PODCAST

LessWrong (30+ Karma)

LessWrong

Audio narrations of LessWrong posts.

technology philosophy

All Episodes

2:43

[Linkpost] “Introducing AI Lab Watch” by Zach...

LessWrong (30+ Karma) ·

2024/04/30

This is a linkpost for https://ailabwatch.orgI'm launching AI Lab Watch. I collected actions for frontier AI labs to improve AI safety, then evaluated some frontier labs accordingly.It's a collection of information on what labs should do and what labs are doing. It also has some adjacent resources, including a list of other safety-ish scorecard-ish stuff.(It's much better on desktop than mobile — don't read it on mobile.)It's in beta—leave feedback here or comment or DM me—but I basically endorse

en-gb

26:21

“Towards a formalization of the agent structure...

LessWrong (30+ Karma) ·

2024/04/30

In Clarifying the Agent-Like Structure Problem (2022), John Wentworth describes a hypothetical instance of what he calls a selection theorem. In Scott Garrabrant's words, the question is, does agent-like behavior imply agent-like architecture? That is, if we take some class of behaving things and apply a filter for agent-like behavior, do we end up selecting things with agent-like architecture (or structure)? Of course, this question is heavily under-specified. So another way to ask this is, under

en-gb

23:55

“Towards Multimodal Interpretability: Learning Sparse...

LessWrong (30+ Karma) ·

2024/04/30

Two Minute SummaryIn this post I present my results from training a Sparse Autoencoder (SAE) on a CLIP Vision Transformer (ViT) using the ImageNet-1k dataset. I have created an interactive web app, 'SAE Explorer', to allow the public to explore the visual features the SAE has learnt, found here: https://sae-explorer.streamlit.app/ (best viewed on a laptop). My results illustrate that SAEs can identify sparse and highly interpretable directions in the residual stream of vision models, enabling

en-gb

18:59

“Ironing Out the Squiggles” by Zack_M_Davis

LessWrong (30+ Karma) ·

2024/04/29

Adversarial Examples: A Problem The apparent successes of the deep learning revolution conceal a dark underbelly. It may seem that we now know how to get computers to (say) check whether a photo is of a bird, but this façade of seemingly good performance is belied by the existence of adversarial examples—specially prepared data that looks ordinary to humans, but is seen radically differently by machine learning models. The differentiable nature of neural networks, which make them possible to be

en-gb

3:20

“List your AI X-Risk cruxes!” by Aryeh Englander

LessWrong (30+ Karma) ·

2024/04/29

[I'm posting this as a very informal community request in lieu of a more detailed writeup, because if I wait to do this in a much more careful fashion then it probably won't happen at all. If someone else wants to do a more careful version that would be great!] By crux I mean some uncertainty you have such that your estimate for the likelihood of existential risk from AI - your "p(doom)" if you like that term -might shift significantly if that uncertainty were resolved. More precisely, let's define

en-gb

4:09

“AISC9 has ended and there will be an AISC10” by...

LessWrong (30+ Karma) ·

2024/04/29

Crossposted from the AI Alignment Forum. May contain more technical jargon than usual.The 9th AI Safety Camp (AISC9) just ended, and as usual, it was a success! Follow this link to find project summaries, links to their outputs, recordings to the end of camp presentations and contact info to all our teams in case you want to engage more.AISC9 both had the largest number of participants (159) and the smallest number of staff (2) of all the camps we’ve done so far. Me and Remmelt have proven that if

en-gb

23:50

“[Aspiration-based designs] 1. Informal introduction”...

LessWrong (30+ Karma) ·

2024/04/29

Sequence Summary. This sequence documents research by SatisfIA, an ongoing project on non-maximizing, aspiration-based designs for AI agents that fulfill goals specified by constraints ("aspirations") rather than maximizing an objective function. We aim to contribute to AI safety by exploring design approaches and their software implementations that we believe might be promising but neglected or novel. Our approach is roughly related to but largely complementary to concepts like quantilization and

en-gb

27:49

“Constructability: Plainly-coded AGIs may be feasible...

LessWrong (30+ Karma) ·

2024/04/28

Charbel-Raphaël Segerie and Épiphanie Gédéon contributed equally to this post. Many thanks to Davidad, Gabriel Alfour, Jérémy Andréoletti, Lucie Philippon, Vladimir Ivanov, Alexandre Variengien, Angélina Gentaz, Léo Dana and Diego Dorn for useful feedback.TLDR: We present a new method for a safer-by design AI development. We think using plainly coded AIs may be feasible in the near future and may be safe. We also present a prototype and research ideas.Epistemic status: Armchair reasoning style. We

en-gb

5:07

“We are headed into an extreme compute overhang” by...

LessWrong (30+ Karma) ·

2024/04/28

If we achieve AGI-level performance using an LLM-like approach, the training hardware will be capable of running ~1,000,000s concurrent instances of the model. DefinitionsAlthough there is some debate about the definition of compute overhang, I believe that the AI Impacts definition matches the original use, and I prefer it: "enough computing hardware to run many powerful AI systems already exists by the time the software to run such systems is developed". A large compute overhang leads to

en-gb

13:09

“So What’s Up With PUFAs Chemically?” by J Bostock

LessWrong (30+ Karma) ·

2024/04/27

This is exploratory investigation of a new-ish hypothesis, it is not intended to be a comprehensive review of the field or even a a full investigation of the hypothesis.I've always been skeptical of the seed-oil theory of obesity. Perhaps this is bad rationality on my part, but I've tended to retreat to the sniff test on issues as charged and confusing as diet. My response to the general seed-oil theory was basically "Really? Seeds and nuts? The things you just find growing on plants, and that our

en-gb

17:08

“Refusal in LLMs is mediated by a single direction”...

LessWrong (30+ Karma) ·

2024/04/27

Crossposted from the AI Alignment Forum. May contain more technical jargon than usual.This work was produced as part of Neel Nanda's stream in the ML Alignment & Theory Scholars Program - Winter 2023-24 Cohort, with co-supervision from Wes Gurnee.This post is a preview for our upcoming paper, which will provide more detail into our current understanding of refusal.We thank Nina Rimsky and Daniel Paleka for the helpful conversations and review. Executive summaryModern LLMs are typically

en-gb

6:52

“D&D.Sci Long War: Defender of Data-mocracy” by aphyer

LessWrong (30+ Karma) ·

2024/04/27

This is an entry in the 'Dungeons & Data Science' series, a set of puzzles where players are given a dataset to analyze and an objective to pursue using information from that dataset. STORY (skippable)You have the excellent fortune to live under the governance of The People's Glorious Free Democratic Republic of Earth, giving you a Glorious life of Freedom and Democracy.Sadly, your cherished values of Democracy and Freedom are under attack by...THE ALIEN MENACE!The typical reaction of an Alien

en-gb

26:22

“Superposition is not ‘just’ neuron polysemanticity”...

LessWrong (30+ Karma) ·

2024/04/27

Crossposted from the AI Alignment Forum. May contain more technical jargon than usual.TL;DR: In this post, I distinguish between two related concepts in neural network interpretability: polysemanticity and superposition. Neuron polysemanticity is the observed phenomena that many neurons seem to fire (have large, positive activations) on multiple unrelated concepts. Superposition is a specific explanation for neuron (or attention head) polysemanticity, where a neural network represents more sparse

en-gb

14:30

“On Not Pulling The Ladder Up Behind You” by Screwtape

LessWrong (30+ Karma) ·

2024/04/27

Epistemic Status: Musing and speculation, but I think there's a real thing here. 1.When I was a kid, a friend of mine had a tree fort. If you've never seen such a fort, imagine a series of wooden boards secured to a tree, creating a platform about fifteen feet off the ground where you can sit or stand and walk around the tree. This one had a rope ladder we used to get up and down, a length of knotted rope that was tied to the tree at the top and dangled over the edge so that it reached the ground

en-gb

8:22

“Duct Tape security” by Isaac King

LessWrong (30+ Karma) ·

2024/04/26

This is a linkpost for On Duct Tape and Fence Posts.Eliezer writes about fence post security. When people think to themselves "in the current system, what's the weakest point?", and then dedicate their resources to shoring up the defenses at that point, not realizing that after the first small improvement in that area, there's likely now a new weakest point somewhere else. Fence post security happens preemptively, when the designers of the system fixate on the most salient aspect(s) and don't

en-gb

7:39

“Scaling of AI training runs will slow down after...

LessWrong (30+ Karma) ·

2024/04/26

My credence: 33% confidence in the claim that the growth in the number of GPUs used for training SOTA AI will slow down significantly directly after GPT-5. It is not higher because of (1) decentralized training is possible, and (2) GPT-5 may be able to increase hardware efficiency significantly, (3) GPT-5 may be smaller than assumed in this post, (4) race dynamics.TLDR: Because of a bottleneck in energy access to data centers and the need to build OOM larger data centers. The reasoning behind the

en-gb

15:19

“Spatial attention as a ‘tell’ for empathetic...

LessWrong (30+ Karma) ·

2024/04/26

(Half-baked work-in-progress. There might be a “version 2” of this post at some point, with fewer mistakes, and more neuroscience details, and nice illustrations and pedagogy etc. But it's fun to chat and see if anyone has thoughts.) 1. BackgroundThere's a neuroscience problem that's had me stumped since almost the very beginning of when I became interested in neuroscience at all (as a lens into AGI safety) back in 2019. But I think I might finally have “a foot in the door” towards a solution!What

en-gb

8:51

“Losing Faith In Contrarianism” by omnizoid

LessWrong (30+ Karma) ·

2024/04/26

Crosspost from my blog. If you spend a lot of time in the blogosphere, you’ll find a great deal of people expressing contrarian views. If you hang out in the circles that I do, you’ll probably have heard of Yudkowsky say that dieting doesn’t really work, Guzey say that sleep is overrated, Hanson argue that medicine doesn’t improve health, various people argue for the lab leak, others argue for hereditarianism, Caplan argue that mental illness is mostly just aberrant preferences and education doesn’t

en-gb

12:15

[Linkpost] “LLMs seem (relatively) safe” by JustisMills

LessWrong (30+ Karma) ·

2024/04/26

This is a linkpost for https://justismills.substack.com/p/llms-seem-relatively-safePost for a somewhat more general audience than the modal LessWrong reader, but gets at my actual thoughts on the topic.In 2018 OpenAI defeated the world champions of Dota 2, a major esports game. This was hot on the heels of DeepMind's AlphaGo performance against Lee Sedol in 2016, achieving superhuman Go performance way before anyone thought that might happen. AI benchmarks were being cleared at a pace which felt

en-gb

1:35

[Linkpost] “Improving Dictionary Learning with Gated...

LessWrong (30+ Karma) ·

2024/04/25

Crossposted from the AI Alignment Forum. May contain more technical jargon than usual.This is a linkpost for https://arxiv.org/abs/2404.16014Authors: Senthooran Rajamanoharan*, Arthur Conmy*, Lewis Smith, Tom Lieberum, Vikrant Varma, János Kramár, Rohin Shah, Neel NandaA new paper from the Google DeepMind mech interp team: Improving Dictionary Learning with Gated Sparse Autoencoders! Gated SAEs are a new Sparse Autoencoder architecture that seems to be a significant Pareto-improvement over normal

en-gb

4,491 results

LessWrong (30+ Karma)

All Episodes

Similar Podcasts

Lead With AI

Tamara Nall

Reflections

Livewire Labs

ForGeeks

ForGeeks

FOX on Tech

FOX News Podcasts

The Linux Cast

The Linux Cast