~www_lesswrong_com | Bookmarks (669)

Published on November 8, 2024 11:47 PM GMTSummaryFour months after my post 'LLM Generality is a Timeline Crux', new research on o1-preview should update us significantly toward LLMs being capable of general reasoning, and hence of scaling straight to AGI, and shorten our timeline estimates.Summary of previous postIn June of 2024, I wrote a post, 'LLM Generality is a Timeline Crux', in which I...

Published on November 8, 2024 10:04 PM GMTLet's consider air purifier design a bit. preface This post is about a potential new category of air purifier, but existing air purifiers are already worth using. You should probably have at least one at your home. Some reasonable options are: the Levoit Vital 200S-P for larger rooms, and the Winix A230 for smaller rooms. Those are...

Published on November 8, 2024 9:50 PM GMTMy husband, Andrew Rettek, has a blog you should read. As he’s gotten into fitness, he’s started following exercise science, which is the (very new!) field of running small controlled experiments on diet and exercise on athletes who do exactly what you tell them to, under strict observation.This is in contrast to fields like nutrition science, which...

Published on November 8, 2024 9:20 PM GMTHey y'all! I just started a rationality group on the UChicago campus and wanted to post it here to advertise it to UChicago-affiliated LessWrong readers. We've had a couple meetings so far which have been great, and I'm excited for more! A few more things: (1) You can join the email list by submitting this form to be updated...

Published on November 7, 2024 6:15 PM GMTN.B. This is a chapter in a planned book about epistemology. Chapters are not necessarily released in order. If you read this, the most helpful comments would be on things you found confusing, things you felt were missing, threads that were hard to follow or seemed irrelevant, and otherwise mid to high level feedback about the content....

Published on November 7, 2024 4:30 PM GMTA lot happened in AI this week, but most people’s focus was very much elsewhere. I’ll start with what Trump might mean for AI policy, then move on to the rest. This is the future we have to live in, and potentially save. Back to work, as they say. Table of Contents Trump Card. What does Trump’s...

Published on November 7, 2024 4:06 PM GMTEpistemic status: This text presents a thought experiment suggested by James Miller, along with Alexey Turchin's musings on possible solutions. While our thoughts are largely aligned (we both accept high chances of quantum Immortality and the timeline selection principle), some ideas are more personal (e.g., Turchin's "transcendental advantage") in Part 2.TL;DR: If quantum immortality is true, I...

Published on November 7, 2024 3:39 PM GMTProduced as part of MATS 6.0 and 6.1.Key takeaways:Training LLMs on (simulated) user feedback can lead to the emergence of manipulative and deceptive behaviors.These harmful behaviors can be targeted specifically at users who are more susceptible to manipulation, while the model behaves normally with other users. This makes such behaviors challenging to detect.Standard model evaluations for sycophancy...

Published on November 7, 2024 3:26 PM GMTThere is some difference between despotic and cosmopolitan agents. A despotic agent builds a universe grabbing Singleton and makes it satisfy its own desires. A cosmopolitan agent builds a universe grabbing Singleton and makes it satisfy the desires of all the things which ought be happy. The cosmopolitan agent will burn some of its precious negentropy on...

Published on November 7, 2024 12:14 PM GMTI've been accepted as a mentor for the next AI Safety Camp. You can apply to work with me and the team. The deadline for applicants is November 17.The program will run from January 11 to April 27. SummaryCore underlying hypothesis - we believe that there is a significant agency overhang in the modern LLMs, meaning there is a potential...

Published on November 7, 2024 9:40 AM GMTI just read the wikipedia article on the evolution of human intelligence, and TBH I wasn't super impressed with the quality of the considerations there.I currently have 3 main (categories of) hypotheses for what caused selection pressure for intelligence in humans. (But please post an answer if you have other hypotheses that seem plausible!):("H" for "hypothesis")H1: social...

Published on November 7, 2024 5:27 AM GMTThis is an excerpt from the Introductions section to a book-length project that was kicked off as a response to the framing of the essay competition on the Automation of Wisdom and Philosophy. Many unrelated-seeming threads open in this post, that will come together by the end of the overall sequence.If you don't like abstractness, the first...

Published on November 7, 2024 5:22 AM GMTThis is an interim report sharing preliminary results. We hope this update will be useful to related research occurring in parallel.Executive SummaryProblem: Qwen1.5 0.5B Chat SAEs trained on the pile (webtext) fail to find sparse, interpretable reconstructions of the refusal direction from Arditi et al. The most refusal-related latent we find is coarse grained and underperforms the refusal direction at...

Published on November 6, 2024 11:01 PM GMTAs Americans know, the electoral college gives disproportionate influence to swing states, which means a vote in the extremely blue state of California was basically wasted in the 2024 election, as are votes in extremely red states like Texas, Oklahoma, and Louisiana. State legislatures have the Constitutional power to assign their state's electoral votes. So why don't...

Published on November 5, 2024 2:51 PM GMTNote: thank you to Brita Belli, senior communications manager at Recursion, for connecting me to Charles Baker, a VP at Recursion, who led a lot of this work I’ll discuss here + allowed me to interview him about it! I am not affiliated with Recursion in any capacity.One more note: a few people have pointed it out...

Published on November 5, 2024 11:37 AM GMTIn our jobs as AI safety researchers, we think a lot about what it means to have reasonable beliefs and to make good decisions. This matters because we want to understand how powerful AI systems might behave. It also matters because we ourselves need to know how to make good decisions in light of tremendous uncertainty about...

Published on November 5, 2024 7:07 AM GMTThe first two sections are below: Increasingly powerful AI systems have the potential to accelerate scientific progress, unlock new medical treatments, and grow the economy. But along with the remarkable new capabilities of these AIs come significant risks. Governments should urgently take action on AI policy in the next eighteen months. The window for proactive risk prevention...

Published on November 5, 2024 1:18 AM GMTIntroductionThis is a short summary of my experience attending the ML4Good UK bootcamp in September 2024. There are 2 previous experience reports I link to at the bottom, but because the program is refined each time, I wanted to describe my experience and add my two cents. This is useful for you if you are contemplating applying...

Published on November 5, 2024 5:07 AM GMTCross posting from my personal blog: https://spiralprogress.com/2024/10/28/the-shallow-bench/Spoilers for "Project Hail Mary", stop reading here if you don't want to be spoiled.Project Hail Mary follows Ryland Grace, a disgraced academic turned high school biology teacher who gets selected as part of a crew of three tasked with saving all of humanity from an impending alien threat.Not even in...

Published on November 5, 2024 1:01 AM GMTThis blog was published by Jonathan Ng, Andrey Anurin, Connor Axiotes, Esben Kran.Apart Research's newest paper, Catastrophic Cyber Capabilities Benchmark (3cb): Robustly Evaluating LLM Agent Cyber Offense Capabilities (website), creates a novel cyber offense capability benchmark that engages with issues of legibility, coverage, and generalization in cyber offense benchmarks.We were moved to create 3cb because a superintelligent AI...

Published on November 4, 2024 11:29 PM GMT(Btw everything I write here about orcas also applies to a slightly lesser extent to pilot whales (especially long finned ones)[1].)(I'm very very far from an orca expert - basically everything I know about them I learned today.)I always thought that bigger animals might have bigger brains than humans but not actually more neurons in their neocortex...

Published on November 4, 2024 10:50 PM GMTMidjourney, “metastatic cancer”Metastatic Cancer Is Usually DeadlyWhen my mom was diagnosed with cancer, it was already metastatic; her lymph nodes, said the nurse, “lit up like a Christmas tree.”Most people with any kind of metastatic cancer — that is, instead of a single tumor, a cancer that has spread to multiple locations in the body — die...

Published on November 4, 2024 10:31 PM GMTHello! I'm looking for community members to read speeches at the Bay Winter Solstice event this year. If you're interested, please email me by the end of November 17 with the speeches you're potentially interested in. I'll ask you to record yourself reading it (no need to be particularly polished).   I am currently auditioning for:Thanksgiving PrayerI Have...

Published on November 4, 2024 9:11 PM GMTThank you to Justis Millis for providing feedback and proofreading on this post. This post is also available on my Substack.TL/DR: Contrary to the theory that neurological sex differences and autism both involve the same tradeoff of systemic versus empathic thinking, I found complex differences. It turned out that men were more interested in technology and more...