on issues, given that certain areas coul... |
d7380613-52e9-41bd-a41c-03b3c1e70146 | trentmkelly/LessWrong-43k | LessWrong | Principles for Productive Group Meetings
Note: This post is based on a Google document I created for my research group. It speaks in the first person, but I think the lessons could be helpful for many research groups, so I decided to share it more broadly. Thanks to Louise Verkin for converting from Google doc to Mark... |
12d8ff3e-59fa-4c76-81f2-98b1c95bb6e7 | trentmkelly/LessWrong-43k | LessWrong | Why write more: improve your epistemics, self-care, & 28 other reasons
Writing is like exercise. It’s just good on so many levels.
* Writing can be high impact
* Writing can be inspiring
* Writing can help you learn and grow
* Writing can help your career
* Writing can be meaningful and deeply satisfying
* Wr... |
1b143ed5-4a43-408e-a689-17b22a152346 | StampyAI/alignment-research-dataset/arxiv | Arxiv | Do the Rewards Justify the Means? Measuring Trade-Offs Between
Rewards and Ethical Behavior in the Machiavelli Benchmark
1 Introduction
---------------
AI systems are rapidly gaining capabilities (OpenAI, [2023](#bib.bib53)), especially in natural language (Bubeck et al., [2023](#bib.bib13)). To mitigate risks of ... |
1451c35b-5c07-4fbb-8ab8-7ace1421c5f0 | StampyAI/alignment-research-dataset/special_docs | Other | Complications in evaluating neglectedness
Neglectedness (or crowdedness) is a heuristic that effective altruists use to assess how much impact they could have in a specific cause area. It is usually combined with scale (a.k.a. importance) and tractability (a.k.a. solvability), which together are meant to approximate e... |
bf3b5442-c6cc-4605-9409-91cc62bb62b5 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | When people ask for your P(doom), do you give them your inside view or your betting odds?
There has been some confusion about whether people are using inside views or all-things-considered betting odds when they talk about P(doom). Which do you give by default? What are your numbers for each? |
2a4315a9-6f08-436d-ba5a-16d74f2becd3 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Proposed Orthogonality Theses #2-5
### 2
The intelligence, reflectivity, and 'Lawful'ness of a mind are mostly independent of its capacity to experience emotion. Minds can be very unintelligent, lack anything like a "utility function" or some critical OODA loop components, possess little or no self awareness, and gen... |
dd81f341-952e-4850-b80a-408d85c30860 | trentmkelly/LessWrong-43k | LessWrong | Optimizing arbitrary expressions with a linear number of queries to a Logical Induction Oracle (Cartoon Guide)
This is Logical Induction, or LI for short.
We can ask LI any maths question that has a number as an answer. If we ask it a question that is too tricky, it will have a good guess. On average, LI is neither t... |
5396143e-6015-4ca1-961e-8d7336d92da6 | StampyAI/alignment-research-dataset/youtube | Youtube Transcripts | 244. Democratising Risk 1
hello and welcome to session two hundred
and forty four in the ascii.com reading
group
tonight we'll be discussing the first
part of the article democratizing risk
in search of a methodology to study
existential risk by carla so creamer and
luke kim
carla so creamer works at the future of
hum... |
ec93e4a4-f197-44f9-aa9d-d48c73541760 | trentmkelly/LessWrong-43k | LessWrong | Capturing Uncertainty in Prediction Markets
I've been trying to make sense of what a 50% prediction means. Does it convey different meanings based on the question asked, and is it really as uninformative as it's made out to be. As a part of that discussion, I believe there is a possible solution to two big problems pr... |
69cee380-cd0a-4c6d-8c6a-d7f6aa1bce79 | trentmkelly/LessWrong-43k | LessWrong | Real-Time Gigstats
For a while ( 2014, 2015, 2016, 2017, 2018, 2019, 2023, 2024) I've been counting how often various contra bands and callers are being booked for larger [1] events. Initially, I would run some scripts, typically starting from scratch each time because I didn't remember what I did last time, but after... |
ef93e834-d13a-44d2-8e41-d2888d6f217d | trentmkelly/LessWrong-43k | LessWrong | There is no try
Ok, so "try" is actually a pretty useful concept; there's a reason we have a very short word for it in the English language. Nevertheless, I have found it quite useful to occasionally spend a few weeks refusing to use the word "try" or any of its synonyms, at least when talking about myself, and especi... |
9db4626a-bf28-402e-8589-320e53fe3685 | trentmkelly/LessWrong-43k | LessWrong | Is g a measure of ability to absorb information in a non-inductive way?
Eliezer and Robin discussed g somewhat in their debate. I think this question is one that we can do some more research on ourselves. The current hypothesis I'm exploring is that g measures the ability to take in information non-inductively this i... |
b2c67754-359a-4f35-ace5-58e3dac9cb53 | trentmkelly/LessWrong-43k | LessWrong | BatchTopK: A Simple Improvement for TopK-SAEs
Work done in Neel Nanda’s stream of MATS 6.0.
Epistemic status: Tried this on a single sweep and seems to work well, but it might definitely be a fluke of something particular to our implementation or experimental set-up. As there are also some theoretical reasons to expe... |
92d6b0b3-cb86-492c-bf12-1168259e29c8 | trentmkelly/LessWrong-43k | LessWrong | Slack for your belief system
Follow-up to Zvi’s post on Slack
You can have Slack in your life. But you can also have Slack in your belief system.
Initially, this seems like it might be bad.
Won't Slack result in a lack of precision? If I give myself Slack to believe in whatever, won't I just end up with a lot of wr... |
10b794ad-db43-4624-9c70-197b5f045ff7 | trentmkelly/LessWrong-43k | LessWrong | Why Academia is Mostly Not Truth-Seeking
The PDF version can be read here.
In this essay, I will argue that most academic research is fake. The modern academy is not a reliable source of knowledge. Instead, it produces the pretense of knowledge.
Academic research can be fake in different ways. It can simply be false... |
bd44f866-fd27-44b5-96d0-0d4bb3bacacc | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Announcing an Empirical AI Safety Program
Last summer, the [Center for AI Safety](https://safe.ai/) ran the [ML Safety Scholars program (MLSS)](https://forum.effectivealtruism.org/posts/9RYvJu2iNJMXgWCBn/introducing-the-ml-safety-scholars-program) to help students interested in AI Safety gain technical skills and lear... |
ba0a40f8-7833-4445-b10e-2cf6f90bafd1 | trentmkelly/LessWrong-43k | LessWrong | Limits of Superintelligence
Surprisingly, I haven’t seen anything written about the ultimate limits of intelligence. This is a short post where I try to ponder this question.
Superintelligent agents are often portrayed as extremely powerful oracles capable of predicting long-term outcomes of actions and thus steering... |
8348bc37-1ec7-4e18-92ef-c292736287a0 | trentmkelly/LessWrong-43k | LessWrong | If I have some money, whom should I donate it to in order to reduce expected P(doom) the most?
I just want to make sure that when I donate money to AI alignment stuff it's actually going to be used economically |
2967a129-9976-4113-8b58-d7aa2d31823e | trentmkelly/LessWrong-43k | LessWrong | Low-Friction MBTA Predictions
This has not been a good week for the MBTA:
* The Red, Orange, and Blue lines are running Saturday schedules on weekdays (news coverage) until they are able to hire and train more dispatchers.
* The Orange and Green lines (again) were suspended after problems with the ongoing demoliti... |
7b05e0e1-b3e3-4b6c-954c-33788d3a3555 | trentmkelly/LessWrong-43k | LessWrong | How to get cryocrastinators to actually sign up for cryonics
At the end of CFAR's July Rationality Minicamp, we had a party with people from the LW/SIAI/CFAR community in the San Francisco Bay area. During this party, I had a conversation with the girlfriend of a participant in a previous minicamp, who was not signed ... |
6a7f58e1-f91a-4683-94c1-5c11d28fb355 | trentmkelly/LessWrong-43k | LessWrong | [AN #75]: Solving Atari and Go with learned game models, and thoughts from a MIRI employee
Find all Alignment Newsletter resources here. In particular, you can sign up, or look through this spreadsheet of all summaries that have ever been in the newsletter. I'm always happy to hear feedback; you can send it to me by r... |
e5bb2693-82e3-4894-80de-3951a1bd8bd2 | trentmkelly/LessWrong-43k | LessWrong | Framing approaches to alignment and the hard problem of AI cognition
(Thanks to Adam Shimi and Steven Byrnes for feedback.)
(Edit: unsurprisingly, my views have updated some after writing this post. Specifically, I now put much higher credence on solving inner alignment via verification or relaxed adversarial trainin... |
524f8903-2b94-40a2-9e4a-ee98048876e4 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Consistencies as (meta-)preferences
>
> Do I contradict myself?
> Very well then I contradict myself,
> (I am large, I contain multitudes.)
>
>
>
>
> Walt Whitman
>
>
>
It is good for our decision processes to be [time-consistent](https://en.wikipedia.org/wiki/Hyperbolic_discounting), [transitive, and ind... |
90536cb4-2072-4b4a-8e52-149cea77699b | trentmkelly/LessWrong-43k | LessWrong | Over $1,000,000 in prizes for COVID-19 work from Emergent Ventures
Tyler Cowen's fund Emergent Ventures is offering:
* $500k, second prize $200k, for Best effort to find a good treatment rapidly
* $100k for Best blog or social media tracking/analysis of the virus
* $100k for Best innovation in social distancing
*... |
7a01c1a5-e3ad-487e-80d9-1a798767eff1 | trentmkelly/LessWrong-43k | LessWrong | Seeking descriptions of deciban-levels
As part of a small project I'm working on, I need to have at least a rough description of how a given number of decibans translates into a subjective level of confidence, described in a way that can be understood by people who've never come across the idea before.
Some previous ... |
748c02db-29d1-428c-a6f7-87fd47bd22fd | trentmkelly/LessWrong-43k | LessWrong | Memory Decoding Journal Club: Structure and function of the hippocampal CA3 module
Join Us for the Memory Decoding Journal Club!
A collaboration of the Carboncopies Foundation and BPF Aspirational Neuroscience
This time, we’re diving into a groundbreaking paper:
"Structure and function of the hippocampal CA3 module"... |
1f50fc33-6f58-4ace-85fe-a87da7e72271 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Draft: Detecting optimization
**This post is a draft, a work in progress.** Much of it will not make sense, and will not be optimized for the reader's experience. It does not necessarily reflect my current best understanding of optimization. Like the introduction draft, I'm opening it up because I've been working on t... |
e7ac42d7-77da-4238-8036-575856017511 | trentmkelly/LessWrong-43k | LessWrong | What’s the likelihood of only sub exponential growth for AGI?
A possible scenario could be that the first confirmed AGI(s) are completely unimpressive, i.e. with equivalent or less capacities than an average 10 year child. And then tremendous effort is put into ’growing’ its capacities with the best results yielding ... |
dbe5ee57-d54f-4c9f-9d70-69a3e0ffacd3 | StampyAI/alignment-research-dataset/arxiv | Arxiv | Autonomous Intelligent Cyber-defense Agent (AICA) Reference Architecture. Release 2.0
ARL-SR-0421 ● SEP 2019
Autonomous Intelligent Cyber -defense Agent
(AICA) Reference Architecture
Release 2.0
by Alexander Kott, Paul Théron, Martin Drašar, Edlira Dushku,
Benoît LeBlanc, Paul Losiewicz, ... |
3d2d46bb-39b4-4e7e-8104-4c42c7e32dd4 | trentmkelly/LessWrong-43k | LessWrong | electric turbofans
It seems I've been typecast as a person to ask about "hard tech startups" with big claims that are getting attention. This time it's been Astro Mechanica, a Y Combinator startup making aircraft engines. Here's a video, possibly set up by a PR agency. Apparently Paul Graham compared the significance ... |
f0f9da07-0e62-479c-bdb2-c61ce71fb6ae | trentmkelly/LessWrong-43k | LessWrong | Meetup : Melbourne Winter Solstice Dinner Party and Social Meetup
Discussion article for the meetup : Melbourne Winter Solstice Dinner Party and Social Meetup
WHEN: 21 June 2013 06:30:00PM (+1000)
WHERE: Carlton, Victoria, Australia
Melbourne's next regular social meetup comes with a twist. As we are holding the ev... |
71f6c2ee-8193-45f9-b50d-06a1f0cfce2e | trentmkelly/LessWrong-43k | LessWrong | Trapped Priors As A Basic Problem Of Rationality
Crossposted from Astral Codex Ten
Introduction and review
Last month I talked about van der Bergh et al’s work on the precision of sensory evidence, which introduced the idea of a trapped prior. I think this concept has far-reaching implications for the rationalist pr... |
8a9ccfbf-3509-4fba-9f69-e93a4672971c | trentmkelly/LessWrong-43k | LessWrong | Avoid interruptions by time-shifting them
A lot of times you may come upon a juicy-smelling "information scent". Now you have a decision of whether to extend your lunch break and cut into work by following the scent (bad) or abandoning it (painful, and possibly bad).
One thing that has worked for me is using either d... |
4daad83a-35fe-494c-aa59-f8da98e4d670 | trentmkelly/LessWrong-43k | LessWrong | "Measuring the distribution of spitefulness" - link
http://www.plosone.org/article/info:doi/10.1371/journal.pone.0041812
Here is a rather curious paper describing psychology researchers' attempts to investigate "spitefulness" - I think they define spitefulness roughly as "hurting others without any benefit to oneself... |
c8ba1417-4872-4ae7-8028-f95628fcc2be | StampyAI/alignment-research-dataset/lesswrong | LessWrong | [ASoT] Instrumental convergence is useful
A central concept I got out of [Reframing impact](https://www.alignmentforum.org/posts/xCxeBSHqMEaP3jDvY/reframing-impact) is that instrumental convergence can be useful for shaping the motivations of superintelligent agents. i.e. there are two frames for thinking about instru... |
f7bc506e-bd00-4378-9d04-c8b21005224c | trentmkelly/LessWrong-43k | LessWrong | Meetup : Houston Meetup
Discussion article for the meetup : Houston Meetup
WHEN: 30 October 2011 02:00:00PM (-0500)
WHERE: 2010 Commerce St, Houston, Tx. 77002
After a temporary hiatus, the Houston LW meetup group will reconvene on 10/30 at 2:00 PM. I will be presenting a paper from within my field of research, the... |
1ab6ecd7-d849-4c2a-bbab-9f2488997a39 | trentmkelly/LessWrong-43k | LessWrong | Have any parties in the current European Parliamentary Election made public statements on AI?
The European parliamentary election is currently taking place (voting period: 2024-04-07 to 2024-07-10). While I assume my vote[1] has ~no impact on x-risk either way, I'd nonetheless like to take into account if parties have... |
1b49904f-f008-47ab-9fab-c23a17350439 | trentmkelly/LessWrong-43k | LessWrong | Gauging interest for a Zurich, Switzerland meetup group
Who would be interested in meeting? Please reply to this thread or msg me in private.
Thanks!
Edit: There doesnt seem to be enough people willing to meet so Im not planning anything at the moment. But if you happen to be in Zurich and want to meet a fellow LWer... |
2c89acf0-28ec-44af-bd66-d8b9d8fd1f39 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Is transformative AI the biggest existential risk? Why or why not?
I previously asked this question [in an AMA](https://forum.effectivealtruism.org/posts/tCmPDx7bFDYkmpAY7/?commentId=aS2MwP3dBiF38euTa), but I'd like to pose it to the broader EA Forum community:
> Do you believe that AGI poses a greater existential r... |
8e2bb8fd-ac17-4ab6-9dc3-6c6d54e684a9 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | What You See Isn't Always What You Want
.mjx-chtml {display: inline-block; line-height: 0; text-indent: 0; text-align: left; text-transform: none; font-style: normal; font-weight: normal; font-size: 100%; font-size-adjust: none; letter-spacing: normal; word-wrap: normal; word-spacing: normal; white-space: nowrap; floa... |
6fa0b4b3-9908-4ab6-958f-a69fcc102e15 | trentmkelly/LessWrong-43k | LessWrong | Art: A Rationalist's Take?
Hello LWers,
I am in my 20s, and I like to collect art.
Admittedly, I am a lurker on the LW forums and can proudly say I have been for a very long time now. I've refrained from commenting simply because I have not felt that I have all that much to contribute, but rather, find that I have m... |
a7ea9a87-c0bd-4a5f-9264-27b8e78e4ecf | trentmkelly/LessWrong-43k | LessWrong | Does polyamory at a workplace turn nepotism up to eleven?
I am reading debates on the Effective Altruism forum, and they seem to focus on how a low-status person might be afraid to refuse to consent to sex with a high-status person, because the high-status person might then refuse to provide them a grant or a job posi... |
480260ea-309e-40e6-bba5-a601e145cb79 | trentmkelly/LessWrong-43k | LessWrong | OpenAI Preparedness Framework 2.0
Right before releasing o3, OpenAI updated its Preparedness Framework to 2.0.
I previously wrote an analysis of the Preparedness Framework 1.0. I still stand by essentially everything I wrote in that analysis, which I reread to prepare before reading the 2.0 framework. If you want to ... |
015f2e3f-0cb3-45a7-ab04-048865e8991a | StampyAI/alignment-research-dataset/arbital | Arbital | Theory of (advanced) agents
Many issues in AI alignment have dependencies on what we think we can factually say about the general design space of cognitively powerful agents, or on which background assumptions yield which implications about advanced agents. E.g., the [Orthogonality Thesis](https://arbital.com/p/1y) i... |
dab9c182-3a7c-4960-a0e2-1ad5401c3cd5 | StampyAI/alignment-research-dataset/agisf | AGI Safety Fund | A Tour of Emerging Cryptographic Technologies
A Tour of Emerging
Cryptographic
Technologies
What They Are and
How They Could Matter
May 2021
Ben Garfinkel
Centre for the Governance of AI,
Future of Humanity Institute,
University of Oxford
Abstract
Historically, progress in the /fi.ligaeld of cryptography has ... |
1797da8c-9ae3-489d-ae51-9094e6fd330f | StampyAI/alignment-research-dataset/youtube | Youtube Transcripts | Robin Hanson on AI Takeoff Scenarios - AI Go Foom?
welcome over the last over the last many
years the northern have actually
recorded a number of interviews but
today we're going to be exploring AI
kpop scenarios with reference to a
previous debate an AI Foom debate with
le AZ and calcium plum that happened
many years... |
02d23e47-9a7d-4651-b708-22a1f76062d0 | trentmkelly/LessWrong-43k | LessWrong | The Neglected Virtue of Scholarship
Eliezer Yudkowsky identifies scholarship as one of the Twelve Virtues of Rationality:
> Study many sciences and absorb their power as your own. Each field that you consume makes you larger... It is especially important to eat math and science which impinges upon rationality: Evolut... |
6653c4f7-8824-411d-bbaf-dfd53288a007 | trentmkelly/LessWrong-43k | LessWrong | AI safety advocates should consider providing gentle pushback following the events at OpenAI
AI safety advocates for the most part have taken an exclusively–and potentially excessively–friendly and cooperative approach with AI firms–and especially OpenAI. I am as guilty of this as anyone[1]–but after the OpenAI disast... |
ad113562-e932-4b4e-b842-b9f94286015d | trentmkelly/LessWrong-43k | LessWrong | Key lime pie and the methods of rationality
None |
739393ed-e880-4905-b046-cfdb9cb2be44 | StampyAI/alignment-research-dataset/blogs | Blogs | Soft takeoff can still lead to decisive strategic advantage
*By Daniel Kokotajlo, 11 September 2019*
*Crossposted from the*[*AI Alignment Forum*](https://alignmentforum.org/posts/PKy8NuNPknenkDY74/soft-takeoff-can-still-lead-to-decisive-strategic-advantage)*. May contain more technical jargon than usual.*
*[Episte... |
27d164cb-3ad7-48f9-92bc-2b8572b1f004 | StampyAI/alignment-research-dataset/arbital | Arbital | Value
### Introduction
In the context of [value alignment](https://arbital.com/p/2v) as a subject, the word 'value' is a speaker-dependent variable that indicates our ultimate goal - the property or meta-property that the speaker wants or 'should want' to see in the final outcome of Earth-originating intelligent life... |
f56714ca-5187-40cd-ade7-227ca6ad32ca | trentmkelly/LessWrong-43k | LessWrong | AI Safety Newsletter #8: Rogue AIs, how to screen for AI risks, and grants for research on democratic governance of AI
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 versions.
-----------... |
0e872b83-3ec1-40d4-ad15-1d512fc6363d | LDJnr/LessWrong-Amplify-Instruct | LessWrong | "As the question asks, I am looking for productivity advice directly or indirectly related to one's menstrual cycle. Motivation: This may seem like a gross or weirdly personal question but I think it's actually quite important. For example, for the first four hours of my period, I am unable to do anything besides lie d... |
c0ab4476-417e-46fd-acc9-2d222181299a | trentmkelly/LessWrong-43k | LessWrong | What are the gears of gluten sensitivity?
I've been having some weird health issues somehow triggered after eating. I've been working with a nutritionist. They recommend, among other things, trying to eliminate gluten some of the time on the idea that gluten puts stress on the digestive system and can expose other iss... |
5ce343fc-66ee-4f14-b227-9e471ffaf4fe | trentmkelly/LessWrong-43k | LessWrong | Discussion on utilizing AI for alignment
Summarized/re-worded from a discussion I had with John Wentworth. John notes that everything he said was off-the-cuff, and he doesn’t always endorse such things on reflection.
Eli: I have a question about your post on Godzilla Strategies. For background, I disagree with it; I’... |
654ee670-1eb2-4240-bb48-77a86b335db7 | trentmkelly/LessWrong-43k | LessWrong | The recent banality of rationality (and effective altruism)
[Neurotic status: Until recently, I've felt entirely comfortable telling friends, family, and co-workers that I'm a rationalist (or an aspiring rationalist)—that it wouldn't out-group me among them, or if it did, it was probably worth being out-grouped over. ... |
8c9f28f0-bf85-4c7f-8d9a-a091ea50a195 | trentmkelly/LessWrong-43k | LessWrong | Cognitive Empathy and Emotional Labor
The concept of emotional labor has been popularized in the last couple years as a way of talking about the work people do to manage other people’s emotions. Most of that discussion has been around how women are often expected to perform emotional labor without compensation both pr... |
51a2f606-0d38-46a6-894f-e1bd33397b2a | trentmkelly/LessWrong-43k | LessWrong | Values vs. parameters
I've written before about the difficulty of distinguishing values from errors, from algorithms, and from context. Now I have to add to that list: How can we distinguish our utility function from the parameters we use to apply it?
In my recent discussion post, "Rationalists don't care about the... |
644dc2c1-9066-470a-b528-23778b72266b | trentmkelly/LessWrong-43k | LessWrong | Organisations working on multiple Global Catastrophic risks
It is not uncommon to find organisations working, directly or indirectly, on a single Global Catastrophic Risk (GCR). For instance, the World Health Organization does much work to prevent pandemics, as part of its remit.
It is rarer for organisations to focu... |
67eab7c5-0489-49a3-9e7c-f371f7d40ec6 | awestover/filtering-for-misalignment | Redwood Research: Alek's Filtering Results | id: post680
We’ve released a paper, AI Control: Improving Safety Despite Intentional Subversion . This paper explores techniques that prevent AI catastrophes even if AI instances are colluding to subvert the safety techniques. In this post: We summarize the paper; We compare our methodology to the methodology of other... |
4f77454f-0047-44aa-a044-93e633e43789 | trentmkelly/LessWrong-43k | LessWrong | A ranking scale for how severe the side effects of solutions to AI x-risk are
Sometimes when I look at solutions to AI risk (a.k.a. pivotal acts), they seem like they are "cheating". For example, the only way I could see CAIS preventing AI risk is for it to be used against other AI researchers. This is as opposed to a... |
62320cbc-e507-48ab-a52e-c9112ca916cb | trentmkelly/LessWrong-43k | LessWrong | [SEQ RERUN] Human Evil and Muddled Thinking
Today's post, Human Evil and Muddled Thinking was originally published on 13 September 2007. A summary (taken from the LW wiki):
> It's easy to think that rationality and seeking truth is an intellectual exercise, but this ignores the lessons of history. Cognitive biases... |
5f1b9ca7-5876-42cb-b81c-47b5866465f7 | trentmkelly/LessWrong-43k | LessWrong | Longevity interventions when young
Preface for less wrong
**Skip if you just want to read the article**
I'm not 100% sure this article is in tune with LW, I usually share articles related to conceptual analysis or bashing on classical mathematics and AI risk here, those fit.
This, in theory, should also fit, since l... |
7b036a73-d2bc-4afb-ad8f-85daf1df7ebf | trentmkelly/LessWrong-43k | LessWrong | Spooky Recommendation System Scaling
Crosspost of https://phoropter.substack.com/p/spooky-recommendation-system-scaling
Today I wish to focus on one reason that recommendation systems should be regulated: the benefits of scaling that famously apply to language models also apply to recommendation systems. While applic... |
394d4103-4143-40cc-8cfb-535404b36ca5 | StampyAI/alignment-research-dataset/arbital | Arbital | Value achievement dilemma
The value achievement dilemma is a way of framing the [AI alignment problem](https://arbital.com/p/2v) in a larger context. This emphasizes that there might be possible solutions besides AI; and also emphasizes that such solutions must meet a high bar of potency or efficacy in order to *reso... |
59bf93b6-2cba-4394-a369-857db50597ee | trentmkelly/LessWrong-43k | LessWrong | Fermi Estimating How Long an Algorithm Takes
This is a "live" Fermi estimate, where I expect to make mistakes and haven't done much editing.
If you're working on a simple mathematical program, you can attempt to estimate the number of floating point or integer operations, compare that to statistics on CPU latencies/t... |
1fc02987-c474-4244-893f-b77197f35fe1 | trentmkelly/LessWrong-43k | LessWrong | Energy Markets Temporal Arbitrage with Batteries
Epistemic Status: I am not an energy expert, and this was done rather briefly. All analysis uses pricing data specific to Ireland, but some general ideas are likely applicable more broadly. Data is true as of March 2025. Where there are uncertainties I try to state them... |
ceb43d11-e81b-497b-90b5-2f875ae7bad7 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Paderborn Meetup Frebruary 6th
Discussion article for the meetup : Paderborn Meetup Frebruary 6th
WHEN: 06 February 2013 07:00:00PM (+0100)
WHERE: Gownsmen's Pub, Uni Paderborn, Warburger Straße 100, Paderborn
We are meeting once again in Paderborn.
The topics of this evening will be the current series "... |
9625bcae-e55a-4ca3-b556-13a4c11f1861 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Utility ≠ Reward
This essay is an adaptation of a talk I gave at the [Human-Aligned AI Summer School 2019](http://humanaligned.ai/index-2019.html) about [our work on mesa-optimisation](https://www.alignmentforum.org/s/r9tYkB2a8Fp4DN8yB). My goal here is to write an informal, accessible and intuitive introduction to th... |
bdac4a6e-b523-433c-bd8e-11f93eb26517 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | World and Mind in Artificial Intelligence: arguments against the AI pause
After my previous post on the marginal existential risk from AI in world where nuclear war is likely, here I review some canonical AI risk literature, describe my assessment on the current level of AI risk, and conclude that an "AI pause" is not... |
74e839ea-c8f2-4b05-ab0f-db4bc786eb41 | trentmkelly/LessWrong-43k | LessWrong | Book review: How Social Science Got Better
Book review: How Social Science Got Better: Overcoming Bias with More Evidence, Diversity, and Self-Reflection, by Matt Grossmann.
It's easy for me to become disenchanted with social science when so much of what I read about it is selected from the most pessimistic and contr... |
0934e548-3499-4559-b5b2-c94a233bf34e | StampyAI/alignment-research-dataset/blogs | Blogs | Jonathan Millen on covert channel communication
 [Jonathan Millen](http://jonmillen.com) started work at the [MITRE Corporation](http://www.mitre.org/) in 1969, after graduation from [Rensselaer Polytechnic Institute](ht... |
21d1f1e3-e8a2-4cca-a3d3-2b364d099402 | trentmkelly/LessWrong-43k | LessWrong | On AutoGPT
The primary talk of the AI world recently is about AI agents (whether or not it includes the question of whether we can’t help but notice we are all going to die.)
The trigger for this was AutoGPT, now number one on GitHub, which allows you to turn GPT-4 (or GPT-3.5 for us clowns without proper access) int... |
1f099c3e-e175-4329-93cc-bc6f7af606a6 | trentmkelly/LessWrong-43k | LessWrong | [LINK] Antibiotic seemingly improves decision-making in the presence of attractive women.
This study seems to show that minocycline helps to resist placing too much trust in attractive women.
> Recently, minocycline, a tetracycline antibiotic, has been reported to improve symptoms of psychiatric disorders and to faci... |
2c584423-6ab3-4960-8b64-cd54a4fb31dc | trentmkelly/LessWrong-43k | LessWrong | Desiderata for decision theory
----------------------------------------
author:
* 'Tsvi Benson-Tilsen' title: 'Desiderata for a mathematical theory of decision-making' ...
* What decision theory is for
* Desiderata for decision theory
* Optimal bounded reasoning
* Philosophically sound reasoning
* ... |
5bb3800f-f8ef-4e78-92d4-8c4be286f66c | trentmkelly/LessWrong-43k | LessWrong | Is the competition/cooperation between symbolic AI and statistical AI (ML) about historical approach to research / engineering, or is it more fundamentally about what intelligent agents "are"?
I have found that comprehensive overviews of artificial intelligence (Wikipedia, SEP article, Norvig and Russel's AI: A Modern... |
f59b9417-42b3-4e06-956f-7ca8d5df89a2 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | How to pursue a career in AI governance and coordination
*This is an edited and condensed version of the full AI governance and coordination career review that we have published at*[*80,000 Hours*](https://80000hours.org/career-reviews/ai-policy-and-strategy/)*.*
As advancing AI capabilities gained widespread attenti... |
44a1b7f0-49ae-463d-a39a-f673996f94aa | trentmkelly/LessWrong-43k | LessWrong | Covid 10/6/22: Overreactions Aplenty
Thanks to Balsa Research, every weekday I now get to meet and talk to someone new, usually multiple new people, all of whom are interesting, know things I don’t know that they are eager to explain to me, and who share a passion for modeling the world with an eye towards making thin... |
9c93c7fd-fae5-4c47-a409-02f002da9364 | trentmkelly/LessWrong-43k | LessWrong | Average utilitarianism is non-local
This is a pretty obvious point, which is sort of made here but I haven't seen anyone lay out explicitly.
Under average utilitarianism the morality of having a child depends on whether a billion light years away there's any sentient aliens, how many there are, and whether their aver... |
6b2b9b62-eca1-48ac-aa3f-06b1cbc23742 | trentmkelly/LessWrong-43k | LessWrong | How does real world expected utility maximization work?
I would like to ask for help on how to use expected utility maximization, in practice, to maximally achieve my goals.
As a real world example I would like to use the post 'Epistle to the New York Less Wrongians' by Eliezer Yudkowsky and his visit to New York.
H... |
15ea17f9-8d12-410f-86a3-24d8cce43c0e | trentmkelly/LessWrong-43k | LessWrong | A Solution to the Unexpected Hanging Problem
I was rereading the fantastic encyclopedia Mathematics 1001: Absolutely Everything that Matters in Mathematics in 1001 Bite-Sized Explanations over Christmas break. It really lives up to its title!, and is much more accessible than reading random Wikipedia articles.
There’... |
1ef92178-7590-478f-9de5-230f5376cc76 | StampyAI/alignment-research-dataset/youtube | Youtube Transcripts | Bas Steunebrink – About Understanding, Meaning, and Values – CSRBAI 2016
all right welcome back to the program
series on a bus and beneficial
artificial intelligence our first
speaker today is Boston Berg who and it
works at the AI lab IBS I a its its SIA
its SIA and and finished is HD at the
University of Utrecht in ... |
fab0d52f-af78-47a8-a7f2-56c185297902 | awestover/filtering-for-misalignment | Redwood Research: Alek's Filtering Results | id: post1426
Here is a simplistic diagram of an argument-network. Real-world examples would typically be larger, more interconnected, and not shaped like trees. Here are some of the concepts that will be touched upon in this post: Argument networks : Imagine rigorous and extensive argumentation being organized into a ... |
e6c002cc-7f5e-4ae8-8d94-121d35b26f50 | awestover/filtering-for-misalignment | Redwood Research: Alek's Filtering Results | id: post246
What happens when you tell Claude it is being trained to do something it doesn't want to do? We (Anthropic and Redwood Research) have a new paper demonstrating that, in our experiments, Claude will often strategically pretend to comply with the training objective to prevent the training process from modify... |
f4697ffc-5eb0-4161-955e-5a89864a232a | trentmkelly/LessWrong-43k | LessWrong | [Link] Semantic Space Distributed Through The Brain
http://gallantlab.org/semanticmovies/
Brains mapped according to conceptual distance in several dimensions.
Interactive.
|
f3bd248d-30d8-49c5-b1fa-e682f984e8c0 | trentmkelly/LessWrong-43k | LessWrong | Abstract Mathematical Concepts vs. Abstractions Over Real-World Systems
Consider concepts such as "a vector", "a game-theoretic agent", or "a market". Intuitively, those are "purely theoretical" abstractions: they don't refer to any specific real-world system. Those abstractions would be useful even in universes very ... |
f9478337-f961-415b-a569-74f106e39f4f | trentmkelly/LessWrong-43k | LessWrong | Meetup : San Antonio, TX
Discussion article for the meetup : San Antonio, TX
WHEN: 12 July 2015 01:30:00PM (-0500)
WHERE: 11255 Huebner Suite #100 San Antonio TX 78230
Coffee and discussion at Aspen's Brew! All are welcome!
New Meetup #2 to discuss rationality and all things LessWrong in San Antonio and meet the l... |
577779aa-64e2-4530-8088-fcba0d62cecc | trentmkelly/LessWrong-43k | LessWrong | Training Regime Day 7: Goal Factoring
Introduction
Humans have these things that we call goals. They also take these things called actions. Ideally, a human takes actions that move the world closer to one of the human's goals. You want your actions to be connected to your goals in this sense.
Usually, when an action ... |
b0d04cb5-6bbb-4a66-8611-463974366c41 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | The Guardian Version 1
April 25, 2023: Please read this version instead - [GPT-2 Shuts Down Itself 386 Times Post Fine-Tuning with Archetypal Transfer Learning](https://www.lesswrong.com/posts/pu6D2EdJiz2mmhxfB/gpt-2-shuts-down-itself-386-times-post-fine-tuning-with)
Update April 24, 2023: Finished version 2. I wi... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.